diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,176542 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 50.0,
+  "eval_steps": 500,
+  "global_step": 25000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002,
+      "grad_norm": 100.7913589477539,
+      "learning_rate": 2e-05,
+      "loss": 3.10091186,
+      "step": 1
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 40.7878532409668,
+      "learning_rate": 2e-05,
+      "loss": 1.36577916,
+      "step": 2
+    },
+    {
+      "epoch": 0.006,
+      "grad_norm": 17.934667587280273,
+      "learning_rate": 2e-05,
+      "loss": 0.80921072,
+      "step": 3
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 29.937301635742188,
+      "learning_rate": 2e-05,
+      "loss": 0.85672188,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.57142448425293,
+      "learning_rate": 2e-05,
+      "loss": 0.69120318,
+      "step": 5
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 7.896178245544434,
+      "learning_rate": 2e-05,
+      "loss": 0.47293058,
+      "step": 6
+    },
+    {
+      "epoch": 0.014,
+      "grad_norm": 8.591035842895508,
+      "learning_rate": 2e-05,
+      "loss": 0.50240922,
+      "step": 7
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 10.887709617614746,
+      "learning_rate": 2e-05,
+      "loss": 0.51509768,
+      "step": 8
+    },
+    {
+      "epoch": 0.018,
+      "grad_norm": 134.73768615722656,
+      "learning_rate": 2e-05,
+      "loss": 0.49554652,
+      "step": 9
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 55.33390426635742,
+      "learning_rate": 2e-05,
+      "loss": 0.62111574,
+      "step": 10
+    },
+    {
+      "epoch": 0.022,
+      "grad_norm": 12.508940696716309,
+      "learning_rate": 2e-05,
+      "loss": 0.43727028,
+      "step": 11
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 8.327451705932617,
+      "learning_rate": 2e-05,
+      "loss": 0.4894059,
+      "step": 12
+    },
+    {
+      "epoch": 0.026,
+      "grad_norm": 4.562747001647949,
+      "learning_rate": 2e-05,
+      "loss": 0.42299265,
+      "step": 13
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 5.968645095825195,
+      "learning_rate": 2e-05,
+      "loss": 0.3810643,
+      "step": 14
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.443109035491943,
+      "learning_rate": 2e-05,
+      "loss": 0.41381907,
+      "step": 15
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 3.4299163818359375,
+      "learning_rate": 2e-05,
+      "loss": 0.42420715,
+      "step": 16
+    },
+    {
+      "epoch": 0.034,
+      "grad_norm": 4.944918155670166,
+      "learning_rate": 2e-05,
+      "loss": 0.46888059,
+      "step": 17
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 4.6365180015563965,
+      "learning_rate": 2e-05,
+      "loss": 0.45326883,
+      "step": 18
+    },
+    {
+      "epoch": 0.038,
+      "grad_norm": 2.8536622524261475,
+      "learning_rate": 2e-05,
+      "loss": 0.44977522,
+      "step": 19
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.375406265258789,
+      "learning_rate": 2e-05,
+      "loss": 0.44011864,
+      "step": 20
+    },
+    {
+      "epoch": 0.042,
+      "grad_norm": 3.37801456451416,
+      "learning_rate": 2e-05,
+      "loss": 0.40136099,
+      "step": 21
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 3.705557346343994,
+      "learning_rate": 2e-05,
+      "loss": 0.37761164,
+      "step": 22
+    },
+    {
+      "epoch": 0.046,
+      "grad_norm": 3.8013367652893066,
+      "learning_rate": 2e-05,
+      "loss": 0.44508934,
+      "step": 23
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 3.566617250442505,
+      "learning_rate": 2e-05,
+      "loss": 0.38218161,
+      "step": 24
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.859208345413208,
+      "learning_rate": 2e-05,
+      "loss": 0.3878693,
+      "step": 25
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 3.2768349647521973,
+      "learning_rate": 2e-05,
+      "loss": 0.37782881,
+      "step": 26
+    },
+    {
+      "epoch": 0.054,
+      "grad_norm": 2.3881168365478516,
+      "learning_rate": 2e-05,
+      "loss": 0.42464909,
+      "step": 27
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 2.54347562789917,
+      "learning_rate": 2e-05,
+      "loss": 0.39481583,
+      "step": 28
+    },
+    {
+      "epoch": 0.058,
+      "grad_norm": 7.55833101272583,
+      "learning_rate": 2e-05,
+      "loss": 0.42484623,
+      "step": 29
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.715081214904785,
+      "learning_rate": 2e-05,
+      "loss": 0.46655717,
+      "step": 30
+    },
+    {
+      "epoch": 0.062,
+      "grad_norm": 2.705334424972534,
+      "learning_rate": 2e-05,
+      "loss": 0.38881761,
+      "step": 31
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 3.3475492000579834,
+      "learning_rate": 2e-05,
+      "loss": 0.42681402,
+      "step": 32
+    },
+    {
+      "epoch": 0.066,
+      "grad_norm": 3.141242504119873,
+      "learning_rate": 2e-05,
+      "loss": 0.42609936,
+      "step": 33
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 2.4876580238342285,
+      "learning_rate": 2e-05,
+      "loss": 0.37850282,
+      "step": 34
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4182939529418945,
+      "learning_rate": 2e-05,
+      "loss": 0.377267,
+      "step": 35
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 3.404860258102417,
+      "learning_rate": 2e-05,
+      "loss": 0.41409212,
+      "step": 36
+    },
+    {
+      "epoch": 0.074,
+      "grad_norm": 4.462337970733643,
+      "learning_rate": 2e-05,
+      "loss": 0.41622919,
+      "step": 37
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 3.895780563354492,
+      "learning_rate": 2e-05,
+      "loss": 0.40852085,
+      "step": 38
+    },
+    {
+      "epoch": 0.078,
+      "grad_norm": 4.350463390350342,
+      "learning_rate": 2e-05,
+      "loss": 0.44786674,
+      "step": 39
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5350425243377686,
+      "learning_rate": 2e-05,
+      "loss": 0.40965152,
+      "step": 40
+    },
+    {
+      "epoch": 0.082,
+      "grad_norm": 5.017084121704102,
+      "learning_rate": 2e-05,
+      "loss": 0.45135465,
+      "step": 41
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 3.472752332687378,
+      "learning_rate": 2e-05,
+      "loss": 0.42319882,
+      "step": 42
+    },
+    {
+      "epoch": 0.086,
+      "grad_norm": 3.5228331089019775,
+      "learning_rate": 2e-05,
+      "loss": 0.31802225,
+      "step": 43
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 4.688898086547852,
+      "learning_rate": 2e-05,
+      "loss": 0.46733904,
+      "step": 44
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.554978370666504,
+      "learning_rate": 2e-05,
+      "loss": 0.35162151,
+      "step": 45
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 3.005284070968628,
+      "learning_rate": 2e-05,
+      "loss": 0.40723544,
+      "step": 46
+    },
+    {
+      "epoch": 0.094,
+      "grad_norm": 3.2923007011413574,
+      "learning_rate": 2e-05,
+      "loss": 0.35407171,
+      "step": 47
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 7.087942600250244,
+      "learning_rate": 2e-05,
+      "loss": 0.43687314,
+      "step": 48
+    },
+    {
+      "epoch": 0.098,
+      "grad_norm": 3.748598098754883,
+      "learning_rate": 2e-05,
+      "loss": 0.40476981,
+      "step": 49
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1440889835357666,
+      "learning_rate": 2e-05,
+      "loss": 0.34471345,
+      "step": 50
+    },
+    {
+      "epoch": 0.102,
+      "grad_norm": 3.0694971084594727,
+      "learning_rate": 2e-05,
+      "loss": 0.39447641,
+      "step": 51
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 2.8853402137756348,
+      "learning_rate": 2e-05,
+      "loss": 0.4363516,
+      "step": 52
+    },
+    {
+      "epoch": 0.106,
+      "grad_norm": 2.3027663230895996,
+      "learning_rate": 2e-05,
+      "loss": 0.42112815,
+      "step": 53
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 3.492051124572754,
+      "learning_rate": 2e-05,
+      "loss": 0.36403522,
+      "step": 54
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2932403087615967,
+      "learning_rate": 2e-05,
+      "loss": 0.37265083,
+      "step": 55
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 2.5150811672210693,
+      "learning_rate": 2e-05,
+      "loss": 0.35803241,
+      "step": 56
+    },
+    {
+      "epoch": 0.114,
+      "grad_norm": 2.370377540588379,
+      "learning_rate": 2e-05,
+      "loss": 0.4209469,
+      "step": 57
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 2.392200469970703,
+      "learning_rate": 2e-05,
+      "loss": 0.44664249,
+      "step": 58
+    },
+    {
+      "epoch": 0.118,
+      "grad_norm": 2.7458958625793457,
+      "learning_rate": 2e-05,
+      "loss": 0.41989598,
+      "step": 59
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5858850479125977,
+      "learning_rate": 2e-05,
+      "loss": 0.4336924,
+      "step": 60
+    },
+    {
+      "epoch": 0.122,
+      "grad_norm": 3.472402334213257,
+      "learning_rate": 2e-05,
+      "loss": 0.45059329,
+      "step": 61
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 2.9056155681610107,
+      "learning_rate": 2e-05,
+      "loss": 0.43756476,
+      "step": 62
+    },
+    {
+      "epoch": 0.126,
+      "grad_norm": 2.1384313106536865,
+      "learning_rate": 2e-05,
+      "loss": 0.37061867,
+      "step": 63
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 2.247955560684204,
+      "learning_rate": 2e-05,
+      "loss": 0.4254021,
+      "step": 64
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1384289264678955,
+      "learning_rate": 2e-05,
+      "loss": 0.48214421,
+      "step": 65
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 3.4709699153900146,
+      "learning_rate": 2e-05,
+      "loss": 0.42113096,
+      "step": 66
+    },
+    {
+      "epoch": 0.134,
+      "grad_norm": 2.4829514026641846,
+      "learning_rate": 2e-05,
+      "loss": 0.3783536,
+      "step": 67
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 3.276494264602661,
+      "learning_rate": 2e-05,
+      "loss": 0.42203033,
+      "step": 68
+    },
+    {
+      "epoch": 0.138,
+      "grad_norm": 2.3716046810150146,
+      "learning_rate": 2e-05,
+      "loss": 0.37840182,
+      "step": 69
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6660404205322266,
+      "learning_rate": 2e-05,
+      "loss": 0.3909995,
+      "step": 70
+    },
+    {
+      "epoch": 0.142,
+      "grad_norm": 2.4495251178741455,
+      "learning_rate": 2e-05,
+      "loss": 0.39900306,
+      "step": 71
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 2.6454856395721436,
+      "learning_rate": 2e-05,
+      "loss": 0.40227687,
+      "step": 72
+    },
+    {
+      "epoch": 0.146,
+      "grad_norm": 3.020829916000366,
+      "learning_rate": 2e-05,
+      "loss": 0.43715149,
+      "step": 73
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 2.465714454650879,
+      "learning_rate": 2e-05,
+      "loss": 0.43334347,
+      "step": 74
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1749234199523926,
+      "learning_rate": 2e-05,
+      "loss": 0.39349914,
+      "step": 75
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 2.522357225418091,
+      "learning_rate": 2e-05,
+      "loss": 0.40576932,
+      "step": 76
+    },
+    {
+      "epoch": 0.154,
+      "grad_norm": 2.7155542373657227,
+      "learning_rate": 2e-05,
+      "loss": 0.45305899,
+      "step": 77
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 2.727293014526367,
+      "learning_rate": 2e-05,
+      "loss": 0.44109389,
+      "step": 78
+    },
+    {
+      "epoch": 0.158,
+      "grad_norm": 2.575566291809082,
+      "learning_rate": 2e-05,
+      "loss": 0.44445091,
+      "step": 79
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.941967487335205,
+      "learning_rate": 2e-05,
+      "loss": 0.41291177,
+      "step": 80
+    },
+    {
+      "epoch": 0.162,
+      "grad_norm": 2.702101707458496,
+      "learning_rate": 2e-05,
+      "loss": 0.3560887,
+      "step": 81
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 2.959989547729492,
+      "learning_rate": 2e-05,
+      "loss": 0.3968443,
+      "step": 82
+    },
+    {
+      "epoch": 0.166,
+      "grad_norm": 3.733569860458374,
+      "learning_rate": 2e-05,
+      "loss": 0.4232977,
+      "step": 83
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 2.6571216583251953,
+      "learning_rate": 2e-05,
+      "loss": 0.40444511,
+      "step": 84
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2582576274871826,
+      "learning_rate": 2e-05,
+      "loss": 0.40064648,
+      "step": 85
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 2.748850107192993,
+      "learning_rate": 2e-05,
+      "loss": 0.38729197,
+      "step": 86
+    },
+    {
+      "epoch": 0.174,
+      "grad_norm": 2.8563621044158936,
+      "learning_rate": 2e-05,
+      "loss": 0.41891837,
+      "step": 87
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 2.7641937732696533,
+      "learning_rate": 2e-05,
+      "loss": 0.36975527,
+      "step": 88
+    },
+    {
+      "epoch": 0.178,
+      "grad_norm": 3.3558907508850098,
+      "learning_rate": 2e-05,
+      "loss": 0.40827322,
+      "step": 89
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4098260402679443,
+      "learning_rate": 2e-05,
+      "loss": 0.3943249,
+      "step": 90
+    },
+    {
+      "epoch": 0.182,
+      "grad_norm": 2.5855796337127686,
+      "learning_rate": 2e-05,
+      "loss": 0.43265003,
+      "step": 91
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 3.2219607830047607,
+      "learning_rate": 2e-05,
+      "loss": 0.44435838,
+      "step": 92
+    },
+    {
+      "epoch": 0.186,
+      "grad_norm": 2.994154453277588,
+      "learning_rate": 2e-05,
+      "loss": 0.43490282,
+      "step": 93
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 3.057532548904419,
+      "learning_rate": 2e-05,
+      "loss": 0.42345771,
+      "step": 94
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1149864196777344,
+      "learning_rate": 2e-05,
+      "loss": 0.40705955,
+      "step": 95
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 3.158325672149658,
+      "learning_rate": 2e-05,
+      "loss": 0.46287918,
+      "step": 96
+    },
+    {
+      "epoch": 0.194,
+      "grad_norm": 2.7815473079681396,
+      "learning_rate": 2e-05,
+      "loss": 0.39676839,
+      "step": 97
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 2.1724812984466553,
+      "learning_rate": 2e-05,
+      "loss": 0.38517368,
+      "step": 98
+    },
+    {
+      "epoch": 0.198,
+      "grad_norm": 2.216989040374756,
+      "learning_rate": 2e-05,
+      "loss": 0.41994014,
+      "step": 99
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4955761432647705,
+      "learning_rate": 2e-05,
+      "loss": 0.44885725,
+      "step": 100
+    },
+    {
+      "epoch": 0.202,
+      "grad_norm": 2.3539469242095947,
+      "learning_rate": 2e-05,
+      "loss": 0.39006793,
+      "step": 101
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 2.2714924812316895,
+      "learning_rate": 2e-05,
+      "loss": 0.34156144,
+      "step": 102
+    },
+    {
+      "epoch": 0.206,
+      "grad_norm": 2.3460693359375,
+      "learning_rate": 2e-05,
+      "loss": 0.43049014,
+      "step": 103
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 3.0970299243927,
+      "learning_rate": 2e-05,
+      "loss": 0.42005956,
+      "step": 104
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.211514711380005,
+      "learning_rate": 2e-05,
+      "loss": 0.38698167,
+      "step": 105
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 2.8566031455993652,
+      "learning_rate": 2e-05,
+      "loss": 0.39388683,
+      "step": 106
+    },
+    {
+      "epoch": 0.214,
+      "grad_norm": 4.965332508087158,
+      "learning_rate": 2e-05,
+      "loss": 0.32786798,
+      "step": 107
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 2.8506662845611572,
+      "learning_rate": 2e-05,
+      "loss": 0.39046088,
+      "step": 108
+    },
+    {
+      "epoch": 0.218,
+      "grad_norm": 2.1420042514801025,
+      "learning_rate": 2e-05,
+      "loss": 0.39321035,
+      "step": 109
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7388463020324707,
+      "learning_rate": 2e-05,
+      "loss": 0.39163101,
+      "step": 110
+    },
+    {
+      "epoch": 0.222,
+      "grad_norm": 2.2771174907684326,
+      "learning_rate": 2e-05,
+      "loss": 0.41349089,
+      "step": 111
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 4.137855529785156,
+      "learning_rate": 2e-05,
+      "loss": 0.37277463,
+      "step": 112
+    },
+    {
+      "epoch": 0.226,
+      "grad_norm": 2.8623104095458984,
+      "learning_rate": 2e-05,
+      "loss": 0.38881728,
+      "step": 113
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 2.295194387435913,
+      "learning_rate": 2e-05,
+      "loss": 0.3589763,
+      "step": 114
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1582581996917725,
+      "learning_rate": 2e-05,
+      "loss": 0.42957473,
+      "step": 115
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 3.635145664215088,
+      "learning_rate": 2e-05,
+      "loss": 0.39942896,
+      "step": 116
+    },
+    {
+      "epoch": 0.234,
+      "grad_norm": 3.435088872909546,
+      "learning_rate": 2e-05,
+      "loss": 0.41030282,
+      "step": 117
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 2.8914260864257812,
+      "learning_rate": 2e-05,
+      "loss": 0.38662139,
+      "step": 118
+    },
+    {
+      "epoch": 0.238,
+      "grad_norm": 3.000392198562622,
+      "learning_rate": 2e-05,
+      "loss": 0.42550755,
+      "step": 119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0846173763275146,
+      "learning_rate": 2e-05,
+      "loss": 0.3794331,
+      "step": 120
+    },
+    {
+      "epoch": 0.242,
+      "grad_norm": 2.7421340942382812,
+      "learning_rate": 2e-05,
+      "loss": 0.40051925,
+      "step": 121
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 2.206716537475586,
+      "learning_rate": 2e-05,
+      "loss": 0.41522628,
+      "step": 122
+    },
+    {
+      "epoch": 0.246,
+      "grad_norm": 2.345612049102783,
+      "learning_rate": 2e-05,
+      "loss": 0.39287353,
+      "step": 123
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 2.5863916873931885,
+      "learning_rate": 2e-05,
+      "loss": 0.44706243,
+      "step": 124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4030227661132812,
+      "learning_rate": 2e-05,
+      "loss": 0.39214543,
+      "step": 125
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 2.58503794670105,
+      "learning_rate": 2e-05,
+      "loss": 0.40951967,
+      "step": 126
+    },
+    {
+      "epoch": 0.254,
+      "grad_norm": 2.058553457260132,
+      "learning_rate": 2e-05,
+      "loss": 0.40562445,
+      "step": 127
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 2.324352979660034,
+      "learning_rate": 2e-05,
+      "loss": 0.39885789,
+      "step": 128
+    },
+    {
+      "epoch": 0.258,
+      "grad_norm": 2.4435112476348877,
+      "learning_rate": 2e-05,
+      "loss": 0.4411369,
+      "step": 129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4042422771453857,
+      "learning_rate": 2e-05,
+      "loss": 0.42098755,
+      "step": 130
+    },
+    {
+      "epoch": 0.262,
+      "grad_norm": 2.854393482208252,
+      "learning_rate": 2e-05,
+      "loss": 0.40045485,
+      "step": 131
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 2.2807977199554443,
+      "learning_rate": 2e-05,
+      "loss": 0.44947705,
+      "step": 132
+    },
+    {
+      "epoch": 0.266,
+      "grad_norm": 2.3839499950408936,
+      "learning_rate": 2e-05,
+      "loss": 0.3765958,
+      "step": 133
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 2.462369203567505,
+      "learning_rate": 2e-05,
+      "loss": 0.34250346,
+      "step": 134
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4267525672912598,
+      "learning_rate": 2e-05,
+      "loss": 0.4025071,
+      "step": 135
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 2.5603864192962646,
+      "learning_rate": 2e-05,
+      "loss": 0.35927606,
+      "step": 136
+    },
+    {
+      "epoch": 0.274,
+      "grad_norm": 3.1467013359069824,
+      "learning_rate": 2e-05,
+      "loss": 0.42534155,
+      "step": 137
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 2.4989264011383057,
+      "learning_rate": 2e-05,
+      "loss": 0.38834065,
+      "step": 138
+    },
+    {
+      "epoch": 0.278,
+      "grad_norm": 5.77341890335083,
+      "learning_rate": 2e-05,
+      "loss": 0.3826952,
+      "step": 139
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.720933675765991,
+      "learning_rate": 2e-05,
+      "loss": 0.39199528,
+      "step": 140
+    },
+    {
+      "epoch": 0.282,
+      "grad_norm": 6.706570625305176,
+      "learning_rate": 2e-05,
+      "loss": 0.43876618,
+      "step": 141
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 22.753585815429688,
+      "learning_rate": 2e-05,
+      "loss": 0.58754456,
+      "step": 142
+    },
+    {
+      "epoch": 0.286,
+      "grad_norm": 951.1945190429688,
+      "learning_rate": 2e-05,
+      "loss": 0.81678319,
+      "step": 143
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 52.611961364746094,
+      "learning_rate": 2e-05,
+      "loss": 0.69098788,
+      "step": 144
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 45.12958526611328,
+      "learning_rate": 2e-05,
+      "loss": 0.76437581,
+      "step": 145
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 33.77717971801758,
+      "learning_rate": 2e-05,
+      "loss": 0.75502336,
+      "step": 146
+    },
+    {
+      "epoch": 0.294,
+      "grad_norm": 78.94849395751953,
+      "learning_rate": 2e-05,
+      "loss": 0.98386419,
+      "step": 147
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 99.15904998779297,
+      "learning_rate": 2e-05,
+      "loss": 0.7885493,
+      "step": 148
+    },
+    {
+      "epoch": 0.298,
+      "grad_norm": 95.97003173828125,
+      "learning_rate": 2e-05,
+      "loss": 1.16757083,
+      "step": 149
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 395.0550537109375,
+      "learning_rate": 2e-05,
+      "loss": 0.89327443,
+      "step": 150
+    },
+    {
+      "epoch": 0.302,
+      "grad_norm": 34.62553787231445,
+      "learning_rate": 2e-05,
+      "loss": 0.80242836,
+      "step": 151
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 49.44704818725586,
+      "learning_rate": 2e-05,
+      "loss": 0.63106126,
+      "step": 152
+    },
+    {
+      "epoch": 0.306,
+      "grad_norm": 171.4626007080078,
+      "learning_rate": 2e-05,
+      "loss": 0.57824063,
+      "step": 153
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 93.4944839477539,
+      "learning_rate": 2e-05,
+      "loss": 0.52180779,
+      "step": 154
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 112.07191467285156,
+      "learning_rate": 2e-05,
+      "loss": 0.57057035,
+      "step": 155
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 23.541261672973633,
+      "learning_rate": 2e-05,
+      "loss": 0.6415171,
+      "step": 156
+    },
+    {
+      "epoch": 0.314,
+      "grad_norm": 121.93998718261719,
+      "learning_rate": 2e-05,
+      "loss": 0.81909418,
+      "step": 157
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 34.9930305480957,
+      "learning_rate": 2e-05,
+      "loss": 0.68764293,
+      "step": 158
+    },
+    {
+      "epoch": 0.318,
+      "grad_norm": 38.409847259521484,
+      "learning_rate": 2e-05,
+      "loss": 0.52681422,
+      "step": 159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 23.83429718017578,
+      "learning_rate": 2e-05,
+      "loss": 0.57993436,
+      "step": 160
+    },
+    {
+      "epoch": 0.322,
+      "grad_norm": 23.568622589111328,
+      "learning_rate": 2e-05,
+      "loss": 0.51327848,
+      "step": 161
+    },
+    {
+      "epoch": 0.324,
+      "grad_norm": 22.358970642089844,
+      "learning_rate": 2e-05,
+      "loss": 0.4694868,
+      "step": 162
+    },
+    {
+      "epoch": 0.326,
+      "grad_norm": 37.23964309692383,
+      "learning_rate": 2e-05,
+      "loss": 0.52104098,
+      "step": 163
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 27.70103645324707,
+      "learning_rate": 2e-05,
+      "loss": 0.5051856,
+      "step": 164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 19.599693298339844,
+      "learning_rate": 2e-05,
+      "loss": 0.48360169,
+      "step": 165
+    },
+    {
+      "epoch": 0.332,
+      "grad_norm": 17.194791793823242,
+      "learning_rate": 2e-05,
+      "loss": 0.50848162,
+      "step": 166
+    },
+    {
+      "epoch": 0.334,
+      "grad_norm": 27.718290328979492,
+      "learning_rate": 2e-05,
+      "loss": 0.46776086,
+      "step": 167
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 29.439424514770508,
+      "learning_rate": 2e-05,
+      "loss": 0.48417586,
+      "step": 168
+    },
+    {
+      "epoch": 0.338,
+      "grad_norm": 13.250664710998535,
+      "learning_rate": 2e-05,
+      "loss": 0.47358498,
+      "step": 169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 33.282222747802734,
+      "learning_rate": 2e-05,
+      "loss": 0.52936995,
+      "step": 170
+    },
+    {
+      "epoch": 0.342,
+      "grad_norm": 30.323564529418945,
+      "learning_rate": 2e-05,
+      "loss": 0.50996685,
+      "step": 171
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 26.364585876464844,
+      "learning_rate": 2e-05,
+      "loss": 0.45880085,
+      "step": 172
+    },
+    {
+      "epoch": 0.346,
+      "grad_norm": 16.831697463989258,
+      "learning_rate": 2e-05,
+      "loss": 0.44258234,
+      "step": 173
+    },
+    {
+      "epoch": 0.348,
+      "grad_norm": 15.406957626342773,
+      "learning_rate": 2e-05,
+      "loss": 0.42777783,
+      "step": 174
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 27.91541862487793,
+      "learning_rate": 2e-05,
+      "loss": 0.44150254,
+      "step": 175
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 11.729520797729492,
+      "learning_rate": 2e-05,
+      "loss": 0.45409214,
+      "step": 176
+    },
+    {
+      "epoch": 0.354,
+      "grad_norm": 10.498335838317871,
+      "learning_rate": 2e-05,
+      "loss": 0.45166379,
+      "step": 177
+    },
+    {
+      "epoch": 0.356,
+      "grad_norm": 13.207727432250977,
+      "learning_rate": 2e-05,
+      "loss": 0.48104641,
+      "step": 178
+    },
+    {
+      "epoch": 0.358,
+      "grad_norm": 38.07108688354492,
+      "learning_rate": 2e-05,
+      "loss": 0.4513554,
+      "step": 179
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.624302864074707,
+      "learning_rate": 2e-05,
+      "loss": 0.41277581,
+      "step": 180
+    },
+    {
+      "epoch": 0.362,
+      "grad_norm": 8.47309398651123,
+      "learning_rate": 2e-05,
+      "loss": 0.42749909,
+      "step": 181
+    },
+    {
+      "epoch": 0.364,
+      "grad_norm": 5.230127811431885,
+      "learning_rate": 2e-05,
+      "loss": 0.39681259,
+      "step": 182
+    },
+    {
+      "epoch": 0.366,
+      "grad_norm": 11.410614013671875,
+      "learning_rate": 2e-05,
+      "loss": 0.39442861,
+      "step": 183
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 12.072883605957031,
+      "learning_rate": 2e-05,
+      "loss": 0.40660879,
+      "step": 184
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.550542831420898,
+      "learning_rate": 2e-05,
+      "loss": 0.40361917,
+      "step": 185
+    },
+    {
+      "epoch": 0.372,
+      "grad_norm": 6.943983554840088,
+      "learning_rate": 2e-05,
+      "loss": 0.45435604,
+      "step": 186
+    },
+    {
+      "epoch": 0.374,
+      "grad_norm": 7.458695888519287,
+      "learning_rate": 2e-05,
+      "loss": 0.36000335,
+      "step": 187
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 8.648634910583496,
+      "learning_rate": 2e-05,
+      "loss": 0.43435773,
+      "step": 188
+    },
+    {
+      "epoch": 0.378,
+      "grad_norm": 5.483082294464111,
+      "learning_rate": 2e-05,
+      "loss": 0.40732422,
+      "step": 189
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.177685737609863,
+      "learning_rate": 2e-05,
+      "loss": 0.42347455,
+      "step": 190
+    },
+    {
+      "epoch": 0.382,
+      "grad_norm": 3.8768444061279297,
+      "learning_rate": 2e-05,
+      "loss": 0.38617033,
+      "step": 191
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 4.045779705047607,
+      "learning_rate": 2e-05,
+      "loss": 0.3738341,
+      "step": 192
+    },
+    {
+      "epoch": 0.386,
+      "grad_norm": 4.49653434753418,
+      "learning_rate": 2e-05,
+      "loss": 0.41778257,
+      "step": 193
+    },
+    {
+      "epoch": 0.388,
+      "grad_norm": 3.0218472480773926,
+      "learning_rate": 2e-05,
+      "loss": 0.37529367,
+      "step": 194
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.7495503425598145,
+      "learning_rate": 2e-05,
+      "loss": 0.39258975,
+      "step": 195
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 4.811206817626953,
+      "learning_rate": 2e-05,
+      "loss": 0.39241675,
+      "step": 196
+    },
+    {
+      "epoch": 0.394,
+      "grad_norm": 2.8001279830932617,
+      "learning_rate": 2e-05,
+      "loss": 0.42924637,
+      "step": 197
+    },
+    {
+      "epoch": 0.396,
+      "grad_norm": 4.613635063171387,
+      "learning_rate": 2e-05,
+      "loss": 0.40108818,
+      "step": 198
+    },
+    {
+      "epoch": 0.398,
+      "grad_norm": 6.350893497467041,
+      "learning_rate": 2e-05,
+      "loss": 0.45137745,
+      "step": 199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.332032203674316,
+      "learning_rate": 2e-05,
+      "loss": 0.3409031,
+      "step": 200
+    },
+    {
+      "epoch": 0.402,
+      "grad_norm": 3.7277584075927734,
+      "learning_rate": 2e-05,
+      "loss": 0.4174028,
+      "step": 201
+    },
+    {
+      "epoch": 0.404,
+      "grad_norm": 2.3304224014282227,
+      "learning_rate": 2e-05,
+      "loss": 0.36742717,
+      "step": 202
+    },
+    {
+      "epoch": 0.406,
+      "grad_norm": 2.5752127170562744,
+      "learning_rate": 2e-05,
+      "loss": 0.36992675,
+      "step": 203
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 2.8590283393859863,
+      "learning_rate": 2e-05,
+      "loss": 0.37675118,
+      "step": 204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1647579669952393,
+      "learning_rate": 2e-05,
+      "loss": 0.3679235,
+      "step": 205
+    },
+    {
+      "epoch": 0.412,
+      "grad_norm": 2.6260170936584473,
+      "learning_rate": 2e-05,
+      "loss": 0.4498952,
+      "step": 206
+    },
+    {
+      "epoch": 0.414,
+      "grad_norm": 2.4622349739074707,
+      "learning_rate": 2e-05,
+      "loss": 0.35507852,
+      "step": 207
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 2.9872968196868896,
+      "learning_rate": 2e-05,
+      "loss": 0.3549549,
+      "step": 208
+    },
+    {
+      "epoch": 0.418,
+      "grad_norm": 3.04270601272583,
+      "learning_rate": 2e-05,
+      "loss": 0.41885364,
+      "step": 209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9570157527923584,
+      "learning_rate": 2e-05,
+      "loss": 0.43402624,
+      "step": 210
+    },
+    {
+      "epoch": 0.422,
+      "grad_norm": 2.0857791900634766,
+      "learning_rate": 2e-05,
+      "loss": 0.40431345,
+      "step": 211
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 2.574413537979126,
+      "learning_rate": 2e-05,
+      "loss": 0.40234259,
+      "step": 212
+    },
+    {
+      "epoch": 0.426,
+      "grad_norm": 2.3002917766571045,
+      "learning_rate": 2e-05,
+      "loss": 0.4213936,
+      "step": 213
+    },
+    {
+      "epoch": 0.428,
+      "grad_norm": 2.598877429962158,
+      "learning_rate": 2e-05,
+      "loss": 0.42018652,
+      "step": 214
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.410801887512207,
+      "learning_rate": 2e-05,
+      "loss": 0.41117987,
+      "step": 215
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 2.146622657775879,
+      "learning_rate": 2e-05,
+      "loss": 0.37472421,
+      "step": 216
+    },
+    {
+      "epoch": 0.434,
+      "grad_norm": 2.1747183799743652,
+      "learning_rate": 2e-05,
+      "loss": 0.42943805,
+      "step": 217
+    },
+    {
+      "epoch": 0.436,
+      "grad_norm": 2.395761013031006,
+      "learning_rate": 2e-05,
+      "loss": 0.39487824,
+      "step": 218
+    },
+    {
+      "epoch": 0.438,
+      "grad_norm": 2.3509767055511475,
+      "learning_rate": 2e-05,
+      "loss": 0.33290654,
+      "step": 219
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0364584922790527,
+      "learning_rate": 2e-05,
+      "loss": 0.40460128,
+      "step": 220
+    },
+    {
+      "epoch": 0.442,
+      "grad_norm": 2.566840648651123,
+      "learning_rate": 2e-05,
+      "loss": 0.41224802,
+      "step": 221
+    },
+    {
+      "epoch": 0.444,
+      "grad_norm": 3.774826765060425,
+      "learning_rate": 2e-05,
+      "loss": 0.43019855,
+      "step": 222
+    },
+    {
+      "epoch": 0.446,
+      "grad_norm": 2.1061086654663086,
+      "learning_rate": 2e-05,
+      "loss": 0.4085083,
+      "step": 223
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 2.4577176570892334,
+      "learning_rate": 2e-05,
+      "loss": 0.39225131,
+      "step": 224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.246354579925537,
+      "learning_rate": 2e-05,
+      "loss": 0.43242174,
+      "step": 225
+    },
+    {
+      "epoch": 0.452,
+      "grad_norm": 2.795057773590088,
+      "learning_rate": 2e-05,
+      "loss": 0.34006116,
+      "step": 226
+    },
+    {
+      "epoch": 0.454,
+      "grad_norm": 2.231876850128174,
+      "learning_rate": 2e-05,
+      "loss": 0.39721149,
+      "step": 227
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 3.058765172958374,
+      "learning_rate": 2e-05,
+      "loss": 0.41030541,
+      "step": 228
+    },
+    {
+      "epoch": 0.458,
+      "grad_norm": 2.9272921085357666,
+      "learning_rate": 2e-05,
+      "loss": 0.40642214,
+      "step": 229
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.118684768676758,
+      "learning_rate": 2e-05,
+      "loss": 0.40627092,
+      "step": 230
+    },
+    {
+      "epoch": 0.462,
+      "grad_norm": 4.35416841506958,
+      "learning_rate": 2e-05,
+      "loss": 0.41149405,
+      "step": 231
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 3.272524356842041,
+      "learning_rate": 2e-05,
+      "loss": 0.40240556,
+      "step": 232
+    },
+    {
+      "epoch": 0.466,
+      "grad_norm": 2.587944746017456,
+      "learning_rate": 2e-05,
+      "loss": 0.374861,
+      "step": 233
+    },
+    {
+      "epoch": 0.468,
+      "grad_norm": 2.573385238647461,
+      "learning_rate": 2e-05,
+      "loss": 0.42833674,
+      "step": 234
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.941406726837158,
+      "learning_rate": 2e-05,
+      "loss": 0.44332218,
+      "step": 235
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 2.2706453800201416,
+      "learning_rate": 2e-05,
+      "loss": 0.38916123,
+      "step": 236
+    },
+    {
+      "epoch": 0.474,
+      "grad_norm": 2.572369337081909,
+      "learning_rate": 2e-05,
+      "loss": 0.38100535,
+      "step": 237
+    },
+    {
+      "epoch": 0.476,
+      "grad_norm": 2.6199276447296143,
+      "learning_rate": 2e-05,
+      "loss": 0.39440346,
+      "step": 238
+    },
+    {
+      "epoch": 0.478,
+      "grad_norm": 2.281733512878418,
+      "learning_rate": 2e-05,
+      "loss": 0.3807506,
+      "step": 239
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7786262035369873,
+      "learning_rate": 2e-05,
+      "loss": 0.34910616,
+      "step": 240
+    },
+    {
+      "epoch": 0.482,
+      "grad_norm": 2.339578866958618,
+      "learning_rate": 2e-05,
+      "loss": 0.40123379,
+      "step": 241
+    },
+    {
+      "epoch": 0.484,
+      "grad_norm": 2.5612998008728027,
+      "learning_rate": 2e-05,
+      "loss": 0.34685874,
+      "step": 242
+    },
+    {
+      "epoch": 0.486,
+      "grad_norm": 2.650993824005127,
+      "learning_rate": 2e-05,
+      "loss": 0.37038237,
+      "step": 243
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 2.3199408054351807,
+      "learning_rate": 2e-05,
+      "loss": 0.35883853,
+      "step": 244
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2961573600769043,
+      "learning_rate": 2e-05,
+      "loss": 0.4362112,
+      "step": 245
+    },
+    {
+      "epoch": 0.492,
+      "grad_norm": 2.3096096515655518,
+      "learning_rate": 2e-05,
+      "loss": 0.3606168,
+      "step": 246
+    },
+    {
+      "epoch": 0.494,
+      "grad_norm": 2.3645904064178467,
+      "learning_rate": 2e-05,
+      "loss": 0.42863727,
+      "step": 247
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 2.2203831672668457,
+      "learning_rate": 2e-05,
+      "loss": 0.40489531,
+      "step": 248
+    },
+    {
+      "epoch": 0.498,
+      "grad_norm": 2.45977783203125,
+      "learning_rate": 2e-05,
+      "loss": 0.39940965,
+      "step": 249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9246718883514404,
+      "learning_rate": 2e-05,
+      "loss": 0.39140785,
+      "step": 250
+    },
+    {
+      "epoch": 0.502,
+      "grad_norm": 1.8212493658065796,
+      "learning_rate": 2e-05,
+      "loss": 0.36203432,
+      "step": 251
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 2.2326316833496094,
+      "learning_rate": 2e-05,
+      "loss": 0.44940746,
+      "step": 252
+    },
+    {
+      "epoch": 0.506,
+      "grad_norm": 2.1888906955718994,
+      "learning_rate": 2e-05,
+      "loss": 0.42821601,
+      "step": 253
+    },
+    {
+      "epoch": 0.508,
+      "grad_norm": 3.0814597606658936,
+      "learning_rate": 2e-05,
+      "loss": 0.36414945,
+      "step": 254
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7043652534484863,
+      "learning_rate": 2e-05,
+      "loss": 0.39156815,
+      "step": 255
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 2.3231751918792725,
+      "learning_rate": 2e-05,
+      "loss": 0.37385294,
+      "step": 256
+    },
+    {
+      "epoch": 0.514,
+      "grad_norm": 2.275028705596924,
+      "learning_rate": 2e-05,
+      "loss": 0.38484564,
+      "step": 257
+    },
+    {
+      "epoch": 0.516,
+      "grad_norm": 2.472454071044922,
+      "learning_rate": 2e-05,
+      "loss": 0.36479205,
+      "step": 258
+    },
+    {
+      "epoch": 0.518,
+      "grad_norm": 2.479954481124878,
+      "learning_rate": 2e-05,
+      "loss": 0.38072658,
+      "step": 259
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.311441659927368,
+      "learning_rate": 2e-05,
+      "loss": 0.33453172,
+      "step": 260
+    },
+    {
+      "epoch": 0.522,
+      "grad_norm": 3.8335723876953125,
+      "learning_rate": 2e-05,
+      "loss": 0.40032426,
+      "step": 261
+    },
+    {
+      "epoch": 0.524,
+      "grad_norm": 2.4983038902282715,
+      "learning_rate": 2e-05,
+      "loss": 0.37842286,
+      "step": 262
+    },
+    {
+      "epoch": 0.526,
+      "grad_norm": 2.8506362438201904,
+      "learning_rate": 2e-05,
+      "loss": 0.45623779,
+      "step": 263
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 2.5922443866729736,
+      "learning_rate": 2e-05,
+      "loss": 0.42091221,
+      "step": 264
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4378294944763184,
+      "learning_rate": 2e-05,
+      "loss": 0.43036759,
+      "step": 265
+    },
+    {
+      "epoch": 0.532,
+      "grad_norm": 2.5051655769348145,
+      "learning_rate": 2e-05,
+      "loss": 0.36193722,
+      "step": 266
+    },
+    {
+      "epoch": 0.534,
+      "grad_norm": 2.4639880657196045,
+      "learning_rate": 2e-05,
+      "loss": 0.42388517,
+      "step": 267
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 2.43228816986084,
+      "learning_rate": 2e-05,
+      "loss": 0.37180698,
+      "step": 268
+    },
+    {
+      "epoch": 0.538,
+      "grad_norm": 2.8706986904144287,
+      "learning_rate": 2e-05,
+      "loss": 0.3466678,
+      "step": 269
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8321847915649414,
+      "learning_rate": 2e-05,
+      "loss": 0.36745179,
+      "step": 270
+    },
+    {
+      "epoch": 0.542,
+      "grad_norm": 2.8276920318603516,
+      "learning_rate": 2e-05,
+      "loss": 0.38181525,
+      "step": 271
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 2.267299175262451,
+      "learning_rate": 2e-05,
+      "loss": 0.36299157,
+      "step": 272
+    },
+    {
+      "epoch": 0.546,
+      "grad_norm": 2.176030158996582,
+      "learning_rate": 2e-05,
+      "loss": 0.35883152,
+      "step": 273
+    },
+    {
+      "epoch": 0.548,
+      "grad_norm": 2.4151179790496826,
+      "learning_rate": 2e-05,
+      "loss": 0.35068169,
+      "step": 274
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5002832412719727,
+      "learning_rate": 2e-05,
+      "loss": 0.4389441,
+      "step": 275
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 3.062483072280884,
+      "learning_rate": 2e-05,
+      "loss": 0.38645649,
+      "step": 276
+    },
+    {
+      "epoch": 0.554,
+      "grad_norm": 1.8937181234359741,
+      "learning_rate": 2e-05,
+      "loss": 0.32340169,
+      "step": 277
+    },
+    {
+      "epoch": 0.556,
+      "grad_norm": 2.4653971195220947,
+      "learning_rate": 2e-05,
+      "loss": 0.44503152,
+      "step": 278
+    },
+    {
+      "epoch": 0.558,
+      "grad_norm": 2.399894952774048,
+      "learning_rate": 2e-05,
+      "loss": 0.3736915,
+      "step": 279
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2674710750579834,
+      "learning_rate": 2e-05,
+      "loss": 0.37925249,
+      "step": 280
+    },
+    {
+      "epoch": 0.562,
+      "grad_norm": 2.1981282234191895,
+      "learning_rate": 2e-05,
+      "loss": 0.379641,
+      "step": 281
+    },
+    {
+      "epoch": 0.564,
+      "grad_norm": 2.4101972579956055,
+      "learning_rate": 2e-05,
+      "loss": 0.32225233,
+      "step": 282
+    },
+    {
+      "epoch": 0.566,
+      "grad_norm": 3.3641390800476074,
+      "learning_rate": 2e-05,
+      "loss": 0.40978536,
+      "step": 283
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 2.453429937362671,
+      "learning_rate": 2e-05,
+      "loss": 0.42054862,
+      "step": 284
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.228393793106079,
+      "learning_rate": 2e-05,
+      "loss": 0.36940455,
+      "step": 285
+    },
+    {
+      "epoch": 0.572,
+      "grad_norm": 1.8759377002716064,
+      "learning_rate": 2e-05,
+      "loss": 0.3323741,
+      "step": 286
+    },
+    {
+      "epoch": 0.574,
+      "grad_norm": 2.5758562088012695,
+      "learning_rate": 2e-05,
+      "loss": 0.40134084,
+      "step": 287
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 2.2973790168762207,
+      "learning_rate": 2e-05,
+      "loss": 0.42725891,
+      "step": 288
+    },
+    {
+      "epoch": 0.578,
+      "grad_norm": 2.4652743339538574,
+      "learning_rate": 2e-05,
+      "loss": 0.35494244,
+      "step": 289
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2309014797210693,
+      "learning_rate": 2e-05,
+      "loss": 0.3866846,
+      "step": 290
+    },
+    {
+      "epoch": 0.582,
+      "grad_norm": 2.242001533508301,
+      "learning_rate": 2e-05,
+      "loss": 0.4193109,
+      "step": 291
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 2.306638479232788,
+      "learning_rate": 2e-05,
+      "loss": 0.38435796,
+      "step": 292
+    },
+    {
+      "epoch": 0.586,
+      "grad_norm": 2.3204331398010254,
+      "learning_rate": 2e-05,
+      "loss": 0.32084846,
+      "step": 293
+    },
+    {
+      "epoch": 0.588,
+      "grad_norm": 2.4214060306549072,
+      "learning_rate": 2e-05,
+      "loss": 0.36136317,
+      "step": 294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7441887855529785,
+      "learning_rate": 2e-05,
+      "loss": 0.38891983,
+      "step": 295
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 2.1641783714294434,
+      "learning_rate": 2e-05,
+      "loss": 0.3675043,
+      "step": 296
+    },
+    {
+      "epoch": 0.594,
+      "grad_norm": 2.537519931793213,
+      "learning_rate": 2e-05,
+      "loss": 0.39061338,
+      "step": 297
+    },
+    {
+      "epoch": 0.596,
+      "grad_norm": 2.3807148933410645,
+      "learning_rate": 2e-05,
+      "loss": 0.38095194,
+      "step": 298
+    },
+    {
+      "epoch": 0.598,
+      "grad_norm": 3.760620594024658,
+      "learning_rate": 2e-05,
+      "loss": 0.38023266,
+      "step": 299
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8382818698883057,
+      "learning_rate": 2e-05,
+      "loss": 0.42183822,
+      "step": 300
+    },
+    {
+      "epoch": 0.602,
+      "grad_norm": 2.5104598999023438,
+      "learning_rate": 2e-05,
+      "loss": 0.38760763,
+      "step": 301
+    },
+    {
+      "epoch": 0.604,
+      "grad_norm": 3.4479098320007324,
+      "learning_rate": 2e-05,
+      "loss": 0.38792706,
+      "step": 302
+    },
+    {
+      "epoch": 0.606,
+      "grad_norm": 2.2696542739868164,
+      "learning_rate": 2e-05,
+      "loss": 0.43671882,
+      "step": 303
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 2.4829628467559814,
+      "learning_rate": 2e-05,
+      "loss": 0.38020101,
+      "step": 304
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.053079843521118,
+      "learning_rate": 2e-05,
+      "loss": 0.42556402,
+      "step": 305
+    },
+    {
+      "epoch": 0.612,
+      "grad_norm": 2.130146026611328,
+      "learning_rate": 2e-05,
+      "loss": 0.36177719,
+      "step": 306
+    },
+    {
+      "epoch": 0.614,
+      "grad_norm": 5.296567440032959,
+      "learning_rate": 2e-05,
+      "loss": 0.40961123,
+      "step": 307
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 2.4534523487091064,
+      "learning_rate": 2e-05,
+      "loss": 0.34303677,
+      "step": 308
+    },
+    {
+      "epoch": 0.618,
+      "grad_norm": 1.9672064781188965,
+      "learning_rate": 2e-05,
+      "loss": 0.42068005,
+      "step": 309
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3204848766326904,
+      "learning_rate": 2e-05,
+      "loss": 0.37758303,
+      "step": 310
+    },
+    {
+      "epoch": 0.622,
+      "grad_norm": 2.4542860984802246,
+      "learning_rate": 2e-05,
+      "loss": 0.41424572,
+      "step": 311
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 1.9260770082473755,
+      "learning_rate": 2e-05,
+      "loss": 0.34970552,
+      "step": 312
+    },
+    {
+      "epoch": 0.626,
+      "grad_norm": 2.6891448497772217,
+      "learning_rate": 2e-05,
+      "loss": 0.30088019,
+      "step": 313
+    },
+    {
+      "epoch": 0.628,
+      "grad_norm": 2.968409538269043,
+      "learning_rate": 2e-05,
+      "loss": 0.42999855,
+      "step": 314
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5964770317077637,
+      "learning_rate": 2e-05,
+      "loss": 0.41451907,
+      "step": 315
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 2.4311039447784424,
+      "learning_rate": 2e-05,
+      "loss": 0.35747087,
+      "step": 316
+    },
+    {
+      "epoch": 0.634,
+      "grad_norm": 2.2850985527038574,
+      "learning_rate": 2e-05,
+      "loss": 0.45336699,
+      "step": 317
+    },
+    {
+      "epoch": 0.636,
+      "grad_norm": 2.128378391265869,
+      "learning_rate": 2e-05,
+      "loss": 0.40923813,
+      "step": 318
+    },
+    {
+      "epoch": 0.638,
+      "grad_norm": 2.9572834968566895,
+      "learning_rate": 2e-05,
+      "loss": 0.40831214,
+      "step": 319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1294198036193848,
+      "learning_rate": 2e-05,
+      "loss": 0.352382,
+      "step": 320
+    },
+    {
+      "epoch": 0.642,
+      "grad_norm": 2.146904706954956,
+      "learning_rate": 2e-05,
+      "loss": 0.36986923,
+      "step": 321
+    },
+    {
+      "epoch": 0.644,
+      "grad_norm": 2.226712465286255,
+      "learning_rate": 2e-05,
+      "loss": 0.37009352,
+      "step": 322
+    },
+    {
+      "epoch": 0.646,
+      "grad_norm": 2.8748672008514404,
+      "learning_rate": 2e-05,
+      "loss": 0.39577782,
+      "step": 323
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 2.0993311405181885,
+      "learning_rate": 2e-05,
+      "loss": 0.34025472,
+      "step": 324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.553697347640991,
+      "learning_rate": 2e-05,
+      "loss": 0.35979444,
+      "step": 325
+    },
+    {
+      "epoch": 0.652,
+      "grad_norm": 3.0230658054351807,
+      "learning_rate": 2e-05,
+      "loss": 0.38101768,
+      "step": 326
+    },
+    {
+      "epoch": 0.654,
+      "grad_norm": 2.227163076400757,
+      "learning_rate": 2e-05,
+      "loss": 0.36153871,
+      "step": 327
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 2.366971731185913,
+      "learning_rate": 2e-05,
+      "loss": 0.40203893,
+      "step": 328
+    },
+    {
+      "epoch": 0.658,
+      "grad_norm": 4.39686393737793,
+      "learning_rate": 2e-05,
+      "loss": 0.34971178,
+      "step": 329
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1364448070526123,
+      "learning_rate": 2e-05,
+      "loss": 0.38524896,
+      "step": 330
+    },
+    {
+      "epoch": 0.662,
+      "grad_norm": 2.4799602031707764,
+      "learning_rate": 2e-05,
+      "loss": 0.3510325,
+      "step": 331
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 3.8624725341796875,
+      "learning_rate": 2e-05,
+      "loss": 0.40343612,
+      "step": 332
+    },
+    {
+      "epoch": 0.666,
+      "grad_norm": 2.8651304244995117,
+      "learning_rate": 2e-05,
+      "loss": 0.3543312,
+      "step": 333
+    },
+    {
+      "epoch": 0.668,
+      "grad_norm": 2.507993221282959,
+      "learning_rate": 2e-05,
+      "loss": 0.4071638,
+      "step": 334
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.307523488998413,
+      "learning_rate": 2e-05,
+      "loss": 0.34446114,
+      "step": 335
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 2.8034613132476807,
+      "learning_rate": 2e-05,
+      "loss": 0.36103964,
+      "step": 336
+    },
+    {
+      "epoch": 0.674,
+      "grad_norm": 2.804739236831665,
+      "learning_rate": 2e-05,
+      "loss": 0.34058389,
+      "step": 337
+    },
+    {
+      "epoch": 0.676,
+      "grad_norm": 2.3788864612579346,
+      "learning_rate": 2e-05,
+      "loss": 0.3461957,
+      "step": 338
+    },
+    {
+      "epoch": 0.678,
+      "grad_norm": 2.2598109245300293,
+      "learning_rate": 2e-05,
+      "loss": 0.34844282,
+      "step": 339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7016749382019043,
+      "learning_rate": 2e-05,
+      "loss": 0.37224805,
+      "step": 340
+    },
+    {
+      "epoch": 0.682,
+      "grad_norm": 2.102294445037842,
+      "learning_rate": 2e-05,
+      "loss": 0.34660944,
+      "step": 341
+    },
+    {
+      "epoch": 0.684,
+      "grad_norm": 2.256808280944824,
+      "learning_rate": 2e-05,
+      "loss": 0.36872256,
+      "step": 342
+    },
+    {
+      "epoch": 0.686,
+      "grad_norm": 2.219933032989502,
+      "learning_rate": 2e-05,
+      "loss": 0.41074491,
+      "step": 343
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 38.59788131713867,
+      "learning_rate": 2e-05,
+      "loss": 0.42532402,
+      "step": 344
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8445427417755127,
+      "learning_rate": 2e-05,
+      "loss": 0.40610889,
+      "step": 345
+    },
+    {
+      "epoch": 0.692,
+      "grad_norm": 3.2422025203704834,
+      "learning_rate": 2e-05,
+      "loss": 0.34015438,
+      "step": 346
+    },
+    {
+      "epoch": 0.694,
+      "grad_norm": 2.0103511810302734,
+      "learning_rate": 2e-05,
+      "loss": 0.33063939,
+      "step": 347
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 2.821288585662842,
+      "learning_rate": 2e-05,
+      "loss": 0.40264255,
+      "step": 348
+    },
+    {
+      "epoch": 0.698,
+      "grad_norm": 3.3211405277252197,
+      "learning_rate": 2e-05,
+      "loss": 0.33633679,
+      "step": 349
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1281816959381104,
+      "learning_rate": 2e-05,
+      "loss": 0.36737266,
+      "step": 350
+    },
+    {
+      "epoch": 0.702,
+      "grad_norm": 2.546674966812134,
+      "learning_rate": 2e-05,
+      "loss": 0.39834434,
+      "step": 351
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 1.9590989351272583,
+      "learning_rate": 2e-05,
+      "loss": 0.376955,
+      "step": 352
+    },
+    {
+      "epoch": 0.706,
+      "grad_norm": 2.5581297874450684,
+      "learning_rate": 2e-05,
+      "loss": 0.39899454,
+      "step": 353
+    },
+    {
+      "epoch": 0.708,
+      "grad_norm": 2.3324179649353027,
+      "learning_rate": 2e-05,
+      "loss": 0.37603098,
+      "step": 354
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.334880828857422,
+      "learning_rate": 2e-05,
+      "loss": 0.42339247,
+      "step": 355
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 3.104797601699829,
+      "learning_rate": 2e-05,
+      "loss": 0.36481744,
+      "step": 356
+    },
+    {
+      "epoch": 0.714,
+      "grad_norm": 6.291801929473877,
+      "learning_rate": 2e-05,
+      "loss": 0.40678573,
+      "step": 357
+    },
+    {
+      "epoch": 0.716,
+      "grad_norm": 4.675421237945557,
+      "learning_rate": 2e-05,
+      "loss": 0.39416665,
+      "step": 358
+    },
+    {
+      "epoch": 0.718,
+      "grad_norm": 2.531383752822876,
+      "learning_rate": 2e-05,
+      "loss": 0.35706043,
+      "step": 359
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.778651237487793,
+      "learning_rate": 2e-05,
+      "loss": 0.41085088,
+      "step": 360
+    },
+    {
+      "epoch": 0.722,
+      "grad_norm": 2.785738945007324,
+      "learning_rate": 2e-05,
+      "loss": 0.31312498,
+      "step": 361
+    },
+    {
+      "epoch": 0.724,
+      "grad_norm": 4.2192063331604,
+      "learning_rate": 2e-05,
+      "loss": 0.43665931,
+      "step": 362
+    },
+    {
+      "epoch": 0.726,
+      "grad_norm": 2.268357515335083,
+      "learning_rate": 2e-05,
+      "loss": 0.41825864,
+      "step": 363
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 2.246572494506836,
+      "learning_rate": 2e-05,
+      "loss": 0.36334115,
+      "step": 364
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2913174629211426,
+      "learning_rate": 2e-05,
+      "loss": 0.40367627,
+      "step": 365
+    },
+    {
+      "epoch": 0.732,
+      "grad_norm": 1.9627586603164673,
+      "learning_rate": 2e-05,
+      "loss": 0.36439764,
+      "step": 366
+    },
+    {
+      "epoch": 0.734,
+      "grad_norm": 2.0793967247009277,
+      "learning_rate": 2e-05,
+      "loss": 0.3633121,
+      "step": 367
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 2.039069175720215,
+      "learning_rate": 2e-05,
+      "loss": 0.37875551,
+      "step": 368
+    },
+    {
+      "epoch": 0.738,
+      "grad_norm": 2.103626251220703,
+      "learning_rate": 2e-05,
+      "loss": 0.34807205,
+      "step": 369
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.898775339126587,
+      "learning_rate": 2e-05,
+      "loss": 0.3396126,
+      "step": 370
+    },
+    {
+      "epoch": 0.742,
+      "grad_norm": 2.118276596069336,
+      "learning_rate": 2e-05,
+      "loss": 0.38108289,
+      "step": 371
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 2.3276379108428955,
+      "learning_rate": 2e-05,
+      "loss": 0.35690206,
+      "step": 372
+    },
+    {
+      "epoch": 0.746,
+      "grad_norm": 2.173872232437134,
+      "learning_rate": 2e-05,
+      "loss": 0.35335249,
+      "step": 373
+    },
+    {
+      "epoch": 0.748,
+      "grad_norm": 1.8903744220733643,
+      "learning_rate": 2e-05,
+      "loss": 0.37995228,
+      "step": 374
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2106189727783203,
+      "learning_rate": 2e-05,
+      "loss": 0.34882003,
+      "step": 375
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 2.048971652984619,
+      "learning_rate": 2e-05,
+      "loss": 0.43542331,
+      "step": 376
+    },
+    {
+      "epoch": 0.754,
+      "grad_norm": 2.003777503967285,
+      "learning_rate": 2e-05,
+      "loss": 0.3117795,
+      "step": 377
+    },
+    {
+      "epoch": 0.756,
+      "grad_norm": 2.0449531078338623,
+      "learning_rate": 2e-05,
+      "loss": 0.33903271,
+      "step": 378
+    },
+    {
+      "epoch": 0.758,
+      "grad_norm": 2.3183586597442627,
+      "learning_rate": 2e-05,
+      "loss": 0.33581644,
+      "step": 379
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9782938957214355,
+      "learning_rate": 2e-05,
+      "loss": 0.31719434,
+      "step": 380
+    },
+    {
+      "epoch": 0.762,
+      "grad_norm": 2.3408687114715576,
+      "learning_rate": 2e-05,
+      "loss": 0.38454038,
+      "step": 381
+    },
+    {
+      "epoch": 0.764,
+      "grad_norm": 2.642451524734497,
+      "learning_rate": 2e-05,
+      "loss": 0.36350057,
+      "step": 382
+    },
+    {
+      "epoch": 0.766,
+      "grad_norm": 2.2716588973999023,
+      "learning_rate": 2e-05,
+      "loss": 0.3672176,
+      "step": 383
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 2.143385410308838,
+      "learning_rate": 2e-05,
+      "loss": 0.35499257,
+      "step": 384
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.251404285430908,
+      "learning_rate": 2e-05,
+      "loss": 0.33705157,
+      "step": 385
+    },
+    {
+      "epoch": 0.772,
+      "grad_norm": 2.599787712097168,
+      "learning_rate": 2e-05,
+      "loss": 0.40010357,
+      "step": 386
+    },
+    {
+      "epoch": 0.774,
+      "grad_norm": 2.932671308517456,
+      "learning_rate": 2e-05,
+      "loss": 0.41798162,
+      "step": 387
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 3.084031343460083,
+      "learning_rate": 2e-05,
+      "loss": 0.40057978,
+      "step": 388
+    },
+    {
+      "epoch": 0.778,
+      "grad_norm": 2.370199680328369,
+      "learning_rate": 2e-05,
+      "loss": 0.37256229,
+      "step": 389
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.715414524078369,
+      "learning_rate": 2e-05,
+      "loss": 0.3550342,
+      "step": 390
+    },
+    {
+      "epoch": 0.782,
+      "grad_norm": 2.370290756225586,
+      "learning_rate": 2e-05,
+      "loss": 0.40933335,
+      "step": 391
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 2.753520965576172,
+      "learning_rate": 2e-05,
+      "loss": 0.37647349,
+      "step": 392
+    },
+    {
+      "epoch": 0.786,
+      "grad_norm": 3.428513526916504,
+      "learning_rate": 2e-05,
+      "loss": 0.37809077,
+      "step": 393
+    },
+    {
+      "epoch": 0.788,
+      "grad_norm": 3.0542285442352295,
+      "learning_rate": 2e-05,
+      "loss": 0.3635264,
+      "step": 394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.631666421890259,
+      "learning_rate": 2e-05,
+      "loss": 0.37626997,
+      "step": 395
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 2.9206936359405518,
+      "learning_rate": 2e-05,
+      "loss": 0.3362987,
+      "step": 396
+    },
+    {
+      "epoch": 0.794,
+      "grad_norm": 2.600062608718872,
+      "learning_rate": 2e-05,
+      "loss": 0.36531377,
+      "step": 397
+    },
+    {
+      "epoch": 0.796,
+      "grad_norm": 2.959347724914551,
+      "learning_rate": 2e-05,
+      "loss": 0.39553091,
+      "step": 398
+    },
+    {
+      "epoch": 0.798,
+      "grad_norm": 2.826603651046753,
+      "learning_rate": 2e-05,
+      "loss": 0.35759783,
+      "step": 399
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.756749153137207,
+      "learning_rate": 2e-05,
+      "loss": 0.3761026,
+      "step": 400
+    },
+    {
+      "epoch": 0.802,
+      "grad_norm": 2.015024185180664,
+      "learning_rate": 2e-05,
+      "loss": 0.37341845,
+      "step": 401
+    },
+    {
+      "epoch": 0.804,
+      "grad_norm": 2.473043203353882,
+      "learning_rate": 2e-05,
+      "loss": 0.38090366,
+      "step": 402
+    },
+    {
+      "epoch": 0.806,
+      "grad_norm": 3.260671854019165,
+      "learning_rate": 2e-05,
+      "loss": 0.40409699,
+      "step": 403
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 1.9450372457504272,
+      "learning_rate": 2e-05,
+      "loss": 0.38324416,
+      "step": 404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.213120460510254,
+      "learning_rate": 2e-05,
+      "loss": 0.32400131,
+      "step": 405
+    },
+    {
+      "epoch": 0.812,
+      "grad_norm": 2.4603614807128906,
+      "learning_rate": 2e-05,
+      "loss": 0.38184336,
+      "step": 406
+    },
+    {
+      "epoch": 0.814,
+      "grad_norm": 2.590217351913452,
+      "learning_rate": 2e-05,
+      "loss": 0.31220454,
+      "step": 407
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 2.3522443771362305,
+      "learning_rate": 2e-05,
+      "loss": 0.37040508,
+      "step": 408
+    },
+    {
+      "epoch": 0.818,
+      "grad_norm": 3.8462178707122803,
+      "learning_rate": 2e-05,
+      "loss": 0.31150436,
+      "step": 409
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3633556365966797,
+      "learning_rate": 2e-05,
+      "loss": 0.33486691,
+      "step": 410
+    },
+    {
+      "epoch": 0.822,
+      "grad_norm": 2.797116756439209,
+      "learning_rate": 2e-05,
+      "loss": 0.35024345,
+      "step": 411
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 2.787532329559326,
+      "learning_rate": 2e-05,
+      "loss": 0.34601441,
+      "step": 412
+    },
+    {
+      "epoch": 0.826,
+      "grad_norm": 2.7806966304779053,
+      "learning_rate": 2e-05,
+      "loss": 0.37785235,
+      "step": 413
+    },
+    {
+      "epoch": 0.828,
+      "grad_norm": 3.3908019065856934,
+      "learning_rate": 2e-05,
+      "loss": 0.40772134,
+      "step": 414
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3474032878875732,
+      "learning_rate": 2e-05,
+      "loss": 0.31730652,
+      "step": 415
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 3.695180892944336,
+      "learning_rate": 2e-05,
+      "loss": 0.38893342,
+      "step": 416
+    },
+    {
+      "epoch": 0.834,
+      "grad_norm": 2.969644784927368,
+      "learning_rate": 2e-05,
+      "loss": 0.37266415,
+      "step": 417
+    },
+    {
+      "epoch": 0.836,
+      "grad_norm": 3.3439037799835205,
+      "learning_rate": 2e-05,
+      "loss": 0.36170414,
+      "step": 418
+    },
+    {
+      "epoch": 0.838,
+      "grad_norm": 2.669370651245117,
+      "learning_rate": 2e-05,
+      "loss": 0.41537088,
+      "step": 419
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2042794227600098,
+      "learning_rate": 2e-05,
+      "loss": 0.34706631,
+      "step": 420
+    },
+    {
+      "epoch": 0.842,
+      "grad_norm": 3.0707359313964844,
+      "learning_rate": 2e-05,
+      "loss": 0.36334276,
+      "step": 421
+    },
+    {
+      "epoch": 0.844,
+      "grad_norm": 3.0460245609283447,
+      "learning_rate": 2e-05,
+      "loss": 0.3588101,
+      "step": 422
+    },
+    {
+      "epoch": 0.846,
+      "grad_norm": 3.4480361938476562,
+      "learning_rate": 2e-05,
+      "loss": 0.35593536,
+      "step": 423
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 3.668825626373291,
+      "learning_rate": 2e-05,
+      "loss": 0.36293906,
+      "step": 424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.644979953765869,
+      "learning_rate": 2e-05,
+      "loss": 0.33053726,
+      "step": 425
+    },
+    {
+      "epoch": 0.852,
+      "grad_norm": 3.7507691383361816,
+      "learning_rate": 2e-05,
+      "loss": 0.33477077,
+      "step": 426
+    },
+    {
+      "epoch": 0.854,
+      "grad_norm": 3.2668840885162354,
+      "learning_rate": 2e-05,
+      "loss": 0.35934401,
+      "step": 427
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 3.24027681350708,
+      "learning_rate": 2e-05,
+      "loss": 0.28396821,
+      "step": 428
+    },
+    {
+      "epoch": 0.858,
+      "grad_norm": 12.503198623657227,
+      "learning_rate": 2e-05,
+      "loss": 0.32771713,
+      "step": 429
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.297608852386475,
+      "learning_rate": 2e-05,
+      "loss": 0.33745548,
+      "step": 430
+    },
+    {
+      "epoch": 0.862,
+      "grad_norm": 4.472741603851318,
+      "learning_rate": 2e-05,
+      "loss": 0.27140242,
+      "step": 431
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 3.206968069076538,
+      "learning_rate": 2e-05,
+      "loss": 0.28080299,
+      "step": 432
+    },
+    {
+      "epoch": 0.866,
+      "grad_norm": 3.3867924213409424,
+      "learning_rate": 2e-05,
+      "loss": 0.3420102,
+      "step": 433
+    },
+    {
+      "epoch": 0.868,
+      "grad_norm": 3.833103656768799,
+      "learning_rate": 2e-05,
+      "loss": 0.27758431,
+      "step": 434
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7450990676879883,
+      "learning_rate": 2e-05,
+      "loss": 0.27860394,
+      "step": 435
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 4.227891445159912,
+      "learning_rate": 2e-05,
+      "loss": 0.31056997,
+      "step": 436
+    },
+    {
+      "epoch": 0.874,
+      "grad_norm": 3.7188467979431152,
+      "learning_rate": 2e-05,
+      "loss": 0.26568019,
+      "step": 437
+    },
+    {
+      "epoch": 0.876,
+      "grad_norm": 3.5237154960632324,
+      "learning_rate": 2e-05,
+      "loss": 0.25259641,
+      "step": 438
+    },
+    {
+      "epoch": 0.878,
+      "grad_norm": 5.686617851257324,
+      "learning_rate": 2e-05,
+      "loss": 0.2755993,
+      "step": 439
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4197587966918945,
+      "learning_rate": 2e-05,
+      "loss": 0.23646541,
+      "step": 440
+    },
+    {
+      "epoch": 0.882,
+      "grad_norm": 4.551022052764893,
+      "learning_rate": 2e-05,
+      "loss": 0.22122362,
+      "step": 441
+    },
+    {
+      "epoch": 0.884,
+      "grad_norm": 4.930042266845703,
+      "learning_rate": 2e-05,
+      "loss": 0.27152666,
+      "step": 442
+    },
+    {
+      "epoch": 0.886,
+      "grad_norm": 3.824270725250244,
+      "learning_rate": 2e-05,
+      "loss": 0.2230306,
+      "step": 443
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 3.917961597442627,
+      "learning_rate": 2e-05,
+      "loss": 0.27966443,
+      "step": 444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9343419075012207,
+      "learning_rate": 2e-05,
+      "loss": 0.20023456,
+      "step": 445
+    },
+    {
+      "epoch": 0.892,
+      "grad_norm": 4.3064775466918945,
+      "learning_rate": 2e-05,
+      "loss": 0.19676761,
+      "step": 446
+    },
+    {
+      "epoch": 0.894,
+      "grad_norm": 3.1968557834625244,
+      "learning_rate": 2e-05,
+      "loss": 0.19466686,
+      "step": 447
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 4.086765289306641,
+      "learning_rate": 2e-05,
+      "loss": 0.22548294,
+      "step": 448
+    },
+    {
+      "epoch": 0.898,
+      "grad_norm": 4.0280256271362305,
+      "learning_rate": 2e-05,
+      "loss": 0.22720584,
+      "step": 449
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.620728492736816,
+      "learning_rate": 2e-05,
+      "loss": 0.17157443,
+      "step": 450
+    },
+    {
+      "epoch": 0.902,
+      "grad_norm": 3.6249382495880127,
+      "learning_rate": 2e-05,
+      "loss": 0.23641428,
+      "step": 451
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 4.136668682098389,
+      "learning_rate": 2e-05,
+      "loss": 0.21479097,
+      "step": 452
+    },
+    {
+      "epoch": 0.906,
+      "grad_norm": 7.2505879402160645,
+      "learning_rate": 2e-05,
+      "loss": 0.27446824,
+      "step": 453
+    },
+    {
+      "epoch": 0.908,
+      "grad_norm": 5.988163471221924,
+      "learning_rate": 2e-05,
+      "loss": 0.21940503,
+      "step": 454
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.0386505126953125,
+      "learning_rate": 2e-05,
+      "loss": 0.19830205,
+      "step": 455
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 3.1702799797058105,
+      "learning_rate": 2e-05,
+      "loss": 0.1882799,
+      "step": 456
+    },
+    {
+      "epoch": 0.914,
+      "grad_norm": 4.0222015380859375,
+      "learning_rate": 2e-05,
+      "loss": 0.23622099,
+      "step": 457
+    },
+    {
+      "epoch": 0.916,
+      "grad_norm": 3.4282891750335693,
+      "learning_rate": 2e-05,
+      "loss": 0.18106145,
+      "step": 458
+    },
+    {
+      "epoch": 0.918,
+      "grad_norm": 4.8152337074279785,
+      "learning_rate": 2e-05,
+      "loss": 0.17646313,
+      "step": 459
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1678340435028076,
+      "learning_rate": 2e-05,
+      "loss": 0.20366624,
+      "step": 460
+    },
+    {
+      "epoch": 0.922,
+      "grad_norm": 2.8468918800354004,
+      "learning_rate": 2e-05,
+      "loss": 0.20546392,
+      "step": 461
+    },
+    {
+      "epoch": 0.924,
+      "grad_norm": 3.2070858478546143,
+      "learning_rate": 2e-05,
+      "loss": 0.16359358,
+      "step": 462
+    },
+    {
+      "epoch": 0.926,
+      "grad_norm": 3.970043420791626,
+      "learning_rate": 2e-05,
+      "loss": 0.21810345,
+      "step": 463
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 2.7538678646087646,
+      "learning_rate": 2e-05,
+      "loss": 0.15927938,
+      "step": 464
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.887653112411499,
+      "learning_rate": 2e-05,
+      "loss": 0.17965358,
+      "step": 465
+    },
+    {
+      "epoch": 0.932,
+      "grad_norm": 2.4924097061157227,
+      "learning_rate": 2e-05,
+      "loss": 0.14478508,
+      "step": 466
+    },
+    {
+      "epoch": 0.934,
+      "grad_norm": 2.4270923137664795,
+      "learning_rate": 2e-05,
+      "loss": 0.1299592,
+      "step": 467
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 2.980048179626465,
+      "learning_rate": 2e-05,
+      "loss": 0.16589662,
+      "step": 468
+    },
+    {
+      "epoch": 0.938,
+      "grad_norm": 2.9414331912994385,
+      "learning_rate": 2e-05,
+      "loss": 0.15968111,
+      "step": 469
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0330123901367188,
+      "learning_rate": 2e-05,
+      "loss": 0.17940134,
+      "step": 470
+    },
+    {
+      "epoch": 0.942,
+      "grad_norm": 4.506875514984131,
+      "learning_rate": 2e-05,
+      "loss": 0.17710808,
+      "step": 471
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 4.033360481262207,
+      "learning_rate": 2e-05,
+      "loss": 0.16440117,
+      "step": 472
+    },
+    {
+      "epoch": 0.946,
+      "grad_norm": 3.041499376296997,
+      "learning_rate": 2e-05,
+      "loss": 0.18891403,
+      "step": 473
+    },
+    {
+      "epoch": 0.948,
+      "grad_norm": 3.819559335708618,
+      "learning_rate": 2e-05,
+      "loss": 0.17200641,
+      "step": 474
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.71626615524292,
+      "learning_rate": 2e-05,
+      "loss": 0.14698125,
+      "step": 475
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 4.742694854736328,
+      "learning_rate": 2e-05,
+      "loss": 0.18361038,
+      "step": 476
+    },
+    {
+      "epoch": 0.954,
+      "grad_norm": 3.403785228729248,
+      "learning_rate": 2e-05,
+      "loss": 0.15806127,
+      "step": 477
+    },
+    {
+      "epoch": 0.956,
+      "grad_norm": 3.088829278945923,
+      "learning_rate": 2e-05,
+      "loss": 0.19003233,
+      "step": 478
+    },
+    {
+      "epoch": 0.958,
+      "grad_norm": 4.706967353820801,
+      "learning_rate": 2e-05,
+      "loss": 0.19501597,
+      "step": 479
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4652099609375,
+      "learning_rate": 2e-05,
+      "loss": 0.16612351,
+      "step": 480
+    },
+    {
+      "epoch": 0.962,
+      "grad_norm": 4.185473918914795,
+      "learning_rate": 2e-05,
+      "loss": 0.14217728,
+      "step": 481
+    },
+    {
+      "epoch": 0.964,
+      "grad_norm": 3.6370861530303955,
+      "learning_rate": 2e-05,
+      "loss": 0.18404439,
+      "step": 482
+    },
+    {
+      "epoch": 0.966,
+      "grad_norm": 4.554826736450195,
+      "learning_rate": 2e-05,
+      "loss": 0.16499752,
+      "step": 483
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 2.758190393447876,
+      "learning_rate": 2e-05,
+      "loss": 0.13831472,
+      "step": 484
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.261568784713745,
+      "learning_rate": 2e-05,
+      "loss": 0.13929909,
+      "step": 485
+    },
+    {
+      "epoch": 0.972,
+      "grad_norm": 2.8820412158966064,
+      "learning_rate": 2e-05,
+      "loss": 0.14241502,
+      "step": 486
+    },
+    {
+      "epoch": 0.974,
+      "grad_norm": 3.352541208267212,
+      "learning_rate": 2e-05,
+      "loss": 0.16133608,
+      "step": 487
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 3.834998846054077,
+      "learning_rate": 2e-05,
+      "loss": 0.17684533,
+      "step": 488
+    },
+    {
+      "epoch": 0.978,
+      "grad_norm": 2.977851152420044,
+      "learning_rate": 2e-05,
+      "loss": 0.12988587,
+      "step": 489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.011771202087402,
+      "learning_rate": 2e-05,
+      "loss": 0.15765052,
+      "step": 490
+    },
+    {
+      "epoch": 0.982,
+      "grad_norm": 2.6912081241607666,
+      "learning_rate": 2e-05,
+      "loss": 0.12373734,
+      "step": 491
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 2.2780370712280273,
+      "learning_rate": 2e-05,
+      "loss": 0.15556982,
+      "step": 492
+    },
+    {
+      "epoch": 0.986,
+      "grad_norm": 2.763603448867798,
+      "learning_rate": 2e-05,
+      "loss": 0.14312422,
+      "step": 493
+    },
+    {
+      "epoch": 0.988,
+      "grad_norm": 2.51889705657959,
+      "learning_rate": 2e-05,
+      "loss": 0.14164892,
+      "step": 494
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4747464656829834,
+      "learning_rate": 2e-05,
+      "loss": 0.18990183,
+      "step": 495
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 2.276155471801758,
+      "learning_rate": 2e-05,
+      "loss": 0.13020995,
+      "step": 496
+    },
+    {
+      "epoch": 0.994,
+      "grad_norm": 2.6515796184539795,
+      "learning_rate": 2e-05,
+      "loss": 0.15413743,
+      "step": 497
+    },
+    {
+      "epoch": 0.996,
+      "grad_norm": 2.347593069076538,
+      "learning_rate": 2e-05,
+      "loss": 0.15986988,
+      "step": 498
+    },
+    {
+      "epoch": 0.998,
+      "grad_norm": 2.6618576049804688,
+      "learning_rate": 2e-05,
+      "loss": 0.15566903,
+      "step": 499
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4387569427490234,
+      "learning_rate": 2e-05,
+      "loss": 0.15638649,
+      "step": 500
+    },
+    {
+      "epoch": 1.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.976,
+        "AngleClassification_2": 0.634,
+        "AngleClassification_3": 0.499001996007984,
+        "Equal_1": 0.13,
+        "Equal_2": 0.08383233532934131,
+        "Equal_3": 0.14770459081836326,
+        "LineComparison_1": 0.498,
+        "LineComparison_2": 0.47305389221556887,
+        "LineComparison_3": 0.5069860279441117,
+        "Parallel_1": 0.40480961923847697,
+        "Parallel_2": 0.9038076152304609,
+        "Parallel_3": 0.254,
+        "Perpendicular_1": 0.522,
+        "Perpendicular_2": 0.16,
+        "Perpendicular_3": 0.1092184368737475,
+        "PointLiesOnCircle_1": 0.8086172344689379,
+        "PointLiesOnCircle_2": 0.5923333333333334,
+        "PointLiesOnCircle_3": 0.29560000000000003,
+        "PointLiesOnLine_1": 0.4529058116232465,
+        "PointLiesOnLine_2": 0.342685370741483,
+        "PointLiesOnLine_3": 0.2315369261477046
+      },
+      "eval_runtime": 324.0838,
+      "eval_samples_per_second": 32.399,
+      "eval_steps_per_second": 0.648,
+      "step": 500
+    },
+    {
+      "epoch": 1.002,
+      "grad_norm": 4.821013450622559,
+      "learning_rate": 2e-05,
+      "loss": 0.1556485,
+      "step": 501
+    },
+    {
+      "epoch": 1.004,
+      "grad_norm": 1.8802967071533203,
+      "learning_rate": 2e-05,
+      "loss": 0.11685692,
+      "step": 502
+    },
+    {
+      "epoch": 1.006,
+      "grad_norm": 1.7924933433532715,
+      "learning_rate": 2e-05,
+      "loss": 0.11074463,
+      "step": 503
+    },
+    {
+      "epoch": 1.008,
+      "grad_norm": 3.829883337020874,
+      "learning_rate": 2e-05,
+      "loss": 0.1522513,
+      "step": 504
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.246246337890625,
+      "learning_rate": 2e-05,
+      "loss": 0.15618658,
+      "step": 505
+    },
+    {
+      "epoch": 1.012,
+      "grad_norm": 4.451237678527832,
+      "learning_rate": 2e-05,
+      "loss": 0.23163846,
+      "step": 506
+    },
+    {
+      "epoch": 1.014,
+      "grad_norm": 2.5041894912719727,
+      "learning_rate": 2e-05,
+      "loss": 0.13576819,
+      "step": 507
+    },
+    {
+      "epoch": 1.016,
+      "grad_norm": 2.6320042610168457,
+      "learning_rate": 2e-05,
+      "loss": 0.16223896,
+      "step": 508
+    },
+    {
+      "epoch": 1.018,
+      "grad_norm": 1.8667449951171875,
+      "learning_rate": 2e-05,
+      "loss": 0.12035898,
+      "step": 509
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.622072458267212,
+      "learning_rate": 2e-05,
+      "loss": 0.13647205,
+      "step": 510
+    },
+    {
+      "epoch": 1.022,
+      "grad_norm": 4.481700420379639,
+      "learning_rate": 2e-05,
+      "loss": 0.15954015,
+      "step": 511
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 3.5658326148986816,
+      "learning_rate": 2e-05,
+      "loss": 0.12799534,
+      "step": 512
+    },
+    {
+      "epoch": 1.026,
+      "grad_norm": 6.7309370040893555,
+      "learning_rate": 2e-05,
+      "loss": 0.13949288,
+      "step": 513
+    },
+    {
+      "epoch": 1.028,
+      "grad_norm": 3.3907320499420166,
+      "learning_rate": 2e-05,
+      "loss": 0.1515884,
+      "step": 514
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.2120602130889893,
+      "learning_rate": 2e-05,
+      "loss": 0.11528069,
+      "step": 515
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 1.858370065689087,
+      "learning_rate": 2e-05,
+      "loss": 0.10049677,
+      "step": 516
+    },
+    {
+      "epoch": 1.034,
+      "grad_norm": 3.5174007415771484,
+      "learning_rate": 2e-05,
+      "loss": 0.16639367,
+      "step": 517
+    },
+    {
+      "epoch": 1.036,
+      "grad_norm": 2.5854098796844482,
+      "learning_rate": 2e-05,
+      "loss": 0.16602588,
+      "step": 518
+    },
+    {
+      "epoch": 1.038,
+      "grad_norm": 2.3209636211395264,
+      "learning_rate": 2e-05,
+      "loss": 0.11949471,
+      "step": 519
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.098727226257324,
+      "learning_rate": 2e-05,
+      "loss": 0.1452582,
+      "step": 520
+    },
+    {
+      "epoch": 1.042,
+      "grad_norm": 4.866150379180908,
+      "learning_rate": 2e-05,
+      "loss": 0.12613255,
+      "step": 521
+    },
+    {
+      "epoch": 1.044,
+      "grad_norm": 2.6759192943573,
+      "learning_rate": 2e-05,
+      "loss": 0.16882387,
+      "step": 522
+    },
+    {
+      "epoch": 1.046,
+      "grad_norm": 2.8629543781280518,
+      "learning_rate": 2e-05,
+      "loss": 0.17408426,
+      "step": 523
+    },
+    {
+      "epoch": 1.048,
+      "grad_norm": 3.283302068710327,
+      "learning_rate": 2e-05,
+      "loss": 0.13273655,
+      "step": 524
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.690016984939575,
+      "learning_rate": 2e-05,
+      "loss": 0.16916898,
+      "step": 525
+    },
+    {
+      "epoch": 1.052,
+      "grad_norm": 3.075875997543335,
+      "learning_rate": 2e-05,
+      "loss": 0.15706718,
+      "step": 526
+    },
+    {
+      "epoch": 1.054,
+      "grad_norm": 4.240052223205566,
+      "learning_rate": 2e-05,
+      "loss": 0.12036274,
+      "step": 527
+    },
+    {
+      "epoch": 1.056,
+      "grad_norm": 2.4694156646728516,
+      "learning_rate": 2e-05,
+      "loss": 0.16254291,
+      "step": 528
+    },
+    {
+      "epoch": 1.058,
+      "grad_norm": 2.702230930328369,
+      "learning_rate": 2e-05,
+      "loss": 0.14166299,
+      "step": 529
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.753516912460327,
+      "learning_rate": 2e-05,
+      "loss": 0.15594101,
+      "step": 530
+    },
+    {
+      "epoch": 1.062,
+      "grad_norm": 3.5839767456054688,
+      "learning_rate": 2e-05,
+      "loss": 0.12621114,
+      "step": 531
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 2.400475263595581,
+      "learning_rate": 2e-05,
+      "loss": 0.11684637,
+      "step": 532
+    },
+    {
+      "epoch": 1.066,
+      "grad_norm": 2.5850939750671387,
+      "learning_rate": 2e-05,
+      "loss": 0.16528159,
+      "step": 533
+    },
+    {
+      "epoch": 1.068,
+      "grad_norm": 2.623412609100342,
+      "learning_rate": 2e-05,
+      "loss": 0.11953619,
+      "step": 534
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.529339075088501,
+      "learning_rate": 2e-05,
+      "loss": 0.14916751,
+      "step": 535
+    },
+    {
+      "epoch": 1.072,
+      "grad_norm": 3.2517716884613037,
+      "learning_rate": 2e-05,
+      "loss": 0.15048769,
+      "step": 536
+    },
+    {
+      "epoch": 1.074,
+      "grad_norm": 1.9640837907791138,
+      "learning_rate": 2e-05,
+      "loss": 0.13277957,
+      "step": 537
+    },
+    {
+      "epoch": 1.076,
+      "grad_norm": 2.368446111679077,
+      "learning_rate": 2e-05,
+      "loss": 0.14852183,
+      "step": 538
+    },
+    {
+      "epoch": 1.078,
+      "grad_norm": 2.9297749996185303,
+      "learning_rate": 2e-05,
+      "loss": 0.15348649,
+      "step": 539
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.759192943572998,
+      "learning_rate": 2e-05,
+      "loss": 0.13191144,
+      "step": 540
+    },
+    {
+      "epoch": 1.082,
+      "grad_norm": 2.021772623062134,
+      "learning_rate": 2e-05,
+      "loss": 0.1124662,
+      "step": 541
+    },
+    {
+      "epoch": 1.084,
+      "grad_norm": 2.8374252319335938,
+      "learning_rate": 2e-05,
+      "loss": 0.1071616,
+      "step": 542
+    },
+    {
+      "epoch": 1.086,
+      "grad_norm": 2.1634042263031006,
+      "learning_rate": 2e-05,
+      "loss": 0.11521625,
+      "step": 543
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 2.589329719543457,
+      "learning_rate": 2e-05,
+      "loss": 0.13090719,
+      "step": 544
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.291887044906616,
+      "learning_rate": 2e-05,
+      "loss": 0.09445122,
+      "step": 545
+    },
+    {
+      "epoch": 1.092,
+      "grad_norm": 2.3259284496307373,
+      "learning_rate": 2e-05,
+      "loss": 0.12547138,
+      "step": 546
+    },
+    {
+      "epoch": 1.094,
+      "grad_norm": 2.191845417022705,
+      "learning_rate": 2e-05,
+      "loss": 0.12708214,
+      "step": 547
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 3.3253958225250244,
+      "learning_rate": 2e-05,
+      "loss": 0.14231913,
+      "step": 548
+    },
+    {
+      "epoch": 1.098,
+      "grad_norm": 2.9672670364379883,
+      "learning_rate": 2e-05,
+      "loss": 0.09965955,
+      "step": 549
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.215210199356079,
+      "learning_rate": 2e-05,
+      "loss": 0.13290197,
+      "step": 550
+    },
+    {
+      "epoch": 1.102,
+      "grad_norm": 3.2850470542907715,
+      "learning_rate": 2e-05,
+      "loss": 0.10610727,
+      "step": 551
+    },
+    {
+      "epoch": 1.104,
+      "grad_norm": 2.7427141666412354,
+      "learning_rate": 2e-05,
+      "loss": 0.12695783,
+      "step": 552
+    },
+    {
+      "epoch": 1.106,
+      "grad_norm": 3.2531254291534424,
+      "learning_rate": 2e-05,
+      "loss": 0.12038823,
+      "step": 553
+    },
+    {
+      "epoch": 1.108,
+      "grad_norm": 2.5022897720336914,
+      "learning_rate": 2e-05,
+      "loss": 0.13614482,
+      "step": 554
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.687258720397949,
+      "learning_rate": 2e-05,
+      "loss": 0.11169507,
+      "step": 555
+    },
+    {
+      "epoch": 1.112,
+      "grad_norm": 2.6494035720825195,
+      "learning_rate": 2e-05,
+      "loss": 0.11928535,
+      "step": 556
+    },
+    {
+      "epoch": 1.114,
+      "grad_norm": 2.6407148838043213,
+      "learning_rate": 2e-05,
+      "loss": 0.12640992,
+      "step": 557
+    },
+    {
+      "epoch": 1.116,
+      "grad_norm": 4.039032936096191,
+      "learning_rate": 2e-05,
+      "loss": 0.16989604,
+      "step": 558
+    },
+    {
+      "epoch": 1.1179999999999999,
+      "grad_norm": 3.1691837310791016,
+      "learning_rate": 2e-05,
+      "loss": 0.1739091,
+      "step": 559
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.1099650859832764,
+      "learning_rate": 2e-05,
+      "loss": 0.13166931,
+      "step": 560
+    },
+    {
+      "epoch": 1.1219999999999999,
+      "grad_norm": 2.5065221786499023,
+      "learning_rate": 2e-05,
+      "loss": 0.14683142,
+      "step": 561
+    },
+    {
+      "epoch": 1.124,
+      "grad_norm": 2.24849534034729,
+      "learning_rate": 2e-05,
+      "loss": 0.10228881,
+      "step": 562
+    },
+    {
+      "epoch": 1.126,
+      "grad_norm": 3.08331298828125,
+      "learning_rate": 2e-05,
+      "loss": 0.15210506,
+      "step": 563
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 2.5049123764038086,
+      "learning_rate": 2e-05,
+      "loss": 0.12219332,
+      "step": 564
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.139686346054077,
+      "learning_rate": 2e-05,
+      "loss": 0.11474234,
+      "step": 565
+    },
+    {
+      "epoch": 1.1320000000000001,
+      "grad_norm": 1.576886534690857,
+      "learning_rate": 2e-05,
+      "loss": 0.09722944,
+      "step": 566
+    },
+    {
+      "epoch": 1.134,
+      "grad_norm": 2.270977735519409,
+      "learning_rate": 2e-05,
+      "loss": 0.13902695,
+      "step": 567
+    },
+    {
+      "epoch": 1.1360000000000001,
+      "grad_norm": 3.4217686653137207,
+      "learning_rate": 2e-05,
+      "loss": 0.13287134,
+      "step": 568
+    },
+    {
+      "epoch": 1.138,
+      "grad_norm": 2.5737199783325195,
+      "learning_rate": 2e-05,
+      "loss": 0.14081581,
+      "step": 569
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 1.9033153057098389,
+      "learning_rate": 2e-05,
+      "loss": 0.13062626,
+      "step": 570
+    },
+    {
+      "epoch": 1.142,
+      "grad_norm": 1.9676927328109741,
+      "learning_rate": 2e-05,
+      "loss": 0.10712667,
+      "step": 571
+    },
+    {
+      "epoch": 1.144,
+      "grad_norm": 2.840505599975586,
+      "learning_rate": 2e-05,
+      "loss": 0.12923452,
+      "step": 572
+    },
+    {
+      "epoch": 1.146,
+      "grad_norm": 2.5398242473602295,
+      "learning_rate": 2e-05,
+      "loss": 0.14771613,
+      "step": 573
+    },
+    {
+      "epoch": 1.148,
+      "grad_norm": 2.549650192260742,
+      "learning_rate": 2e-05,
+      "loss": 0.13140582,
+      "step": 574
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.1299757957458496,
+      "learning_rate": 2e-05,
+      "loss": 0.12547702,
+      "step": 575
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 2.358029842376709,
+      "learning_rate": 2e-05,
+      "loss": 0.12179442,
+      "step": 576
+    },
+    {
+      "epoch": 1.154,
+      "grad_norm": 1.966770052909851,
+      "learning_rate": 2e-05,
+      "loss": 0.10394298,
+      "step": 577
+    },
+    {
+      "epoch": 1.156,
+      "grad_norm": 2.0004608631134033,
+      "learning_rate": 2e-05,
+      "loss": 0.14016056,
+      "step": 578
+    },
+    {
+      "epoch": 1.158,
+      "grad_norm": 3.5331180095672607,
+      "learning_rate": 2e-05,
+      "loss": 0.12845185,
+      "step": 579
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.299360513687134,
+      "learning_rate": 2e-05,
+      "loss": 0.12573363,
+      "step": 580
+    },
+    {
+      "epoch": 1.162,
+      "grad_norm": 2.8737940788269043,
+      "learning_rate": 2e-05,
+      "loss": 0.12317209,
+      "step": 581
+    },
+    {
+      "epoch": 1.164,
+      "grad_norm": 4.00046968460083,
+      "learning_rate": 2e-05,
+      "loss": 0.09280093,
+      "step": 582
+    },
+    {
+      "epoch": 1.166,
+      "grad_norm": 2.854579210281372,
+      "learning_rate": 2e-05,
+      "loss": 0.12138534,
+      "step": 583
+    },
+    {
+      "epoch": 1.168,
+      "grad_norm": 2.668538808822632,
+      "learning_rate": 2e-05,
+      "loss": 0.10895318,
+      "step": 584
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9555225372314453,
+      "learning_rate": 2e-05,
+      "loss": 0.10446753,
+      "step": 585
+    },
+    {
+      "epoch": 1.172,
+      "grad_norm": 2.943380117416382,
+      "learning_rate": 2e-05,
+      "loss": 0.09996414,
+      "step": 586
+    },
+    {
+      "epoch": 1.174,
+      "grad_norm": 2.4844489097595215,
+      "learning_rate": 2e-05,
+      "loss": 0.11357398,
+      "step": 587
+    },
+    {
+      "epoch": 1.176,
+      "grad_norm": 4.776824474334717,
+      "learning_rate": 2e-05,
+      "loss": 0.09684882,
+      "step": 588
+    },
+    {
+      "epoch": 1.178,
+      "grad_norm": 2.3888590335845947,
+      "learning_rate": 2e-05,
+      "loss": 0.10829578,
+      "step": 589
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.8649983406066895,
+      "learning_rate": 2e-05,
+      "loss": 0.10984306,
+      "step": 590
+    },
+    {
+      "epoch": 1.182,
+      "grad_norm": 2.9575302600860596,
+      "learning_rate": 2e-05,
+      "loss": 0.08193485,
+      "step": 591
+    },
+    {
+      "epoch": 1.184,
+      "grad_norm": 3.222970724105835,
+      "learning_rate": 2e-05,
+      "loss": 0.12279116,
+      "step": 592
+    },
+    {
+      "epoch": 1.186,
+      "grad_norm": 2.6744375228881836,
+      "learning_rate": 2e-05,
+      "loss": 0.12870033,
+      "step": 593
+    },
+    {
+      "epoch": 1.188,
+      "grad_norm": 4.028379917144775,
+      "learning_rate": 2e-05,
+      "loss": 0.13311404,
+      "step": 594
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.1691460609436035,
+      "learning_rate": 2e-05,
+      "loss": 0.12187042,
+      "step": 595
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 2.7343945503234863,
+      "learning_rate": 2e-05,
+      "loss": 0.15135875,
+      "step": 596
+    },
+    {
+      "epoch": 1.194,
+      "grad_norm": 2.8838765621185303,
+      "learning_rate": 2e-05,
+      "loss": 0.13852769,
+      "step": 597
+    },
+    {
+      "epoch": 1.196,
+      "grad_norm": 2.592514991760254,
+      "learning_rate": 2e-05,
+      "loss": 0.12420864,
+      "step": 598
+    },
+    {
+      "epoch": 1.198,
+      "grad_norm": 3.1650428771972656,
+      "learning_rate": 2e-05,
+      "loss": 0.1350922,
+      "step": 599
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.812872886657715,
+      "learning_rate": 2e-05,
+      "loss": 0.11503953,
+      "step": 600
+    },
+    {
+      "epoch": 1.202,
+      "grad_norm": 2.564276933670044,
+      "learning_rate": 2e-05,
+      "loss": 0.13502818,
+      "step": 601
+    },
+    {
+      "epoch": 1.204,
+      "grad_norm": 2.3174049854278564,
+      "learning_rate": 2e-05,
+      "loss": 0.1046789,
+      "step": 602
+    },
+    {
+      "epoch": 1.206,
+      "grad_norm": 5.605673789978027,
+      "learning_rate": 2e-05,
+      "loss": 0.14738208,
+      "step": 603
+    },
+    {
+      "epoch": 1.208,
+      "grad_norm": 3.615462303161621,
+      "learning_rate": 2e-05,
+      "loss": 0.15257767,
+      "step": 604
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.023568630218506,
+      "learning_rate": 2e-05,
+      "loss": 0.1027983,
+      "step": 605
+    },
+    {
+      "epoch": 1.212,
+      "grad_norm": 2.8860561847686768,
+      "learning_rate": 2e-05,
+      "loss": 0.14651015,
+      "step": 606
+    },
+    {
+      "epoch": 1.214,
+      "grad_norm": 2.644528388977051,
+      "learning_rate": 2e-05,
+      "loss": 0.13001838,
+      "step": 607
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 1.84329092502594,
+      "learning_rate": 2e-05,
+      "loss": 0.09636261,
+      "step": 608
+    },
+    {
+      "epoch": 1.218,
+      "grad_norm": 2.386714220046997,
+      "learning_rate": 2e-05,
+      "loss": 0.11445861,
+      "step": 609
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.4820950031280518,
+      "learning_rate": 2e-05,
+      "loss": 0.14000396,
+      "step": 610
+    },
+    {
+      "epoch": 1.222,
+      "grad_norm": 2.412814140319824,
+      "learning_rate": 2e-05,
+      "loss": 0.09793176,
+      "step": 611
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 2.0066609382629395,
+      "learning_rate": 2e-05,
+      "loss": 0.11375158,
+      "step": 612
+    },
+    {
+      "epoch": 1.226,
+      "grad_norm": 2.4800400733947754,
+      "learning_rate": 2e-05,
+      "loss": 0.11607377,
+      "step": 613
+    },
+    {
+      "epoch": 1.228,
+      "grad_norm": 2.5596139430999756,
+      "learning_rate": 2e-05,
+      "loss": 0.11540417,
+      "step": 614
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8948832750320435,
+      "learning_rate": 2e-05,
+      "loss": 0.10477018,
+      "step": 615
+    },
+    {
+      "epoch": 1.232,
+      "grad_norm": 2.5590243339538574,
+      "learning_rate": 2e-05,
+      "loss": 0.11847038,
+      "step": 616
+    },
+    {
+      "epoch": 1.234,
+      "grad_norm": 2.4262442588806152,
+      "learning_rate": 2e-05,
+      "loss": 0.1339002,
+      "step": 617
+    },
+    {
+      "epoch": 1.236,
+      "grad_norm": 3.6982271671295166,
+      "learning_rate": 2e-05,
+      "loss": 0.15697673,
+      "step": 618
+    },
+    {
+      "epoch": 1.238,
+      "grad_norm": 2.4189836978912354,
+      "learning_rate": 2e-05,
+      "loss": 0.09398519,
+      "step": 619
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.4480879306793213,
+      "learning_rate": 2e-05,
+      "loss": 0.10857891,
+      "step": 620
+    },
+    {
+      "epoch": 1.242,
+      "grad_norm": 2.9036366939544678,
+      "learning_rate": 2e-05,
+      "loss": 0.1304615,
+      "step": 621
+    },
+    {
+      "epoch": 1.244,
+      "grad_norm": 2.4365622997283936,
+      "learning_rate": 2e-05,
+      "loss": 0.09261293,
+      "step": 622
+    },
+    {
+      "epoch": 1.246,
+      "grad_norm": 3.2014307975769043,
+      "learning_rate": 2e-05,
+      "loss": 0.13648328,
+      "step": 623
+    },
+    {
+      "epoch": 1.248,
+      "grad_norm": 3.503472089767456,
+      "learning_rate": 2e-05,
+      "loss": 0.1255216,
+      "step": 624
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.31048583984375,
+      "learning_rate": 2e-05,
+      "loss": 0.09449267,
+      "step": 625
+    },
+    {
+      "epoch": 1.252,
+      "grad_norm": 3.5570435523986816,
+      "learning_rate": 2e-05,
+      "loss": 0.12387832,
+      "step": 626
+    },
+    {
+      "epoch": 1.254,
+      "grad_norm": 2.5230705738067627,
+      "learning_rate": 2e-05,
+      "loss": 0.11507116,
+      "step": 627
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 2.516810655593872,
+      "learning_rate": 2e-05,
+      "loss": 0.09768143,
+      "step": 628
+    },
+    {
+      "epoch": 1.258,
+      "grad_norm": 2.518480062484741,
+      "learning_rate": 2e-05,
+      "loss": 0.10322925,
+      "step": 629
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.009742259979248,
+      "learning_rate": 2e-05,
+      "loss": 0.13511106,
+      "step": 630
+    },
+    {
+      "epoch": 1.262,
+      "grad_norm": 2.5133047103881836,
+      "learning_rate": 2e-05,
+      "loss": 0.09286143,
+      "step": 631
+    },
+    {
+      "epoch": 1.264,
+      "grad_norm": 1.873865008354187,
+      "learning_rate": 2e-05,
+      "loss": 0.0830982,
+      "step": 632
+    },
+    {
+      "epoch": 1.266,
+      "grad_norm": 2.617849826812744,
+      "learning_rate": 2e-05,
+      "loss": 0.14133078,
+      "step": 633
+    },
+    {
+      "epoch": 1.268,
+      "grad_norm": 2.539656162261963,
+      "learning_rate": 2e-05,
+      "loss": 0.1445739,
+      "step": 634
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.631986379623413,
+      "learning_rate": 2e-05,
+      "loss": 0.1087111,
+      "step": 635
+    },
+    {
+      "epoch": 1.272,
+      "grad_norm": 3.4899914264678955,
+      "learning_rate": 2e-05,
+      "loss": 0.10924721,
+      "step": 636
+    },
+    {
+      "epoch": 1.274,
+      "grad_norm": 2.475717067718506,
+      "learning_rate": 2e-05,
+      "loss": 0.11710069,
+      "step": 637
+    },
+    {
+      "epoch": 1.276,
+      "grad_norm": 2.376640796661377,
+      "learning_rate": 2e-05,
+      "loss": 0.12215403,
+      "step": 638
+    },
+    {
+      "epoch": 1.278,
+      "grad_norm": 3.3351945877075195,
+      "learning_rate": 2e-05,
+      "loss": 0.1109551,
+      "step": 639
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 4.320706844329834,
+      "learning_rate": 2e-05,
+      "loss": 0.15159556,
+      "step": 640
+    },
+    {
+      "epoch": 1.282,
+      "grad_norm": 2.7680718898773193,
+      "learning_rate": 2e-05,
+      "loss": 0.1090942,
+      "step": 641
+    },
+    {
+      "epoch": 1.284,
+      "grad_norm": 3.066657304763794,
+      "learning_rate": 2e-05,
+      "loss": 0.12701407,
+      "step": 642
+    },
+    {
+      "epoch": 1.286,
+      "grad_norm": 2.19394588470459,
+      "learning_rate": 2e-05,
+      "loss": 0.1153975,
+      "step": 643
+    },
+    {
+      "epoch": 1.288,
+      "grad_norm": 3.781139612197876,
+      "learning_rate": 2e-05,
+      "loss": 0.13794988,
+      "step": 644
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.143144369125366,
+      "learning_rate": 2e-05,
+      "loss": 0.10998823,
+      "step": 645
+    },
+    {
+      "epoch": 1.292,
+      "grad_norm": 2.4902939796447754,
+      "learning_rate": 2e-05,
+      "loss": 0.10579651,
+      "step": 646
+    },
+    {
+      "epoch": 1.294,
+      "grad_norm": 3.307321310043335,
+      "learning_rate": 2e-05,
+      "loss": 0.1062723,
+      "step": 647
+    },
+    {
+      "epoch": 1.296,
+      "grad_norm": 2.1090455055236816,
+      "learning_rate": 2e-05,
+      "loss": 0.09155425,
+      "step": 648
+    },
+    {
+      "epoch": 1.298,
+      "grad_norm": 2.296339988708496,
+      "learning_rate": 2e-05,
+      "loss": 0.09818932,
+      "step": 649
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.05061936378479,
+      "learning_rate": 2e-05,
+      "loss": 0.10809691,
+      "step": 650
+    },
+    {
+      "epoch": 1.302,
+      "grad_norm": 3.311122417449951,
+      "learning_rate": 2e-05,
+      "loss": 0.16015843,
+      "step": 651
+    },
+    {
+      "epoch": 1.304,
+      "grad_norm": 3.0931217670440674,
+      "learning_rate": 2e-05,
+      "loss": 0.10492094,
+      "step": 652
+    },
+    {
+      "epoch": 1.306,
+      "grad_norm": 2.87888765335083,
+      "learning_rate": 2e-05,
+      "loss": 0.10102548,
+      "step": 653
+    },
+    {
+      "epoch": 1.308,
+      "grad_norm": 2.918794631958008,
+      "learning_rate": 2e-05,
+      "loss": 0.12097271,
+      "step": 654
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.5513689517974854,
+      "learning_rate": 2e-05,
+      "loss": 0.10996251,
+      "step": 655
+    },
+    {
+      "epoch": 1.312,
+      "grad_norm": 2.10673189163208,
+      "learning_rate": 2e-05,
+      "loss": 0.13955916,
+      "step": 656
+    },
+    {
+      "epoch": 1.314,
+      "grad_norm": 2.614741563796997,
+      "learning_rate": 2e-05,
+      "loss": 0.09799536,
+      "step": 657
+    },
+    {
+      "epoch": 1.316,
+      "grad_norm": 3.142801523208618,
+      "learning_rate": 2e-05,
+      "loss": 0.09701319,
+      "step": 658
+    },
+    {
+      "epoch": 1.318,
+      "grad_norm": 4.541778564453125,
+      "learning_rate": 2e-05,
+      "loss": 0.13997841,
+      "step": 659
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.7341322898864746,
+      "learning_rate": 2e-05,
+      "loss": 0.11460865,
+      "step": 660
+    },
+    {
+      "epoch": 1.322,
+      "grad_norm": 2.7065863609313965,
+      "learning_rate": 2e-05,
+      "loss": 0.11451581,
+      "step": 661
+    },
+    {
+      "epoch": 1.324,
+      "grad_norm": 2.814460515975952,
+      "learning_rate": 2e-05,
+      "loss": 0.11479338,
+      "step": 662
+    },
+    {
+      "epoch": 1.326,
+      "grad_norm": 2.700387477874756,
+      "learning_rate": 2e-05,
+      "loss": 0.11010472,
+      "step": 663
+    },
+    {
+      "epoch": 1.328,
+      "grad_norm": 2.559100866317749,
+      "learning_rate": 2e-05,
+      "loss": 0.11664652,
+      "step": 664
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.2762906551361084,
+      "learning_rate": 2e-05,
+      "loss": 0.09327792,
+      "step": 665
+    },
+    {
+      "epoch": 1.332,
+      "grad_norm": 2.627418279647827,
+      "learning_rate": 2e-05,
+      "loss": 0.08959809,
+      "step": 666
+    },
+    {
+      "epoch": 1.334,
+      "grad_norm": 2.182978868484497,
+      "learning_rate": 2e-05,
+      "loss": 0.09787555,
+      "step": 667
+    },
+    {
+      "epoch": 1.336,
+      "grad_norm": 2.3923563957214355,
+      "learning_rate": 2e-05,
+      "loss": 0.10572146,
+      "step": 668
+    },
+    {
+      "epoch": 1.338,
+      "grad_norm": 2.8589816093444824,
+      "learning_rate": 2e-05,
+      "loss": 0.13722643,
+      "step": 669
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.460916757583618,
+      "learning_rate": 2e-05,
+      "loss": 0.08740199,
+      "step": 670
+    },
+    {
+      "epoch": 1.342,
+      "grad_norm": 2.4553704261779785,
+      "learning_rate": 2e-05,
+      "loss": 0.10862092,
+      "step": 671
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 1.6441093683242798,
+      "learning_rate": 2e-05,
+      "loss": 0.06703743,
+      "step": 672
+    },
+    {
+      "epoch": 1.346,
+      "grad_norm": 2.292106866836548,
+      "learning_rate": 2e-05,
+      "loss": 0.10088767,
+      "step": 673
+    },
+    {
+      "epoch": 1.3479999999999999,
+      "grad_norm": 2.2479658126831055,
+      "learning_rate": 2e-05,
+      "loss": 0.09034087,
+      "step": 674
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.2497737407684326,
+      "learning_rate": 2e-05,
+      "loss": 0.11896604,
+      "step": 675
+    },
+    {
+      "epoch": 1.3519999999999999,
+      "grad_norm": 2.4198803901672363,
+      "learning_rate": 2e-05,
+      "loss": 0.09351024,
+      "step": 676
+    },
+    {
+      "epoch": 1.354,
+      "grad_norm": 2.0135796070098877,
+      "learning_rate": 2e-05,
+      "loss": 0.08445922,
+      "step": 677
+    },
+    {
+      "epoch": 1.3559999999999999,
+      "grad_norm": 3.221177339553833,
+      "learning_rate": 2e-05,
+      "loss": 0.11014754,
+      "step": 678
+    },
+    {
+      "epoch": 1.358,
+      "grad_norm": 2.270320415496826,
+      "learning_rate": 2e-05,
+      "loss": 0.12037845,
+      "step": 679
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 2.6250479221343994,
+      "learning_rate": 2e-05,
+      "loss": 0.11658848,
+      "step": 680
+    },
+    {
+      "epoch": 1.362,
+      "grad_norm": 2.4321134090423584,
+      "learning_rate": 2e-05,
+      "loss": 0.09802853,
+      "step": 681
+    },
+    {
+      "epoch": 1.3639999999999999,
+      "grad_norm": 2.4564032554626465,
+      "learning_rate": 2e-05,
+      "loss": 0.10415519,
+      "step": 682
+    },
+    {
+      "epoch": 1.366,
+      "grad_norm": 1.970099925994873,
+      "learning_rate": 2e-05,
+      "loss": 0.08627912,
+      "step": 683
+    },
+    {
+      "epoch": 1.3679999999999999,
+      "grad_norm": 2.5896036624908447,
+      "learning_rate": 2e-05,
+      "loss": 0.0935052,
+      "step": 684
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.6092915534973145,
+      "learning_rate": 2e-05,
+      "loss": 0.09677845,
+      "step": 685
+    },
+    {
+      "epoch": 1.3719999999999999,
+      "grad_norm": 3.095132827758789,
+      "learning_rate": 2e-05,
+      "loss": 0.11584838,
+      "step": 686
+    },
+    {
+      "epoch": 1.374,
+      "grad_norm": 2.705355167388916,
+      "learning_rate": 2e-05,
+      "loss": 0.10224438,
+      "step": 687
+    },
+    {
+      "epoch": 1.376,
+      "grad_norm": 1.8498618602752686,
+      "learning_rate": 2e-05,
+      "loss": 0.07257505,
+      "step": 688
+    },
+    {
+      "epoch": 1.3780000000000001,
+      "grad_norm": 2.658275842666626,
+      "learning_rate": 2e-05,
+      "loss": 0.095241,
+      "step": 689
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1657066345214844,
+      "learning_rate": 2e-05,
+      "loss": 0.08493382,
+      "step": 690
+    },
+    {
+      "epoch": 1.3820000000000001,
+      "grad_norm": 2.649157762527466,
+      "learning_rate": 2e-05,
+      "loss": 0.10965092,
+      "step": 691
+    },
+    {
+      "epoch": 1.384,
+      "grad_norm": 2.6509430408477783,
+      "learning_rate": 2e-05,
+      "loss": 0.10417461,
+      "step": 692
+    },
+    {
+      "epoch": 1.3860000000000001,
+      "grad_norm": 2.7123641967773438,
+      "learning_rate": 2e-05,
+      "loss": 0.0952239,
+      "step": 693
+    },
+    {
+      "epoch": 1.388,
+      "grad_norm": 3.1654226779937744,
+      "learning_rate": 2e-05,
+      "loss": 0.1278252,
+      "step": 694
+    },
+    {
+      "epoch": 1.3900000000000001,
+      "grad_norm": 3.4855122566223145,
+      "learning_rate": 2e-05,
+      "loss": 0.10266776,
+      "step": 695
+    },
+    {
+      "epoch": 1.392,
+      "grad_norm": 1.523314118385315,
+      "learning_rate": 2e-05,
+      "loss": 0.05538701,
+      "step": 696
+    },
+    {
+      "epoch": 1.3940000000000001,
+      "grad_norm": 2.838423490524292,
+      "learning_rate": 2e-05,
+      "loss": 0.11393388,
+      "step": 697
+    },
+    {
+      "epoch": 1.396,
+      "grad_norm": 3.0423550605773926,
+      "learning_rate": 2e-05,
+      "loss": 0.11277989,
+      "step": 698
+    },
+    {
+      "epoch": 1.3980000000000001,
+      "grad_norm": 2.595038414001465,
+      "learning_rate": 2e-05,
+      "loss": 0.1143982,
+      "step": 699
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.106015682220459,
+      "learning_rate": 2e-05,
+      "loss": 0.08432181,
+      "step": 700
+    },
+    {
+      "epoch": 1.4020000000000001,
+      "grad_norm": 2.072103977203369,
+      "learning_rate": 2e-05,
+      "loss": 0.10295324,
+      "step": 701
+    },
+    {
+      "epoch": 1.404,
+      "grad_norm": 5.35106086730957,
+      "learning_rate": 2e-05,
+      "loss": 0.12211221,
+      "step": 702
+    },
+    {
+      "epoch": 1.4060000000000001,
+      "grad_norm": 2.595242738723755,
+      "learning_rate": 2e-05,
+      "loss": 0.11621975,
+      "step": 703
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 2.3813533782958984,
+      "learning_rate": 2e-05,
+      "loss": 0.07120501,
+      "step": 704
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.7448511123657227,
+      "learning_rate": 2e-05,
+      "loss": 0.08237515,
+      "step": 705
+    },
+    {
+      "epoch": 1.412,
+      "grad_norm": 2.4283218383789062,
+      "learning_rate": 2e-05,
+      "loss": 0.07894477,
+      "step": 706
+    },
+    {
+      "epoch": 1.414,
+      "grad_norm": 2.722013235092163,
+      "learning_rate": 2e-05,
+      "loss": 0.09057313,
+      "step": 707
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 2.1725480556488037,
+      "learning_rate": 2e-05,
+      "loss": 0.07726441,
+      "step": 708
+    },
+    {
+      "epoch": 1.418,
+      "grad_norm": 2.412940502166748,
+      "learning_rate": 2e-05,
+      "loss": 0.07639,
+      "step": 709
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8505913019180298,
+      "learning_rate": 2e-05,
+      "loss": 0.09286863,
+      "step": 710
+    },
+    {
+      "epoch": 1.422,
+      "grad_norm": 2.121731996536255,
+      "learning_rate": 2e-05,
+      "loss": 0.07919374,
+      "step": 711
+    },
+    {
+      "epoch": 1.424,
+      "grad_norm": 2.7666337490081787,
+      "learning_rate": 2e-05,
+      "loss": 0.10220705,
+      "step": 712
+    },
+    {
+      "epoch": 1.426,
+      "grad_norm": 2.482006072998047,
+      "learning_rate": 2e-05,
+      "loss": 0.08181632,
+      "step": 713
+    },
+    {
+      "epoch": 1.428,
+      "grad_norm": 2.693687915802002,
+      "learning_rate": 2e-05,
+      "loss": 0.06798096,
+      "step": 714
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.5155346393585205,
+      "learning_rate": 2e-05,
+      "loss": 0.08410256,
+      "step": 715
+    },
+    {
+      "epoch": 1.432,
+      "grad_norm": 3.095309257507324,
+      "learning_rate": 2e-05,
+      "loss": 0.1147206,
+      "step": 716
+    },
+    {
+      "epoch": 1.434,
+      "grad_norm": 2.1130447387695312,
+      "learning_rate": 2e-05,
+      "loss": 0.08827046,
+      "step": 717
+    },
+    {
+      "epoch": 1.436,
+      "grad_norm": 3.6553380489349365,
+      "learning_rate": 2e-05,
+      "loss": 0.09300891,
+      "step": 718
+    },
+    {
+      "epoch": 1.438,
+      "grad_norm": 2.1224894523620605,
+      "learning_rate": 2e-05,
+      "loss": 0.09277943,
+      "step": 719
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9273908138275146,
+      "learning_rate": 2e-05,
+      "loss": 0.08951817,
+      "step": 720
+    },
+    {
+      "epoch": 1.442,
+      "grad_norm": 2.018853187561035,
+      "learning_rate": 2e-05,
+      "loss": 0.07283738,
+      "step": 721
+    },
+    {
+      "epoch": 1.444,
+      "grad_norm": 2.079155206680298,
+      "learning_rate": 2e-05,
+      "loss": 0.08982038,
+      "step": 722
+    },
+    {
+      "epoch": 1.446,
+      "grad_norm": 2.438000440597534,
+      "learning_rate": 2e-05,
+      "loss": 0.10790306,
+      "step": 723
+    },
+    {
+      "epoch": 1.448,
+      "grad_norm": 2.1911613941192627,
+      "learning_rate": 2e-05,
+      "loss": 0.07532462,
+      "step": 724
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8854244947433472,
+      "learning_rate": 2e-05,
+      "loss": 0.05531625,
+      "step": 725
+    },
+    {
+      "epoch": 1.452,
+      "grad_norm": 2.0435564517974854,
+      "learning_rate": 2e-05,
+      "loss": 0.08569458,
+      "step": 726
+    },
+    {
+      "epoch": 1.454,
+      "grad_norm": 4.157101631164551,
+      "learning_rate": 2e-05,
+      "loss": 0.10588066,
+      "step": 727
+    },
+    {
+      "epoch": 1.456,
+      "grad_norm": 1.762305498123169,
+      "learning_rate": 2e-05,
+      "loss": 0.08237082,
+      "step": 728
+    },
+    {
+      "epoch": 1.458,
+      "grad_norm": 1.7560533285140991,
+      "learning_rate": 2e-05,
+      "loss": 0.08737503,
+      "step": 729
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.803896427154541,
+      "learning_rate": 2e-05,
+      "loss": 0.07071456,
+      "step": 730
+    },
+    {
+      "epoch": 1.462,
+      "grad_norm": 2.671704053878784,
+      "learning_rate": 2e-05,
+      "loss": 0.08183609,
+      "step": 731
+    },
+    {
+      "epoch": 1.464,
+      "grad_norm": 2.725677728652954,
+      "learning_rate": 2e-05,
+      "loss": 0.09709771,
+      "step": 732
+    },
+    {
+      "epoch": 1.466,
+      "grad_norm": 2.230529308319092,
+      "learning_rate": 2e-05,
+      "loss": 0.093163,
+      "step": 733
+    },
+    {
+      "epoch": 1.468,
+      "grad_norm": 2.1992108821868896,
+      "learning_rate": 2e-05,
+      "loss": 0.08508515,
+      "step": 734
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9853242635726929,
+      "learning_rate": 2e-05,
+      "loss": 0.07874966,
+      "step": 735
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 3.195524215698242,
+      "learning_rate": 2e-05,
+      "loss": 0.07465832,
+      "step": 736
+    },
+    {
+      "epoch": 1.474,
+      "grad_norm": 2.578688383102417,
+      "learning_rate": 2e-05,
+      "loss": 0.0810234,
+      "step": 737
+    },
+    {
+      "epoch": 1.476,
+      "grad_norm": 2.219371795654297,
+      "learning_rate": 2e-05,
+      "loss": 0.06290153,
+      "step": 738
+    },
+    {
+      "epoch": 1.478,
+      "grad_norm": 3.0493226051330566,
+      "learning_rate": 2e-05,
+      "loss": 0.10852329,
+      "step": 739
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.894547700881958,
+      "learning_rate": 2e-05,
+      "loss": 0.07198999,
+      "step": 740
+    },
+    {
+      "epoch": 1.482,
+      "grad_norm": 1.9765349626541138,
+      "learning_rate": 2e-05,
+      "loss": 0.07682022,
+      "step": 741
+    },
+    {
+      "epoch": 1.484,
+      "grad_norm": 2.758671522140503,
+      "learning_rate": 2e-05,
+      "loss": 0.08025845,
+      "step": 742
+    },
+    {
+      "epoch": 1.486,
+      "grad_norm": 5.092621803283691,
+      "learning_rate": 2e-05,
+      "loss": 0.11716142,
+      "step": 743
+    },
+    {
+      "epoch": 1.488,
+      "grad_norm": 1.7337188720703125,
+      "learning_rate": 2e-05,
+      "loss": 0.06889806,
+      "step": 744
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.6227397918701172,
+      "learning_rate": 2e-05,
+      "loss": 0.06564939,
+      "step": 745
+    },
+    {
+      "epoch": 1.492,
+      "grad_norm": 1.8530882596969604,
+      "learning_rate": 2e-05,
+      "loss": 0.08244382,
+      "step": 746
+    },
+    {
+      "epoch": 1.494,
+      "grad_norm": 2.1974449157714844,
+      "learning_rate": 2e-05,
+      "loss": 0.10057726,
+      "step": 747
+    },
+    {
+      "epoch": 1.496,
+      "grad_norm": 1.9001625776290894,
+      "learning_rate": 2e-05,
+      "loss": 0.07622299,
+      "step": 748
+    },
+    {
+      "epoch": 1.498,
+      "grad_norm": 2.7725989818573,
+      "learning_rate": 2e-05,
+      "loss": 0.09877149,
+      "step": 749
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.439131498336792,
+      "learning_rate": 2e-05,
+      "loss": 0.08865526,
+      "step": 750
+    },
+    {
+      "epoch": 1.502,
+      "grad_norm": 1.9627726078033447,
+      "learning_rate": 2e-05,
+      "loss": 0.0812942,
+      "step": 751
+    },
+    {
+      "epoch": 1.504,
+      "grad_norm": 2.461146831512451,
+      "learning_rate": 2e-05,
+      "loss": 0.08872318,
+      "step": 752
+    },
+    {
+      "epoch": 1.506,
+      "grad_norm": 2.2327353954315186,
+      "learning_rate": 2e-05,
+      "loss": 0.08371209,
+      "step": 753
+    },
+    {
+      "epoch": 1.508,
+      "grad_norm": 1.7057092189788818,
+      "learning_rate": 2e-05,
+      "loss": 0.08295096,
+      "step": 754
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.592014193534851,
+      "learning_rate": 2e-05,
+      "loss": 0.06427182,
+      "step": 755
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 2.1517910957336426,
+      "learning_rate": 2e-05,
+      "loss": 0.07752314,
+      "step": 756
+    },
+    {
+      "epoch": 1.514,
+      "grad_norm": 3.956221580505371,
+      "learning_rate": 2e-05,
+      "loss": 0.08947916,
+      "step": 757
+    },
+    {
+      "epoch": 1.516,
+      "grad_norm": 2.7721030712127686,
+      "learning_rate": 2e-05,
+      "loss": 0.0844971,
+      "step": 758
+    },
+    {
+      "epoch": 1.518,
+      "grad_norm": 1.7479230165481567,
+      "learning_rate": 2e-05,
+      "loss": 0.07433778,
+      "step": 759
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.2564239501953125,
+      "learning_rate": 2e-05,
+      "loss": 0.08172427,
+      "step": 760
+    },
+    {
+      "epoch": 1.522,
+      "grad_norm": 2.6283817291259766,
+      "learning_rate": 2e-05,
+      "loss": 0.07976112,
+      "step": 761
+    },
+    {
+      "epoch": 1.524,
+      "grad_norm": 2.783998489379883,
+      "learning_rate": 2e-05,
+      "loss": 0.07664889,
+      "step": 762
+    },
+    {
+      "epoch": 1.526,
+      "grad_norm": 3.343437671661377,
+      "learning_rate": 2e-05,
+      "loss": 0.07145402,
+      "step": 763
+    },
+    {
+      "epoch": 1.528,
+      "grad_norm": 2.8911867141723633,
+      "learning_rate": 2e-05,
+      "loss": 0.06805974,
+      "step": 764
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.998328685760498,
+      "learning_rate": 2e-05,
+      "loss": 0.07499643,
+      "step": 765
+    },
+    {
+      "epoch": 1.532,
+      "grad_norm": 2.484273672103882,
+      "learning_rate": 2e-05,
+      "loss": 0.07433078,
+      "step": 766
+    },
+    {
+      "epoch": 1.534,
+      "grad_norm": 2.329362392425537,
+      "learning_rate": 2e-05,
+      "loss": 0.08358801,
+      "step": 767
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 2.202545642852783,
+      "learning_rate": 2e-05,
+      "loss": 0.06542914,
+      "step": 768
+    },
+    {
+      "epoch": 1.538,
+      "grad_norm": 2.0014524459838867,
+      "learning_rate": 2e-05,
+      "loss": 0.08993404,
+      "step": 769
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.2723772525787354,
+      "learning_rate": 2e-05,
+      "loss": 0.10550876,
+      "step": 770
+    },
+    {
+      "epoch": 1.542,
+      "grad_norm": 2.0692672729492188,
+      "learning_rate": 2e-05,
+      "loss": 0.07269676,
+      "step": 771
+    },
+    {
+      "epoch": 1.544,
+      "grad_norm": 2.0737030506134033,
+      "learning_rate": 2e-05,
+      "loss": 0.06703752,
+      "step": 772
+    },
+    {
+      "epoch": 1.546,
+      "grad_norm": 2.1744225025177,
+      "learning_rate": 2e-05,
+      "loss": 0.08306466,
+      "step": 773
+    },
+    {
+      "epoch": 1.548,
+      "grad_norm": 17.611337661743164,
+      "learning_rate": 2e-05,
+      "loss": 0.064059,
+      "step": 774
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.8644952774047852,
+      "learning_rate": 2e-05,
+      "loss": 0.07994568,
+      "step": 775
+    },
+    {
+      "epoch": 1.552,
+      "grad_norm": 1.5328989028930664,
+      "learning_rate": 2e-05,
+      "loss": 0.07587386,
+      "step": 776
+    },
+    {
+      "epoch": 1.554,
+      "grad_norm": 2.235229015350342,
+      "learning_rate": 2e-05,
+      "loss": 0.08034088,
+      "step": 777
+    },
+    {
+      "epoch": 1.556,
+      "grad_norm": 2.2400894165039062,
+      "learning_rate": 2e-05,
+      "loss": 0.07154867,
+      "step": 778
+    },
+    {
+      "epoch": 1.558,
+      "grad_norm": 2.6439497470855713,
+      "learning_rate": 2e-05,
+      "loss": 0.08810341,
+      "step": 779
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.811596632003784,
+      "learning_rate": 2e-05,
+      "loss": 0.09437238,
+      "step": 780
+    },
+    {
+      "epoch": 1.562,
+      "grad_norm": 2.309126138687134,
+      "learning_rate": 2e-05,
+      "loss": 0.07691263,
+      "step": 781
+    },
+    {
+      "epoch": 1.564,
+      "grad_norm": 2.3875679969787598,
+      "learning_rate": 2e-05,
+      "loss": 0.07415307,
+      "step": 782
+    },
+    {
+      "epoch": 1.5659999999999998,
+      "grad_norm": 1.5229976177215576,
+      "learning_rate": 2e-05,
+      "loss": 0.0736395,
+      "step": 783
+    },
+    {
+      "epoch": 1.568,
+      "grad_norm": 3.096762180328369,
+      "learning_rate": 2e-05,
+      "loss": 0.09484315,
+      "step": 784
+    },
+    {
+      "epoch": 1.5699999999999998,
+      "grad_norm": 1.823845386505127,
+      "learning_rate": 2e-05,
+      "loss": 0.08399117,
+      "step": 785
+    },
+    {
+      "epoch": 1.572,
+      "grad_norm": 2.103595018386841,
+      "learning_rate": 2e-05,
+      "loss": 0.060688,
+      "step": 786
+    },
+    {
+      "epoch": 1.5739999999999998,
+      "grad_norm": 2.618736505508423,
+      "learning_rate": 2e-05,
+      "loss": 0.08902743,
+      "step": 787
+    },
+    {
+      "epoch": 1.576,
+      "grad_norm": 2.025015115737915,
+      "learning_rate": 2e-05,
+      "loss": 0.08033721,
+      "step": 788
+    },
+    {
+      "epoch": 1.5779999999999998,
+      "grad_norm": 9.556914329528809,
+      "learning_rate": 2e-05,
+      "loss": 0.11010626,
+      "step": 789
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.4782986640930176,
+      "learning_rate": 2e-05,
+      "loss": 0.12129696,
+      "step": 790
+    },
+    {
+      "epoch": 1.5819999999999999,
+      "grad_norm": 2.86425518989563,
+      "learning_rate": 2e-05,
+      "loss": 0.08272199,
+      "step": 791
+    },
+    {
+      "epoch": 1.584,
+      "grad_norm": 1.8563334941864014,
+      "learning_rate": 2e-05,
+      "loss": 0.06831618,
+      "step": 792
+    },
+    {
+      "epoch": 1.5859999999999999,
+      "grad_norm": 1.709964394569397,
+      "learning_rate": 2e-05,
+      "loss": 0.06362506,
+      "step": 793
+    },
+    {
+      "epoch": 1.588,
+      "grad_norm": 2.137883186340332,
+      "learning_rate": 2e-05,
+      "loss": 0.07587803,
+      "step": 794
+    },
+    {
+      "epoch": 1.5899999999999999,
+      "grad_norm": 2.363016128540039,
+      "learning_rate": 2e-05,
+      "loss": 0.10132524,
+      "step": 795
+    },
+    {
+      "epoch": 1.592,
+      "grad_norm": 2.9902260303497314,
+      "learning_rate": 2e-05,
+      "loss": 0.12351868,
+      "step": 796
+    },
+    {
+      "epoch": 1.5939999999999999,
+      "grad_norm": 1.7475019693374634,
+      "learning_rate": 2e-05,
+      "loss": 0.0698937,
+      "step": 797
+    },
+    {
+      "epoch": 1.596,
+      "grad_norm": 1.7452584505081177,
+      "learning_rate": 2e-05,
+      "loss": 0.07310887,
+      "step": 798
+    },
+    {
+      "epoch": 1.5979999999999999,
+      "grad_norm": 2.2125442028045654,
+      "learning_rate": 2e-05,
+      "loss": 0.08515593,
+      "step": 799
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.9570682048797607,
+      "learning_rate": 2e-05,
+      "loss": 0.09576114,
+      "step": 800
+    },
+    {
+      "epoch": 1.6019999999999999,
+      "grad_norm": 2.4968690872192383,
+      "learning_rate": 2e-05,
+      "loss": 0.08693783,
+      "step": 801
+    },
+    {
+      "epoch": 1.604,
+      "grad_norm": 2.26326847076416,
+      "learning_rate": 2e-05,
+      "loss": 0.08654219,
+      "step": 802
+    },
+    {
+      "epoch": 1.6059999999999999,
+      "grad_norm": 2.3274145126342773,
+      "learning_rate": 2e-05,
+      "loss": 0.10750651,
+      "step": 803
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 1.52241849899292,
+      "learning_rate": 2e-05,
+      "loss": 0.07020888,
+      "step": 804
+    },
+    {
+      "epoch": 1.6099999999999999,
+      "grad_norm": 2.112908363342285,
+      "learning_rate": 2e-05,
+      "loss": 0.07239047,
+      "step": 805
+    },
+    {
+      "epoch": 1.612,
+      "grad_norm": 2.029871702194214,
+      "learning_rate": 2e-05,
+      "loss": 0.08264256,
+      "step": 806
+    },
+    {
+      "epoch": 1.6139999999999999,
+      "grad_norm": 1.7689646482467651,
+      "learning_rate": 2e-05,
+      "loss": 0.08204137,
+      "step": 807
+    },
+    {
+      "epoch": 1.616,
+      "grad_norm": 2.38230562210083,
+      "learning_rate": 2e-05,
+      "loss": 0.07087668,
+      "step": 808
+    },
+    {
+      "epoch": 1.6179999999999999,
+      "grad_norm": 1.8665469884872437,
+      "learning_rate": 2e-05,
+      "loss": 0.0778988,
+      "step": 809
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.146146774291992,
+      "learning_rate": 2e-05,
+      "loss": 0.08147563,
+      "step": 810
+    },
+    {
+      "epoch": 1.6219999999999999,
+      "grad_norm": 1.901824951171875,
+      "learning_rate": 2e-05,
+      "loss": 0.07722423,
+      "step": 811
+    },
+    {
+      "epoch": 1.624,
+      "grad_norm": 1.654176950454712,
+      "learning_rate": 2e-05,
+      "loss": 0.07397038,
+      "step": 812
+    },
+    {
+      "epoch": 1.626,
+      "grad_norm": 2.2095117568969727,
+      "learning_rate": 2e-05,
+      "loss": 0.09214038,
+      "step": 813
+    },
+    {
+      "epoch": 1.6280000000000001,
+      "grad_norm": 2.471964120864868,
+      "learning_rate": 2e-05,
+      "loss": 0.08440986,
+      "step": 814
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.14119815826416,
+      "learning_rate": 2e-05,
+      "loss": 0.08491719,
+      "step": 815
+    },
+    {
+      "epoch": 1.6320000000000001,
+      "grad_norm": 1.472819209098816,
+      "learning_rate": 2e-05,
+      "loss": 0.06080973,
+      "step": 816
+    },
+    {
+      "epoch": 1.634,
+      "grad_norm": 4.0974297523498535,
+      "learning_rate": 2e-05,
+      "loss": 0.11212557,
+      "step": 817
+    },
+    {
+      "epoch": 1.6360000000000001,
+      "grad_norm": 1.893890142440796,
+      "learning_rate": 2e-05,
+      "loss": 0.06430957,
+      "step": 818
+    },
+    {
+      "epoch": 1.638,
+      "grad_norm": 1.8362897634506226,
+      "learning_rate": 2e-05,
+      "loss": 0.07269567,
+      "step": 819
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 2.3635354042053223,
+      "learning_rate": 2e-05,
+      "loss": 0.07982106,
+      "step": 820
+    },
+    {
+      "epoch": 1.642,
+      "grad_norm": 2.827873706817627,
+      "learning_rate": 2e-05,
+      "loss": 0.06442684,
+      "step": 821
+    },
+    {
+      "epoch": 1.6440000000000001,
+      "grad_norm": 1.9763258695602417,
+      "learning_rate": 2e-05,
+      "loss": 0.0594982,
+      "step": 822
+    },
+    {
+      "epoch": 1.646,
+      "grad_norm": 1.7666466236114502,
+      "learning_rate": 2e-05,
+      "loss": 0.06877147,
+      "step": 823
+    },
+    {
+      "epoch": 1.6480000000000001,
+      "grad_norm": 3.103104829788208,
+      "learning_rate": 2e-05,
+      "loss": 0.07696585,
+      "step": 824
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.5277068614959717,
+      "learning_rate": 2e-05,
+      "loss": 0.08776857,
+      "step": 825
+    },
+    {
+      "epoch": 1.6520000000000001,
+      "grad_norm": 2.6822338104248047,
+      "learning_rate": 2e-05,
+      "loss": 0.06033723,
+      "step": 826
+    },
+    {
+      "epoch": 1.654,
+      "grad_norm": 1.966208577156067,
+      "learning_rate": 2e-05,
+      "loss": 0.0782545,
+      "step": 827
+    },
+    {
+      "epoch": 1.6560000000000001,
+      "grad_norm": 2.1555910110473633,
+      "learning_rate": 2e-05,
+      "loss": 0.08536707,
+      "step": 828
+    },
+    {
+      "epoch": 1.658,
+      "grad_norm": 2.096773624420166,
+      "learning_rate": 2e-05,
+      "loss": 0.06903753,
+      "step": 829
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 1.9013687372207642,
+      "learning_rate": 2e-05,
+      "loss": 0.08171882,
+      "step": 830
+    },
+    {
+      "epoch": 1.662,
+      "grad_norm": 1.719090461730957,
+      "learning_rate": 2e-05,
+      "loss": 0.06542725,
+      "step": 831
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 2.0266594886779785,
+      "learning_rate": 2e-05,
+      "loss": 0.08807792,
+      "step": 832
+    },
+    {
+      "epoch": 1.666,
+      "grad_norm": 2.070847511291504,
+      "learning_rate": 2e-05,
+      "loss": 0.07473135,
+      "step": 833
+    },
+    {
+      "epoch": 1.6680000000000001,
+      "grad_norm": 2.2031726837158203,
+      "learning_rate": 2e-05,
+      "loss": 0.05795466,
+      "step": 834
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.9450498819351196,
+      "learning_rate": 2e-05,
+      "loss": 0.05321917,
+      "step": 835
+    },
+    {
+      "epoch": 1.6720000000000002,
+      "grad_norm": 2.383620023727417,
+      "learning_rate": 2e-05,
+      "loss": 0.09668325,
+      "step": 836
+    },
+    {
+      "epoch": 1.674,
+      "grad_norm": 1.911126732826233,
+      "learning_rate": 2e-05,
+      "loss": 0.0709563,
+      "step": 837
+    },
+    {
+      "epoch": 1.6760000000000002,
+      "grad_norm": 1.8981558084487915,
+      "learning_rate": 2e-05,
+      "loss": 0.05752856,
+      "step": 838
+    },
+    {
+      "epoch": 1.678,
+      "grad_norm": 1.5697596073150635,
+      "learning_rate": 2e-05,
+      "loss": 0.06154868,
+      "step": 839
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 2.2556140422821045,
+      "learning_rate": 2e-05,
+      "loss": 0.06484008,
+      "step": 840
+    },
+    {
+      "epoch": 1.682,
+      "grad_norm": 2.297083616256714,
+      "learning_rate": 2e-05,
+      "loss": 0.06691624,
+      "step": 841
+    },
+    {
+      "epoch": 1.6840000000000002,
+      "grad_norm": 2.831334114074707,
+      "learning_rate": 2e-05,
+      "loss": 0.0824251,
+      "step": 842
+    },
+    {
+      "epoch": 1.686,
+      "grad_norm": 2.1341090202331543,
+      "learning_rate": 2e-05,
+      "loss": 0.08404815,
+      "step": 843
+    },
+    {
+      "epoch": 1.688,
+      "grad_norm": 1.4432538747787476,
+      "learning_rate": 2e-05,
+      "loss": 0.05486818,
+      "step": 844
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.392119884490967,
+      "learning_rate": 2e-05,
+      "loss": 0.08042577,
+      "step": 845
+    },
+    {
+      "epoch": 1.692,
+      "grad_norm": 2.140087604522705,
+      "learning_rate": 2e-05,
+      "loss": 0.07316266,
+      "step": 846
+    },
+    {
+      "epoch": 1.694,
+      "grad_norm": 2.8727564811706543,
+      "learning_rate": 2e-05,
+      "loss": 0.09117547,
+      "step": 847
+    },
+    {
+      "epoch": 1.696,
+      "grad_norm": 2.5479319095611572,
+      "learning_rate": 2e-05,
+      "loss": 0.0947205,
+      "step": 848
+    },
+    {
+      "epoch": 1.698,
+      "grad_norm": 2.120222806930542,
+      "learning_rate": 2e-05,
+      "loss": 0.06924388,
+      "step": 849
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.4992620944976807,
+      "learning_rate": 2e-05,
+      "loss": 0.09288818,
+      "step": 850
+    },
+    {
+      "epoch": 1.702,
+      "grad_norm": 1.9825425148010254,
+      "learning_rate": 2e-05,
+      "loss": 0.07824534,
+      "step": 851
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 1.519212245941162,
+      "learning_rate": 2e-05,
+      "loss": 0.05538861,
+      "step": 852
+    },
+    {
+      "epoch": 1.706,
+      "grad_norm": 2.399399518966675,
+      "learning_rate": 2e-05,
+      "loss": 0.0779898,
+      "step": 853
+    },
+    {
+      "epoch": 1.708,
+      "grad_norm": 2.637857437133789,
+      "learning_rate": 2e-05,
+      "loss": 0.10369846,
+      "step": 854
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.903981924057007,
+      "learning_rate": 2e-05,
+      "loss": 0.09455149,
+      "step": 855
+    },
+    {
+      "epoch": 1.712,
+      "grad_norm": 1.9127029180526733,
+      "learning_rate": 2e-05,
+      "loss": 0.06748534,
+      "step": 856
+    },
+    {
+      "epoch": 1.714,
+      "grad_norm": 1.633865237236023,
+      "learning_rate": 2e-05,
+      "loss": 0.07176484,
+      "step": 857
+    },
+    {
+      "epoch": 1.716,
+      "grad_norm": 2.8449621200561523,
+      "learning_rate": 2e-05,
+      "loss": 0.07032914,
+      "step": 858
+    },
+    {
+      "epoch": 1.718,
+      "grad_norm": 3.111711263656616,
+      "learning_rate": 2e-05,
+      "loss": 0.07360119,
+      "step": 859
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.459895610809326,
+      "learning_rate": 2e-05,
+      "loss": 0.10556261,
+      "step": 860
+    },
+    {
+      "epoch": 1.722,
+      "grad_norm": 1.9931344985961914,
+      "learning_rate": 2e-05,
+      "loss": 0.06967217,
+      "step": 861
+    },
+    {
+      "epoch": 1.724,
+      "grad_norm": 1.429177165031433,
+      "learning_rate": 2e-05,
+      "loss": 0.04311872,
+      "step": 862
+    },
+    {
+      "epoch": 1.726,
+      "grad_norm": 1.8553142547607422,
+      "learning_rate": 2e-05,
+      "loss": 0.07448836,
+      "step": 863
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 1.4926773309707642,
+      "learning_rate": 2e-05,
+      "loss": 0.0673956,
+      "step": 864
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.5575432777404785,
+      "learning_rate": 2e-05,
+      "loss": 0.07588162,
+      "step": 865
+    },
+    {
+      "epoch": 1.732,
+      "grad_norm": 2.1007280349731445,
+      "learning_rate": 2e-05,
+      "loss": 0.05418327,
+      "step": 866
+    },
+    {
+      "epoch": 1.734,
+      "grad_norm": 3.0805535316467285,
+      "learning_rate": 2e-05,
+      "loss": 0.08815256,
+      "step": 867
+    },
+    {
+      "epoch": 1.736,
+      "grad_norm": 2.14973783493042,
+      "learning_rate": 2e-05,
+      "loss": 0.07602916,
+      "step": 868
+    },
+    {
+      "epoch": 1.738,
+      "grad_norm": 2.6056084632873535,
+      "learning_rate": 2e-05,
+      "loss": 0.08369519,
+      "step": 869
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.980968952178955,
+      "learning_rate": 2e-05,
+      "loss": 0.06892893,
+      "step": 870
+    },
+    {
+      "epoch": 1.742,
+      "grad_norm": 2.932478427886963,
+      "learning_rate": 2e-05,
+      "loss": 0.08124112,
+      "step": 871
+    },
+    {
+      "epoch": 1.744,
+      "grad_norm": 1.6852787733078003,
+      "learning_rate": 2e-05,
+      "loss": 0.06830005,
+      "step": 872
+    },
+    {
+      "epoch": 1.746,
+      "grad_norm": 1.881437063217163,
+      "learning_rate": 2e-05,
+      "loss": 0.04541746,
+      "step": 873
+    },
+    {
+      "epoch": 1.748,
+      "grad_norm": 2.5663695335388184,
+      "learning_rate": 2e-05,
+      "loss": 0.07354845,
+      "step": 874
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.884338617324829,
+      "learning_rate": 2e-05,
+      "loss": 0.07085122,
+      "step": 875
+    },
+    {
+      "epoch": 1.752,
+      "grad_norm": 2.173144578933716,
+      "learning_rate": 2e-05,
+      "loss": 0.06505815,
+      "step": 876
+    },
+    {
+      "epoch": 1.754,
+      "grad_norm": 2.744832754135132,
+      "learning_rate": 2e-05,
+      "loss": 0.10300224,
+      "step": 877
+    },
+    {
+      "epoch": 1.756,
+      "grad_norm": 2.6979944705963135,
+      "learning_rate": 2e-05,
+      "loss": 0.08862556,
+      "step": 878
+    },
+    {
+      "epoch": 1.758,
+      "grad_norm": 2.6769134998321533,
+      "learning_rate": 2e-05,
+      "loss": 0.07417379,
+      "step": 879
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.0921900272369385,
+      "learning_rate": 2e-05,
+      "loss": 0.08049569,
+      "step": 880
+    },
+    {
+      "epoch": 1.762,
+      "grad_norm": 3.668351411819458,
+      "learning_rate": 2e-05,
+      "loss": 0.07346732,
+      "step": 881
+    },
+    {
+      "epoch": 1.764,
+      "grad_norm": 1.955716609954834,
+      "learning_rate": 2e-05,
+      "loss": 0.0588823,
+      "step": 882
+    },
+    {
+      "epoch": 1.766,
+      "grad_norm": 2.055715560913086,
+      "learning_rate": 2e-05,
+      "loss": 0.06786981,
+      "step": 883
+    },
+    {
+      "epoch": 1.768,
+      "grad_norm": 1.8874297142028809,
+      "learning_rate": 2e-05,
+      "loss": 0.06704556,
+      "step": 884
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.452850103378296,
+      "learning_rate": 2e-05,
+      "loss": 0.0776285,
+      "step": 885
+    },
+    {
+      "epoch": 1.772,
+      "grad_norm": 1.6478520631790161,
+      "learning_rate": 2e-05,
+      "loss": 0.0682406,
+      "step": 886
+    },
+    {
+      "epoch": 1.774,
+      "grad_norm": 1.441072940826416,
+      "learning_rate": 2e-05,
+      "loss": 0.05489372,
+      "step": 887
+    },
+    {
+      "epoch": 1.776,
+      "grad_norm": 2.24617862701416,
+      "learning_rate": 2e-05,
+      "loss": 0.05962018,
+      "step": 888
+    },
+    {
+      "epoch": 1.778,
+      "grad_norm": 2.570594072341919,
+      "learning_rate": 2e-05,
+      "loss": 0.07772797,
+      "step": 889
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.316793441772461,
+      "learning_rate": 2e-05,
+      "loss": 0.07414915,
+      "step": 890
+    },
+    {
+      "epoch": 1.782,
+      "grad_norm": 2.2875797748565674,
+      "learning_rate": 2e-05,
+      "loss": 0.06844224,
+      "step": 891
+    },
+    {
+      "epoch": 1.784,
+      "grad_norm": 3.2194788455963135,
+      "learning_rate": 2e-05,
+      "loss": 0.08942357,
+      "step": 892
+    },
+    {
+      "epoch": 1.786,
+      "grad_norm": 3.111416816711426,
+      "learning_rate": 2e-05,
+      "loss": 0.08543706,
+      "step": 893
+    },
+    {
+      "epoch": 1.788,
+      "grad_norm": 2.2496681213378906,
+      "learning_rate": 2e-05,
+      "loss": 0.06796233,
+      "step": 894
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.393826723098755,
+      "learning_rate": 2e-05,
+      "loss": 0.06307848,
+      "step": 895
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 1.8092612028121948,
+      "learning_rate": 2e-05,
+      "loss": 0.08083902,
+      "step": 896
+    },
+    {
+      "epoch": 1.794,
+      "grad_norm": 2.327549934387207,
+      "learning_rate": 2e-05,
+      "loss": 0.07006175,
+      "step": 897
+    },
+    {
+      "epoch": 1.796,
+      "grad_norm": 2.3291375637054443,
+      "learning_rate": 2e-05,
+      "loss": 0.07644182,
+      "step": 898
+    },
+    {
+      "epoch": 1.798,
+      "grad_norm": 1.8122745752334595,
+      "learning_rate": 2e-05,
+      "loss": 0.05956836,
+      "step": 899
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7686699628829956,
+      "learning_rate": 2e-05,
+      "loss": 0.0774643,
+      "step": 900
+    },
+    {
+      "epoch": 1.802,
+      "grad_norm": 3.383394956588745,
+      "learning_rate": 2e-05,
+      "loss": 0.09732038,
+      "step": 901
+    },
+    {
+      "epoch": 1.804,
+      "grad_norm": 1.4468183517456055,
+      "learning_rate": 2e-05,
+      "loss": 0.05280242,
+      "step": 902
+    },
+    {
+      "epoch": 1.806,
+      "grad_norm": 3.099715232849121,
+      "learning_rate": 2e-05,
+      "loss": 0.08549714,
+      "step": 903
+    },
+    {
+      "epoch": 1.808,
+      "grad_norm": 2.2685530185699463,
+      "learning_rate": 2e-05,
+      "loss": 0.08650636,
+      "step": 904
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.1206886768341064,
+      "learning_rate": 2e-05,
+      "loss": 0.06690793,
+      "step": 905
+    },
+    {
+      "epoch": 1.812,
+      "grad_norm": 2.361654281616211,
+      "learning_rate": 2e-05,
+      "loss": 0.07197607,
+      "step": 906
+    },
+    {
+      "epoch": 1.814,
+      "grad_norm": 2.3998827934265137,
+      "learning_rate": 2e-05,
+      "loss": 0.06678525,
+      "step": 907
+    },
+    {
+      "epoch": 1.8159999999999998,
+      "grad_norm": 1.8265810012817383,
+      "learning_rate": 2e-05,
+      "loss": 0.06269243,
+      "step": 908
+    },
+    {
+      "epoch": 1.818,
+      "grad_norm": 2.0034234523773193,
+      "learning_rate": 2e-05,
+      "loss": 0.07594454,
+      "step": 909
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 1.8148810863494873,
+      "learning_rate": 2e-05,
+      "loss": 0.06071473,
+      "step": 910
+    },
+    {
+      "epoch": 1.822,
+      "grad_norm": 1.751679539680481,
+      "learning_rate": 2e-05,
+      "loss": 0.06768817,
+      "step": 911
+    },
+    {
+      "epoch": 1.8239999999999998,
+      "grad_norm": 2.00803279876709,
+      "learning_rate": 2e-05,
+      "loss": 0.07907202,
+      "step": 912
+    },
+    {
+      "epoch": 1.826,
+      "grad_norm": 2.013383388519287,
+      "learning_rate": 2e-05,
+      "loss": 0.06719398,
+      "step": 913
+    },
+    {
+      "epoch": 1.8279999999999998,
+      "grad_norm": 2.5518107414245605,
+      "learning_rate": 2e-05,
+      "loss": 0.07807682,
+      "step": 914
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.0092356204986572,
+      "learning_rate": 2e-05,
+      "loss": 0.05918414,
+      "step": 915
+    },
+    {
+      "epoch": 1.8319999999999999,
+      "grad_norm": 1.9702318906784058,
+      "learning_rate": 2e-05,
+      "loss": 0.05898762,
+      "step": 916
+    },
+    {
+      "epoch": 1.834,
+      "grad_norm": 2.974558115005493,
+      "learning_rate": 2e-05,
+      "loss": 0.09832944,
+      "step": 917
+    },
+    {
+      "epoch": 1.8359999999999999,
+      "grad_norm": 3.1490416526794434,
+      "learning_rate": 2e-05,
+      "loss": 0.08504972,
+      "step": 918
+    },
+    {
+      "epoch": 1.838,
+      "grad_norm": 3.151014804840088,
+      "learning_rate": 2e-05,
+      "loss": 0.05736318,
+      "step": 919
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 2.9925057888031006,
+      "learning_rate": 2e-05,
+      "loss": 0.07320988,
+      "step": 920
+    },
+    {
+      "epoch": 1.842,
+      "grad_norm": 2.636167287826538,
+      "learning_rate": 2e-05,
+      "loss": 0.09282872,
+      "step": 921
+    },
+    {
+      "epoch": 1.8439999999999999,
+      "grad_norm": 1.984940528869629,
+      "learning_rate": 2e-05,
+      "loss": 0.05574302,
+      "step": 922
+    },
+    {
+      "epoch": 1.846,
+      "grad_norm": 2.508833169937134,
+      "learning_rate": 2e-05,
+      "loss": 0.0758808,
+      "step": 923
+    },
+    {
+      "epoch": 1.8479999999999999,
+      "grad_norm": 2.5532920360565186,
+      "learning_rate": 2e-05,
+      "loss": 0.0828255,
+      "step": 924
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.2903261184692383,
+      "learning_rate": 2e-05,
+      "loss": 0.07507607,
+      "step": 925
+    },
+    {
+      "epoch": 1.8519999999999999,
+      "grad_norm": 2.095050096511841,
+      "learning_rate": 2e-05,
+      "loss": 0.05968579,
+      "step": 926
+    },
+    {
+      "epoch": 1.854,
+      "grad_norm": 1.5813626050949097,
+      "learning_rate": 2e-05,
+      "loss": 0.06536861,
+      "step": 927
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 1.9922809600830078,
+      "learning_rate": 2e-05,
+      "loss": 0.06198128,
+      "step": 928
+    },
+    {
+      "epoch": 1.858,
+      "grad_norm": 2.0554776191711426,
+      "learning_rate": 2e-05,
+      "loss": 0.07751165,
+      "step": 929
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 1.9760997295379639,
+      "learning_rate": 2e-05,
+      "loss": 0.0767539,
+      "step": 930
+    },
+    {
+      "epoch": 1.862,
+      "grad_norm": 1.6336450576782227,
+      "learning_rate": 2e-05,
+      "loss": 0.06495491,
+      "step": 931
+    },
+    {
+      "epoch": 1.8639999999999999,
+      "grad_norm": 1.796708583831787,
+      "learning_rate": 2e-05,
+      "loss": 0.0650342,
+      "step": 932
+    },
+    {
+      "epoch": 1.866,
+      "grad_norm": 1.6166343688964844,
+      "learning_rate": 2e-05,
+      "loss": 0.06242783,
+      "step": 933
+    },
+    {
+      "epoch": 1.8679999999999999,
+      "grad_norm": 2.2365167140960693,
+      "learning_rate": 2e-05,
+      "loss": 0.08445454,
+      "step": 934
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.795891284942627,
+      "learning_rate": 2e-05,
+      "loss": 0.0526274,
+      "step": 935
+    },
+    {
+      "epoch": 1.8719999999999999,
+      "grad_norm": 2.8587825298309326,
+      "learning_rate": 2e-05,
+      "loss": 0.06400409,
+      "step": 936
+    },
+    {
+      "epoch": 1.874,
+      "grad_norm": 2.213303565979004,
+      "learning_rate": 2e-05,
+      "loss": 0.06138328,
+      "step": 937
+    },
+    {
+      "epoch": 1.876,
+      "grad_norm": 2.25402569770813,
+      "learning_rate": 2e-05,
+      "loss": 0.07633629,
+      "step": 938
+    },
+    {
+      "epoch": 1.8780000000000001,
+      "grad_norm": 2.096250057220459,
+      "learning_rate": 2e-05,
+      "loss": 0.07336065,
+      "step": 939
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.6095550060272217,
+      "learning_rate": 2e-05,
+      "loss": 0.057545,
+      "step": 940
+    },
+    {
+      "epoch": 1.8820000000000001,
+      "grad_norm": 2.503938913345337,
+      "learning_rate": 2e-05,
+      "loss": 0.07715003,
+      "step": 941
+    },
+    {
+      "epoch": 1.884,
+      "grad_norm": 1.7810205221176147,
+      "learning_rate": 2e-05,
+      "loss": 0.06386582,
+      "step": 942
+    },
+    {
+      "epoch": 1.8860000000000001,
+      "grad_norm": 3.030251979827881,
+      "learning_rate": 2e-05,
+      "loss": 0.09021285,
+      "step": 943
+    },
+    {
+      "epoch": 1.888,
+      "grad_norm": 2.048762798309326,
+      "learning_rate": 2e-05,
+      "loss": 0.07349489,
+      "step": 944
+    },
+    {
+      "epoch": 1.8900000000000001,
+      "grad_norm": 2.1592016220092773,
+      "learning_rate": 2e-05,
+      "loss": 0.06426996,
+      "step": 945
+    },
+    {
+      "epoch": 1.892,
+      "grad_norm": 1.1694790124893188,
+      "learning_rate": 2e-05,
+      "loss": 0.05598904,
+      "step": 946
+    },
+    {
+      "epoch": 1.8940000000000001,
+      "grad_norm": 2.5215678215026855,
+      "learning_rate": 2e-05,
+      "loss": 0.05859364,
+      "step": 947
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 1.6138578653335571,
+      "learning_rate": 2e-05,
+      "loss": 0.05201582,
+      "step": 948
+    },
+    {
+      "epoch": 1.8980000000000001,
+      "grad_norm": 2.5577175617218018,
+      "learning_rate": 2e-05,
+      "loss": 0.07010947,
+      "step": 949
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.392971992492676,
+      "learning_rate": 2e-05,
+      "loss": 0.07514925,
+      "step": 950
+    },
+    {
+      "epoch": 1.9020000000000001,
+      "grad_norm": 1.6294666528701782,
+      "learning_rate": 2e-05,
+      "loss": 0.06291882,
+      "step": 951
+    },
+    {
+      "epoch": 1.904,
+      "grad_norm": 2.229604482650757,
+      "learning_rate": 2e-05,
+      "loss": 0.06559543,
+      "step": 952
+    },
+    {
+      "epoch": 1.9060000000000001,
+      "grad_norm": 1.924761414527893,
+      "learning_rate": 2e-05,
+      "loss": 0.0862481,
+      "step": 953
+    },
+    {
+      "epoch": 1.908,
+      "grad_norm": 1.9866812229156494,
+      "learning_rate": 2e-05,
+      "loss": 0.06686676,
+      "step": 954
+    },
+    {
+      "epoch": 1.9100000000000001,
+      "grad_norm": 1.7814255952835083,
+      "learning_rate": 2e-05,
+      "loss": 0.05599992,
+      "step": 955
+    },
+    {
+      "epoch": 1.912,
+      "grad_norm": 2.1436808109283447,
+      "learning_rate": 2e-05,
+      "loss": 0.07030681,
+      "step": 956
+    },
+    {
+      "epoch": 1.9140000000000001,
+      "grad_norm": 1.3862833976745605,
+      "learning_rate": 2e-05,
+      "loss": 0.04867277,
+      "step": 957
+    },
+    {
+      "epoch": 1.916,
+      "grad_norm": 1.781009316444397,
+      "learning_rate": 2e-05,
+      "loss": 0.05244367,
+      "step": 958
+    },
+    {
+      "epoch": 1.9180000000000001,
+      "grad_norm": 1.6906291246414185,
+      "learning_rate": 2e-05,
+      "loss": 0.06690039,
+      "step": 959
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.1098814010620117,
+      "learning_rate": 2e-05,
+      "loss": 0.11306722,
+      "step": 960
+    },
+    {
+      "epoch": 1.9220000000000002,
+      "grad_norm": 1.7126940488815308,
+      "learning_rate": 2e-05,
+      "loss": 0.05899595,
+      "step": 961
+    },
+    {
+      "epoch": 1.924,
+      "grad_norm": 2.6065030097961426,
+      "learning_rate": 2e-05,
+      "loss": 0.06164584,
+      "step": 962
+    },
+    {
+      "epoch": 1.9260000000000002,
+      "grad_norm": 1.9014359712600708,
+      "learning_rate": 2e-05,
+      "loss": 0.07033944,
+      "step": 963
+    },
+    {
+      "epoch": 1.928,
+      "grad_norm": 1.3948487043380737,
+      "learning_rate": 2e-05,
+      "loss": 0.05189591,
+      "step": 964
+    },
+    {
+      "epoch": 1.9300000000000002,
+      "grad_norm": 2.2040138244628906,
+      "learning_rate": 2e-05,
+      "loss": 0.05501474,
+      "step": 965
+    },
+    {
+      "epoch": 1.932,
+      "grad_norm": 1.710676908493042,
+      "learning_rate": 2e-05,
+      "loss": 0.07464767,
+      "step": 966
+    },
+    {
+      "epoch": 1.9340000000000002,
+      "grad_norm": 2.508552312850952,
+      "learning_rate": 2e-05,
+      "loss": 0.07813706,
+      "step": 967
+    },
+    {
+      "epoch": 1.936,
+      "grad_norm": 1.4470033645629883,
+      "learning_rate": 2e-05,
+      "loss": 0.0598635,
+      "step": 968
+    },
+    {
+      "epoch": 1.938,
+      "grad_norm": 2.1353957653045654,
+      "learning_rate": 2e-05,
+      "loss": 0.05462658,
+      "step": 969
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.136827230453491,
+      "learning_rate": 2e-05,
+      "loss": 0.0630064,
+      "step": 970
+    },
+    {
+      "epoch": 1.942,
+      "grad_norm": 2.935197591781616,
+      "learning_rate": 2e-05,
+      "loss": 0.09128846,
+      "step": 971
+    },
+    {
+      "epoch": 1.944,
+      "grad_norm": 2.3456857204437256,
+      "learning_rate": 2e-05,
+      "loss": 0.08022948,
+      "step": 972
+    },
+    {
+      "epoch": 1.946,
+      "grad_norm": 1.7174320220947266,
+      "learning_rate": 2e-05,
+      "loss": 0.06144142,
+      "step": 973
+    },
+    {
+      "epoch": 1.948,
+      "grad_norm": 1.8288911581039429,
+      "learning_rate": 2e-05,
+      "loss": 0.04687043,
+      "step": 974
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.6098743677139282,
+      "learning_rate": 2e-05,
+      "loss": 0.06447228,
+      "step": 975
+    },
+    {
+      "epoch": 1.952,
+      "grad_norm": 2.4248886108398438,
+      "learning_rate": 2e-05,
+      "loss": 0.11559688,
+      "step": 976
+    },
+    {
+      "epoch": 1.954,
+      "grad_norm": 2.1920695304870605,
+      "learning_rate": 2e-05,
+      "loss": 0.0743776,
+      "step": 977
+    },
+    {
+      "epoch": 1.956,
+      "grad_norm": 1.9038764238357544,
+      "learning_rate": 2e-05,
+      "loss": 0.07296964,
+      "step": 978
+    },
+    {
+      "epoch": 1.958,
+      "grad_norm": 1.6800254583358765,
+      "learning_rate": 2e-05,
+      "loss": 0.06484336,
+      "step": 979
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.069260597229004,
+      "learning_rate": 2e-05,
+      "loss": 0.07390512,
+      "step": 980
+    },
+    {
+      "epoch": 1.962,
+      "grad_norm": 2.405409336090088,
+      "learning_rate": 2e-05,
+      "loss": 0.09900182,
+      "step": 981
+    },
+    {
+      "epoch": 1.964,
+      "grad_norm": 1.6991225481033325,
+      "learning_rate": 2e-05,
+      "loss": 0.05558601,
+      "step": 982
+    },
+    {
+      "epoch": 1.966,
+      "grad_norm": 1.5244240760803223,
+      "learning_rate": 2e-05,
+      "loss": 0.06104971,
+      "step": 983
+    },
+    {
+      "epoch": 1.968,
+      "grad_norm": 1.8483918905258179,
+      "learning_rate": 2e-05,
+      "loss": 0.0812193,
+      "step": 984
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.1218955516815186,
+      "learning_rate": 2e-05,
+      "loss": 0.08611986,
+      "step": 985
+    },
+    {
+      "epoch": 1.972,
+      "grad_norm": 1.5318834781646729,
+      "learning_rate": 2e-05,
+      "loss": 0.06183597,
+      "step": 986
+    },
+    {
+      "epoch": 1.974,
+      "grad_norm": 1.713119387626648,
+      "learning_rate": 2e-05,
+      "loss": 0.07985615,
+      "step": 987
+    },
+    {
+      "epoch": 1.976,
+      "grad_norm": 1.4061270952224731,
+      "learning_rate": 2e-05,
+      "loss": 0.04497521,
+      "step": 988
+    },
+    {
+      "epoch": 1.978,
+      "grad_norm": 2.804215431213379,
+      "learning_rate": 2e-05,
+      "loss": 0.06925857,
+      "step": 989
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.997461199760437,
+      "learning_rate": 2e-05,
+      "loss": 0.06266791,
+      "step": 990
+    },
+    {
+      "epoch": 1.982,
+      "grad_norm": 1.849177598953247,
+      "learning_rate": 2e-05,
+      "loss": 0.08790208,
+      "step": 991
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 1.7157963514328003,
+      "learning_rate": 2e-05,
+      "loss": 0.06307608,
+      "step": 992
+    },
+    {
+      "epoch": 1.986,
+      "grad_norm": 1.68929123878479,
+      "learning_rate": 2e-05,
+      "loss": 0.06697696,
+      "step": 993
+    },
+    {
+      "epoch": 1.988,
+      "grad_norm": 1.83262300491333,
+      "learning_rate": 2e-05,
+      "loss": 0.06532292,
+      "step": 994
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.3675966262817383,
+      "learning_rate": 2e-05,
+      "loss": 0.09480947,
+      "step": 995
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 2.172999143600464,
+      "learning_rate": 2e-05,
+      "loss": 0.06149952,
+      "step": 996
+    },
+    {
+      "epoch": 1.994,
+      "grad_norm": 1.6202707290649414,
+      "learning_rate": 2e-05,
+      "loss": 0.05982627,
+      "step": 997
+    },
+    {
+      "epoch": 1.996,
+      "grad_norm": 1.9917645454406738,
+      "learning_rate": 2e-05,
+      "loss": 0.07037735,
+      "step": 998
+    },
+    {
+      "epoch": 1.998,
+      "grad_norm": 2.8166284561157227,
+      "learning_rate": 2e-05,
+      "loss": 0.06254306,
+      "step": 999
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6974472999572754,
+      "learning_rate": 2e-05,
+      "loss": 0.07782187,
+      "step": 1000
+    },
+    {
+      "epoch": 2.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.986,
+        "AngleClassification_2": 0.714,
+        "AngleClassification_3": 0.5269461077844312,
+        "Equal_1": 0.846,
+        "Equal_2": 0.6007984031936128,
+        "Equal_3": 0.6926147704590818,
+        "LineComparison_1": 0.966,
+        "LineComparison_2": 0.9600798403193613,
+        "LineComparison_3": 0.8363273453093812,
+        "Parallel_1": 0.8577154308617234,
+        "Parallel_2": 0.9438877755511023,
+        "Parallel_3": 0.444,
+        "Perpendicular_1": 0.888,
+        "Perpendicular_2": 0.342,
+        "Perpendicular_3": 0.08517034068136273,
+        "PointLiesOnCircle_1": 0.995691382765531,
+        "PointLiesOnCircle_2": 0.9887,
+        "PointLiesOnCircle_3": 0.7656,
+        "PointLiesOnLine_1": 0.811623246492986,
+        "PointLiesOnLine_2": 0.48096192384769537,
+        "PointLiesOnLine_3": 0.24750499001996007
+      },
+      "eval_runtime": 320.1345,
+      "eval_samples_per_second": 32.799,
+      "eval_steps_per_second": 0.656,
+      "step": 1000
+    },
+    {
+      "epoch": 2.002,
+      "grad_norm": 1.6161152124404907,
+      "learning_rate": 2e-05,
+      "loss": 0.0708351,
+      "step": 1001
+    },
+    {
+      "epoch": 2.004,
+      "grad_norm": 1.5361851453781128,
+      "learning_rate": 2e-05,
+      "loss": 0.0666085,
+      "step": 1002
+    },
+    {
+      "epoch": 2.006,
+      "grad_norm": 2.824312925338745,
+      "learning_rate": 2e-05,
+      "loss": 0.09088291,
+      "step": 1003
+    },
+    {
+      "epoch": 2.008,
+      "grad_norm": 1.716564416885376,
+      "learning_rate": 2e-05,
+      "loss": 0.0687227,
+      "step": 1004
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.0519144535064697,
+      "learning_rate": 2e-05,
+      "loss": 0.07319454,
+      "step": 1005
+    },
+    {
+      "epoch": 2.012,
+      "grad_norm": 2.324007511138916,
+      "learning_rate": 2e-05,
+      "loss": 0.09257935,
+      "step": 1006
+    },
+    {
+      "epoch": 2.014,
+      "grad_norm": 4.496967792510986,
+      "learning_rate": 2e-05,
+      "loss": 0.08681659,
+      "step": 1007
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 1.5570460557937622,
+      "learning_rate": 2e-05,
+      "loss": 0.04536861,
+      "step": 1008
+    },
+    {
+      "epoch": 2.018,
+      "grad_norm": 1.8645915985107422,
+      "learning_rate": 2e-05,
+      "loss": 0.05488719,
+      "step": 1009
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.016777515411377,
+      "learning_rate": 2e-05,
+      "loss": 0.08436164,
+      "step": 1010
+    },
+    {
+      "epoch": 2.022,
+      "grad_norm": 1.9473544359207153,
+      "learning_rate": 2e-05,
+      "loss": 0.06556801,
+      "step": 1011
+    },
+    {
+      "epoch": 2.024,
+      "grad_norm": 4.963746547698975,
+      "learning_rate": 2e-05,
+      "loss": 0.08209211,
+      "step": 1012
+    },
+    {
+      "epoch": 2.026,
+      "grad_norm": 2.4789822101593018,
+      "learning_rate": 2e-05,
+      "loss": 0.0703003,
+      "step": 1013
+    },
+    {
+      "epoch": 2.028,
+      "grad_norm": 2.126289129257202,
+      "learning_rate": 2e-05,
+      "loss": 0.03831969,
+      "step": 1014
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.6576788425445557,
+      "learning_rate": 2e-05,
+      "loss": 0.07527678,
+      "step": 1015
+    },
+    {
+      "epoch": 2.032,
+      "grad_norm": 1.8639591932296753,
+      "learning_rate": 2e-05,
+      "loss": 0.05922488,
+      "step": 1016
+    },
+    {
+      "epoch": 2.034,
+      "grad_norm": 1.5314511060714722,
+      "learning_rate": 2e-05,
+      "loss": 0.04786742,
+      "step": 1017
+    },
+    {
+      "epoch": 2.036,
+      "grad_norm": 2.3438453674316406,
+      "learning_rate": 2e-05,
+      "loss": 0.07262143,
+      "step": 1018
+    },
+    {
+      "epoch": 2.038,
+      "grad_norm": 1.8147228956222534,
+      "learning_rate": 2e-05,
+      "loss": 0.04851551,
+      "step": 1019
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.4879653453826904,
+      "learning_rate": 2e-05,
+      "loss": 0.0645643,
+      "step": 1020
+    },
+    {
+      "epoch": 2.042,
+      "grad_norm": 2.240687608718872,
+      "learning_rate": 2e-05,
+      "loss": 0.07334122,
+      "step": 1021
+    },
+    {
+      "epoch": 2.044,
+      "grad_norm": 2.220008611679077,
+      "learning_rate": 2e-05,
+      "loss": 0.06366929,
+      "step": 1022
+    },
+    {
+      "epoch": 2.046,
+      "grad_norm": 2.523574113845825,
+      "learning_rate": 2e-05,
+      "loss": 0.07826661,
+      "step": 1023
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 2.2099592685699463,
+      "learning_rate": 2e-05,
+      "loss": 0.09013459,
+      "step": 1024
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.7173805236816406,
+      "learning_rate": 2e-05,
+      "loss": 0.08330725,
+      "step": 1025
+    },
+    {
+      "epoch": 2.052,
+      "grad_norm": 1.9037878513336182,
+      "learning_rate": 2e-05,
+      "loss": 0.04698469,
+      "step": 1026
+    },
+    {
+      "epoch": 2.054,
+      "grad_norm": 1.8339701890945435,
+      "learning_rate": 2e-05,
+      "loss": 0.07493852,
+      "step": 1027
+    },
+    {
+      "epoch": 2.056,
+      "grad_norm": 1.7632319927215576,
+      "learning_rate": 2e-05,
+      "loss": 0.05750106,
+      "step": 1028
+    },
+    {
+      "epoch": 2.058,
+      "grad_norm": 1.5772873163223267,
+      "learning_rate": 2e-05,
+      "loss": 0.05176328,
+      "step": 1029
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.7405070066452026,
+      "learning_rate": 2e-05,
+      "loss": 0.05858688,
+      "step": 1030
+    },
+    {
+      "epoch": 2.062,
+      "grad_norm": 2.498608112335205,
+      "learning_rate": 2e-05,
+      "loss": 0.06766784,
+      "step": 1031
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 1.690112590789795,
+      "learning_rate": 2e-05,
+      "loss": 0.06187538,
+      "step": 1032
+    },
+    {
+      "epoch": 2.066,
+      "grad_norm": 2.2642834186553955,
+      "learning_rate": 2e-05,
+      "loss": 0.07190007,
+      "step": 1033
+    },
+    {
+      "epoch": 2.068,
+      "grad_norm": 1.882744550704956,
+      "learning_rate": 2e-05,
+      "loss": 0.05131867,
+      "step": 1034
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.5934783220291138,
+      "learning_rate": 2e-05,
+      "loss": 0.03696432,
+      "step": 1035
+    },
+    {
+      "epoch": 2.072,
+      "grad_norm": 2.2791152000427246,
+      "learning_rate": 2e-05,
+      "loss": 0.08653103,
+      "step": 1036
+    },
+    {
+      "epoch": 2.074,
+      "grad_norm": 1.565434217453003,
+      "learning_rate": 2e-05,
+      "loss": 0.05390278,
+      "step": 1037
+    },
+    {
+      "epoch": 2.076,
+      "grad_norm": 2.557328939437866,
+      "learning_rate": 2e-05,
+      "loss": 0.07149288,
+      "step": 1038
+    },
+    {
+      "epoch": 2.078,
+      "grad_norm": 2.0665178298950195,
+      "learning_rate": 2e-05,
+      "loss": 0.04702565,
+      "step": 1039
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.5441120862960815,
+      "learning_rate": 2e-05,
+      "loss": 0.04806751,
+      "step": 1040
+    },
+    {
+      "epoch": 2.082,
+      "grad_norm": 1.4570614099502563,
+      "learning_rate": 2e-05,
+      "loss": 0.05136944,
+      "step": 1041
+    },
+    {
+      "epoch": 2.084,
+      "grad_norm": 5.4056243896484375,
+      "learning_rate": 2e-05,
+      "loss": 0.07949159,
+      "step": 1042
+    },
+    {
+      "epoch": 2.086,
+      "grad_norm": 2.241541624069214,
+      "learning_rate": 2e-05,
+      "loss": 0.08863392,
+      "step": 1043
+    },
+    {
+      "epoch": 2.088,
+      "grad_norm": 1.510636568069458,
+      "learning_rate": 2e-05,
+      "loss": 0.05880911,
+      "step": 1044
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.41506028175354,
+      "learning_rate": 2e-05,
+      "loss": 0.05853722,
+      "step": 1045
+    },
+    {
+      "epoch": 2.092,
+      "grad_norm": 1.5590910911560059,
+      "learning_rate": 2e-05,
+      "loss": 0.04493964,
+      "step": 1046
+    },
+    {
+      "epoch": 2.094,
+      "grad_norm": 2.0853965282440186,
+      "learning_rate": 2e-05,
+      "loss": 0.08284545,
+      "step": 1047
+    },
+    {
+      "epoch": 2.096,
+      "grad_norm": 1.5066843032836914,
+      "learning_rate": 2e-05,
+      "loss": 0.05551049,
+      "step": 1048
+    },
+    {
+      "epoch": 2.098,
+      "grad_norm": 2.1711227893829346,
+      "learning_rate": 2e-05,
+      "loss": 0.07818221,
+      "step": 1049
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.308896541595459,
+      "learning_rate": 2e-05,
+      "loss": 0.04287009,
+      "step": 1050
+    },
+    {
+      "epoch": 2.102,
+      "grad_norm": 2.321077346801758,
+      "learning_rate": 2e-05,
+      "loss": 0.0904954,
+      "step": 1051
+    },
+    {
+      "epoch": 2.104,
+      "grad_norm": 1.5421305894851685,
+      "learning_rate": 2e-05,
+      "loss": 0.06068816,
+      "step": 1052
+    },
+    {
+      "epoch": 2.106,
+      "grad_norm": 1.8072997331619263,
+      "learning_rate": 2e-05,
+      "loss": 0.0773791,
+      "step": 1053
+    },
+    {
+      "epoch": 2.108,
+      "grad_norm": 2.3196661472320557,
+      "learning_rate": 2e-05,
+      "loss": 0.06575853,
+      "step": 1054
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.4077317714691162,
+      "learning_rate": 2e-05,
+      "loss": 0.05838005,
+      "step": 1055
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 1.5628201961517334,
+      "learning_rate": 2e-05,
+      "loss": 0.05414519,
+      "step": 1056
+    },
+    {
+      "epoch": 2.114,
+      "grad_norm": 1.9730654954910278,
+      "learning_rate": 2e-05,
+      "loss": 0.063483,
+      "step": 1057
+    },
+    {
+      "epoch": 2.116,
+      "grad_norm": 1.9468834400177002,
+      "learning_rate": 2e-05,
+      "loss": 0.0622715,
+      "step": 1058
+    },
+    {
+      "epoch": 2.118,
+      "grad_norm": 2.9658288955688477,
+      "learning_rate": 2e-05,
+      "loss": 0.09098145,
+      "step": 1059
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.6334218978881836,
+      "learning_rate": 2e-05,
+      "loss": 0.06693131,
+      "step": 1060
+    },
+    {
+      "epoch": 2.122,
+      "grad_norm": 3.37898850440979,
+      "learning_rate": 2e-05,
+      "loss": 0.09015776,
+      "step": 1061
+    },
+    {
+      "epoch": 2.124,
+      "grad_norm": 2.2235124111175537,
+      "learning_rate": 2e-05,
+      "loss": 0.06049117,
+      "step": 1062
+    },
+    {
+      "epoch": 2.126,
+      "grad_norm": 1.7589530944824219,
+      "learning_rate": 2e-05,
+      "loss": 0.05549654,
+      "step": 1063
+    },
+    {
+      "epoch": 2.128,
+      "grad_norm": 2.0518884658813477,
+      "learning_rate": 2e-05,
+      "loss": 0.08317362,
+      "step": 1064
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.3999139070510864,
+      "learning_rate": 2e-05,
+      "loss": 0.04877549,
+      "step": 1065
+    },
+    {
+      "epoch": 2.132,
+      "grad_norm": 2.0118892192840576,
+      "learning_rate": 2e-05,
+      "loss": 0.07196333,
+      "step": 1066
+    },
+    {
+      "epoch": 2.134,
+      "grad_norm": 1.9917757511138916,
+      "learning_rate": 2e-05,
+      "loss": 0.05575627,
+      "step": 1067
+    },
+    {
+      "epoch": 2.136,
+      "grad_norm": 1.17472505569458,
+      "learning_rate": 2e-05,
+      "loss": 0.04250437,
+      "step": 1068
+    },
+    {
+      "epoch": 2.138,
+      "grad_norm": 1.4256974458694458,
+      "learning_rate": 2e-05,
+      "loss": 0.0531082,
+      "step": 1069
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.143972396850586,
+      "learning_rate": 2e-05,
+      "loss": 0.0886039,
+      "step": 1070
+    },
+    {
+      "epoch": 2.142,
+      "grad_norm": 1.9579774141311646,
+      "learning_rate": 2e-05,
+      "loss": 0.06291603,
+      "step": 1071
+    },
+    {
+      "epoch": 2.144,
+      "grad_norm": 1.78086519241333,
+      "learning_rate": 2e-05,
+      "loss": 0.05680231,
+      "step": 1072
+    },
+    {
+      "epoch": 2.146,
+      "grad_norm": 1.8430737257003784,
+      "learning_rate": 2e-05,
+      "loss": 0.08492243,
+      "step": 1073
+    },
+    {
+      "epoch": 2.148,
+      "grad_norm": 1.4731837511062622,
+      "learning_rate": 2e-05,
+      "loss": 0.05535863,
+      "step": 1074
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.6965607404708862,
+      "learning_rate": 2e-05,
+      "loss": 0.05869256,
+      "step": 1075
+    },
+    {
+      "epoch": 2.152,
+      "grad_norm": 1.960614800453186,
+      "learning_rate": 2e-05,
+      "loss": 0.0604171,
+      "step": 1076
+    },
+    {
+      "epoch": 2.154,
+      "grad_norm": 1.3592660427093506,
+      "learning_rate": 2e-05,
+      "loss": 0.04966343,
+      "step": 1077
+    },
+    {
+      "epoch": 2.156,
+      "grad_norm": 1.4302160739898682,
+      "learning_rate": 2e-05,
+      "loss": 0.05037902,
+      "step": 1078
+    },
+    {
+      "epoch": 2.158,
+      "grad_norm": 1.5716873407363892,
+      "learning_rate": 2e-05,
+      "loss": 0.06660844,
+      "step": 1079
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.2582805156707764,
+      "learning_rate": 2e-05,
+      "loss": 0.04393882,
+      "step": 1080
+    },
+    {
+      "epoch": 2.162,
+      "grad_norm": 2.170337200164795,
+      "learning_rate": 2e-05,
+      "loss": 0.0725324,
+      "step": 1081
+    },
+    {
+      "epoch": 2.164,
+      "grad_norm": 3.0252792835235596,
+      "learning_rate": 2e-05,
+      "loss": 0.08293726,
+      "step": 1082
+    },
+    {
+      "epoch": 2.166,
+      "grad_norm": 2.281851053237915,
+      "learning_rate": 2e-05,
+      "loss": 0.09101221,
+      "step": 1083
+    },
+    {
+      "epoch": 2.168,
+      "grad_norm": 2.3761801719665527,
+      "learning_rate": 2e-05,
+      "loss": 0.06198543,
+      "step": 1084
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.4593560695648193,
+      "learning_rate": 2e-05,
+      "loss": 0.08639136,
+      "step": 1085
+    },
+    {
+      "epoch": 2.172,
+      "grad_norm": 2.1723086833953857,
+      "learning_rate": 2e-05,
+      "loss": 0.06393465,
+      "step": 1086
+    },
+    {
+      "epoch": 2.174,
+      "grad_norm": 2.035813570022583,
+      "learning_rate": 2e-05,
+      "loss": 0.07111633,
+      "step": 1087
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 2.0545802116394043,
+      "learning_rate": 2e-05,
+      "loss": 0.06639965,
+      "step": 1088
+    },
+    {
+      "epoch": 2.178,
+      "grad_norm": 2.862257957458496,
+      "learning_rate": 2e-05,
+      "loss": 0.08525625,
+      "step": 1089
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.8956446647644043,
+      "learning_rate": 2e-05,
+      "loss": 0.06492867,
+      "step": 1090
+    },
+    {
+      "epoch": 2.182,
+      "grad_norm": 1.6777955293655396,
+      "learning_rate": 2e-05,
+      "loss": 0.06760434,
+      "step": 1091
+    },
+    {
+      "epoch": 2.184,
+      "grad_norm": 1.5188251733779907,
+      "learning_rate": 2e-05,
+      "loss": 0.06885065,
+      "step": 1092
+    },
+    {
+      "epoch": 2.186,
+      "grad_norm": 1.5363314151763916,
+      "learning_rate": 2e-05,
+      "loss": 0.06016498,
+      "step": 1093
+    },
+    {
+      "epoch": 2.188,
+      "grad_norm": 1.5654257535934448,
+      "learning_rate": 2e-05,
+      "loss": 0.0583441,
+      "step": 1094
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.322394371032715,
+      "learning_rate": 2e-05,
+      "loss": 0.07740622,
+      "step": 1095
+    },
+    {
+      "epoch": 2.192,
+      "grad_norm": 1.8494588136672974,
+      "learning_rate": 2e-05,
+      "loss": 0.06826154,
+      "step": 1096
+    },
+    {
+      "epoch": 2.194,
+      "grad_norm": 1.8143563270568848,
+      "learning_rate": 2e-05,
+      "loss": 0.07409322,
+      "step": 1097
+    },
+    {
+      "epoch": 2.196,
+      "grad_norm": 2.239018440246582,
+      "learning_rate": 2e-05,
+      "loss": 0.06441543,
+      "step": 1098
+    },
+    {
+      "epoch": 2.198,
+      "grad_norm": 1.3814536333084106,
+      "learning_rate": 2e-05,
+      "loss": 0.06695756,
+      "step": 1099
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.9067416191101074,
+      "learning_rate": 2e-05,
+      "loss": 0.0664328,
+      "step": 1100
+    },
+    {
+      "epoch": 2.202,
+      "grad_norm": 2.319135904312134,
+      "learning_rate": 2e-05,
+      "loss": 0.07544907,
+      "step": 1101
+    },
+    {
+      "epoch": 2.204,
+      "grad_norm": 1.6520764827728271,
+      "learning_rate": 2e-05,
+      "loss": 0.04925615,
+      "step": 1102
+    },
+    {
+      "epoch": 2.206,
+      "grad_norm": 2.0650670528411865,
+      "learning_rate": 2e-05,
+      "loss": 0.08087505,
+      "step": 1103
+    },
+    {
+      "epoch": 2.208,
+      "grad_norm": 2.419358730316162,
+      "learning_rate": 2e-05,
+      "loss": 0.08194488,
+      "step": 1104
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.1177220344543457,
+      "learning_rate": 2e-05,
+      "loss": 0.06622218,
+      "step": 1105
+    },
+    {
+      "epoch": 2.212,
+      "grad_norm": 1.8905670642852783,
+      "learning_rate": 2e-05,
+      "loss": 0.05285282,
+      "step": 1106
+    },
+    {
+      "epoch": 2.214,
+      "grad_norm": 1.7442760467529297,
+      "learning_rate": 2e-05,
+      "loss": 0.07012412,
+      "step": 1107
+    },
+    {
+      "epoch": 2.216,
+      "grad_norm": 2.560086965560913,
+      "learning_rate": 2e-05,
+      "loss": 0.07056236,
+      "step": 1108
+    },
+    {
+      "epoch": 2.218,
+      "grad_norm": 1.3487662076950073,
+      "learning_rate": 2e-05,
+      "loss": 0.04359243,
+      "step": 1109
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.0357956886291504,
+      "learning_rate": 2e-05,
+      "loss": 0.06766598,
+      "step": 1110
+    },
+    {
+      "epoch": 2.222,
+      "grad_norm": 1.6275434494018555,
+      "learning_rate": 2e-05,
+      "loss": 0.06916457,
+      "step": 1111
+    },
+    {
+      "epoch": 2.224,
+      "grad_norm": 1.9597185850143433,
+      "learning_rate": 2e-05,
+      "loss": 0.07099155,
+      "step": 1112
+    },
+    {
+      "epoch": 2.226,
+      "grad_norm": 1.4252787828445435,
+      "learning_rate": 2e-05,
+      "loss": 0.04932921,
+      "step": 1113
+    },
+    {
+      "epoch": 2.228,
+      "grad_norm": 1.5756709575653076,
+      "learning_rate": 2e-05,
+      "loss": 0.05278656,
+      "step": 1114
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.331275701522827,
+      "learning_rate": 2e-05,
+      "loss": 0.07562015,
+      "step": 1115
+    },
+    {
+      "epoch": 2.232,
+      "grad_norm": 1.697662115097046,
+      "learning_rate": 2e-05,
+      "loss": 0.06566669,
+      "step": 1116
+    },
+    {
+      "epoch": 2.234,
+      "grad_norm": 1.760780692100525,
+      "learning_rate": 2e-05,
+      "loss": 0.04810528,
+      "step": 1117
+    },
+    {
+      "epoch": 2.2359999999999998,
+      "grad_norm": 2.7786638736724854,
+      "learning_rate": 2e-05,
+      "loss": 0.05388356,
+      "step": 1118
+    },
+    {
+      "epoch": 2.238,
+      "grad_norm": 2.2772700786590576,
+      "learning_rate": 2e-05,
+      "loss": 0.07156706,
+      "step": 1119
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.3025233745574951,
+      "learning_rate": 2e-05,
+      "loss": 0.04904636,
+      "step": 1120
+    },
+    {
+      "epoch": 2.242,
+      "grad_norm": 2.5209176540374756,
+      "learning_rate": 2e-05,
+      "loss": 0.0652664,
+      "step": 1121
+    },
+    {
+      "epoch": 2.2439999999999998,
+      "grad_norm": 1.7902709245681763,
+      "learning_rate": 2e-05,
+      "loss": 0.061213,
+      "step": 1122
+    },
+    {
+      "epoch": 2.246,
+      "grad_norm": 1.2348299026489258,
+      "learning_rate": 2e-05,
+      "loss": 0.03768958,
+      "step": 1123
+    },
+    {
+      "epoch": 2.248,
+      "grad_norm": 1.4644815921783447,
+      "learning_rate": 2e-05,
+      "loss": 0.05811838,
+      "step": 1124
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.8537580966949463,
+      "learning_rate": 2e-05,
+      "loss": 0.06129285,
+      "step": 1125
+    },
+    {
+      "epoch": 2.252,
+      "grad_norm": 1.9889532327651978,
+      "learning_rate": 2e-05,
+      "loss": 0.05362317,
+      "step": 1126
+    },
+    {
+      "epoch": 2.254,
+      "grad_norm": 1.2981350421905518,
+      "learning_rate": 2e-05,
+      "loss": 0.03556494,
+      "step": 1127
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 2.6438069343566895,
+      "learning_rate": 2e-05,
+      "loss": 0.07400218,
+      "step": 1128
+    },
+    {
+      "epoch": 2.258,
+      "grad_norm": 2.0551059246063232,
+      "learning_rate": 2e-05,
+      "loss": 0.06481735,
+      "step": 1129
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.9357151985168457,
+      "learning_rate": 2e-05,
+      "loss": 0.07889327,
+      "step": 1130
+    },
+    {
+      "epoch": 2.262,
+      "grad_norm": 1.4854387044906616,
+      "learning_rate": 2e-05,
+      "loss": 0.04724164,
+      "step": 1131
+    },
+    {
+      "epoch": 2.2640000000000002,
+      "grad_norm": 3.739743947982788,
+      "learning_rate": 2e-05,
+      "loss": 0.07142067,
+      "step": 1132
+    },
+    {
+      "epoch": 2.266,
+      "grad_norm": 2.309699773788452,
+      "learning_rate": 2e-05,
+      "loss": 0.05937993,
+      "step": 1133
+    },
+    {
+      "epoch": 2.268,
+      "grad_norm": 3.0048446655273438,
+      "learning_rate": 2e-05,
+      "loss": 0.08490255,
+      "step": 1134
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.9927393198013306,
+      "learning_rate": 2e-05,
+      "loss": 0.05297955,
+      "step": 1135
+    },
+    {
+      "epoch": 2.2720000000000002,
+      "grad_norm": 2.1314926147460938,
+      "learning_rate": 2e-05,
+      "loss": 0.05550206,
+      "step": 1136
+    },
+    {
+      "epoch": 2.274,
+      "grad_norm": 1.467367172241211,
+      "learning_rate": 2e-05,
+      "loss": 0.06153769,
+      "step": 1137
+    },
+    {
+      "epoch": 2.276,
+      "grad_norm": 2.6547863483428955,
+      "learning_rate": 2e-05,
+      "loss": 0.06249413,
+      "step": 1138
+    },
+    {
+      "epoch": 2.278,
+      "grad_norm": 2.3911125659942627,
+      "learning_rate": 2e-05,
+      "loss": 0.04478404,
+      "step": 1139
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 2.295572280883789,
+      "learning_rate": 2e-05,
+      "loss": 0.05000008,
+      "step": 1140
+    },
+    {
+      "epoch": 2.282,
+      "grad_norm": 1.823384165763855,
+      "learning_rate": 2e-05,
+      "loss": 0.05099558,
+      "step": 1141
+    },
+    {
+      "epoch": 2.284,
+      "grad_norm": 2.4514079093933105,
+      "learning_rate": 2e-05,
+      "loss": 0.07286972,
+      "step": 1142
+    },
+    {
+      "epoch": 2.286,
+      "grad_norm": 1.5815463066101074,
+      "learning_rate": 2e-05,
+      "loss": 0.06633377,
+      "step": 1143
+    },
+    {
+      "epoch": 2.288,
+      "grad_norm": 1.3394488096237183,
+      "learning_rate": 2e-05,
+      "loss": 0.05137707,
+      "step": 1144
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.6430034637451172,
+      "learning_rate": 2e-05,
+      "loss": 0.03971995,
+      "step": 1145
+    },
+    {
+      "epoch": 2.292,
+      "grad_norm": 2.691917657852173,
+      "learning_rate": 2e-05,
+      "loss": 0.06252594,
+      "step": 1146
+    },
+    {
+      "epoch": 2.294,
+      "grad_norm": 1.7687281370162964,
+      "learning_rate": 2e-05,
+      "loss": 0.05269225,
+      "step": 1147
+    },
+    {
+      "epoch": 2.296,
+      "grad_norm": 1.8105900287628174,
+      "learning_rate": 2e-05,
+      "loss": 0.07036023,
+      "step": 1148
+    },
+    {
+      "epoch": 2.298,
+      "grad_norm": 1.4458359479904175,
+      "learning_rate": 2e-05,
+      "loss": 0.05376993,
+      "step": 1149
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.6199760437011719,
+      "learning_rate": 2e-05,
+      "loss": 0.06054214,
+      "step": 1150
+    },
+    {
+      "epoch": 2.302,
+      "grad_norm": 1.98055100440979,
+      "learning_rate": 2e-05,
+      "loss": 0.07045031,
+      "step": 1151
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 1.7091319561004639,
+      "learning_rate": 2e-05,
+      "loss": 0.04949348,
+      "step": 1152
+    },
+    {
+      "epoch": 2.306,
+      "grad_norm": 2.5169119834899902,
+      "learning_rate": 2e-05,
+      "loss": 0.06985944,
+      "step": 1153
+    },
+    {
+      "epoch": 2.308,
+      "grad_norm": 1.5572229623794556,
+      "learning_rate": 2e-05,
+      "loss": 0.0572869,
+      "step": 1154
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.3380517959594727,
+      "learning_rate": 2e-05,
+      "loss": 0.04210208,
+      "step": 1155
+    },
+    {
+      "epoch": 2.312,
+      "grad_norm": 2.2991576194763184,
+      "learning_rate": 2e-05,
+      "loss": 0.06161058,
+      "step": 1156
+    },
+    {
+      "epoch": 2.314,
+      "grad_norm": 1.362855076789856,
+      "learning_rate": 2e-05,
+      "loss": 0.05341607,
+      "step": 1157
+    },
+    {
+      "epoch": 2.316,
+      "grad_norm": 1.6171643733978271,
+      "learning_rate": 2e-05,
+      "loss": 0.05781926,
+      "step": 1158
+    },
+    {
+      "epoch": 2.318,
+      "grad_norm": 2.0822646617889404,
+      "learning_rate": 2e-05,
+      "loss": 0.05547912,
+      "step": 1159
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.1096858978271484,
+      "learning_rate": 2e-05,
+      "loss": 0.06335936,
+      "step": 1160
+    },
+    {
+      "epoch": 2.322,
+      "grad_norm": 1.6990431547164917,
+      "learning_rate": 2e-05,
+      "loss": 0.06178097,
+      "step": 1161
+    },
+    {
+      "epoch": 2.324,
+      "grad_norm": 1.849997878074646,
+      "learning_rate": 2e-05,
+      "loss": 0.06346186,
+      "step": 1162
+    },
+    {
+      "epoch": 2.326,
+      "grad_norm": 1.6597445011138916,
+      "learning_rate": 2e-05,
+      "loss": 0.05579802,
+      "step": 1163
+    },
+    {
+      "epoch": 2.328,
+      "grad_norm": 1.805808663368225,
+      "learning_rate": 2e-05,
+      "loss": 0.08137181,
+      "step": 1164
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.5715245008468628,
+      "learning_rate": 2e-05,
+      "loss": 0.07559416,
+      "step": 1165
+    },
+    {
+      "epoch": 2.332,
+      "grad_norm": 1.6368138790130615,
+      "learning_rate": 2e-05,
+      "loss": 0.06028969,
+      "step": 1166
+    },
+    {
+      "epoch": 2.334,
+      "grad_norm": 2.369986057281494,
+      "learning_rate": 2e-05,
+      "loss": 0.05975366,
+      "step": 1167
+    },
+    {
+      "epoch": 2.336,
+      "grad_norm": 1.567192792892456,
+      "learning_rate": 2e-05,
+      "loss": 0.05366561,
+      "step": 1168
+    },
+    {
+      "epoch": 2.338,
+      "grad_norm": 1.529199242591858,
+      "learning_rate": 2e-05,
+      "loss": 0.05094039,
+      "step": 1169
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.299103260040283,
+      "learning_rate": 2e-05,
+      "loss": 0.05972379,
+      "step": 1170
+    },
+    {
+      "epoch": 2.342,
+      "grad_norm": 2.0813450813293457,
+      "learning_rate": 2e-05,
+      "loss": 0.06179533,
+      "step": 1171
+    },
+    {
+      "epoch": 2.344,
+      "grad_norm": 1.8993735313415527,
+      "learning_rate": 2e-05,
+      "loss": 0.05901102,
+      "step": 1172
+    },
+    {
+      "epoch": 2.346,
+      "grad_norm": 1.505812644958496,
+      "learning_rate": 2e-05,
+      "loss": 0.05369399,
+      "step": 1173
+    },
+    {
+      "epoch": 2.348,
+      "grad_norm": 42.221553802490234,
+      "learning_rate": 2e-05,
+      "loss": 0.1112899,
+      "step": 1174
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.6444224119186401,
+      "learning_rate": 2e-05,
+      "loss": 0.05893474,
+      "step": 1175
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 1.4802619218826294,
+      "learning_rate": 2e-05,
+      "loss": 0.05816791,
+      "step": 1176
+    },
+    {
+      "epoch": 2.354,
+      "grad_norm": 1.5738928318023682,
+      "learning_rate": 2e-05,
+      "loss": 0.07226036,
+      "step": 1177
+    },
+    {
+      "epoch": 2.356,
+      "grad_norm": 1.077085018157959,
+      "learning_rate": 2e-05,
+      "loss": 0.03205248,
+      "step": 1178
+    },
+    {
+      "epoch": 2.358,
+      "grad_norm": 48.9713134765625,
+      "learning_rate": 2e-05,
+      "loss": 0.17173615,
+      "step": 1179
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.1683356761932373,
+      "learning_rate": 2e-05,
+      "loss": 0.04456889,
+      "step": 1180
+    },
+    {
+      "epoch": 2.362,
+      "grad_norm": 8.09849739074707,
+      "learning_rate": 2e-05,
+      "loss": 0.29480407,
+      "step": 1181
+    },
+    {
+      "epoch": 2.364,
+      "grad_norm": 1.4625880718231201,
+      "learning_rate": 2e-05,
+      "loss": 0.04769106,
+      "step": 1182
+    },
+    {
+      "epoch": 2.366,
+      "grad_norm": 1.876531720161438,
+      "learning_rate": 2e-05,
+      "loss": 0.0816292,
+      "step": 1183
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 1.9798403978347778,
+      "learning_rate": 2e-05,
+      "loss": 0.06011814,
+      "step": 1184
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.5905330181121826,
+      "learning_rate": 2e-05,
+      "loss": 0.05314508,
+      "step": 1185
+    },
+    {
+      "epoch": 2.372,
+      "grad_norm": 1.7121469974517822,
+      "learning_rate": 2e-05,
+      "loss": 0.04185162,
+      "step": 1186
+    },
+    {
+      "epoch": 2.374,
+      "grad_norm": 1.7982616424560547,
+      "learning_rate": 2e-05,
+      "loss": 0.04177887,
+      "step": 1187
+    },
+    {
+      "epoch": 2.376,
+      "grad_norm": 2.124553680419922,
+      "learning_rate": 2e-05,
+      "loss": 0.06069058,
+      "step": 1188
+    },
+    {
+      "epoch": 2.378,
+      "grad_norm": 1.8068010807037354,
+      "learning_rate": 2e-05,
+      "loss": 0.05539012,
+      "step": 1189
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.0612502098083496,
+      "learning_rate": 2e-05,
+      "loss": 0.07466509,
+      "step": 1190
+    },
+    {
+      "epoch": 2.382,
+      "grad_norm": 1.1964912414550781,
+      "learning_rate": 2e-05,
+      "loss": 0.04185656,
+      "step": 1191
+    },
+    {
+      "epoch": 2.384,
+      "grad_norm": 2.028778314590454,
+      "learning_rate": 2e-05,
+      "loss": 0.05173714,
+      "step": 1192
+    },
+    {
+      "epoch": 2.386,
+      "grad_norm": 1.7989981174468994,
+      "learning_rate": 2e-05,
+      "loss": 0.05134147,
+      "step": 1193
+    },
+    {
+      "epoch": 2.388,
+      "grad_norm": 1.6559616327285767,
+      "learning_rate": 2e-05,
+      "loss": 0.06251637,
+      "step": 1194
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.4277832508087158,
+      "learning_rate": 2e-05,
+      "loss": 0.05279864,
+      "step": 1195
+    },
+    {
+      "epoch": 2.392,
+      "grad_norm": 1.6468921899795532,
+      "learning_rate": 2e-05,
+      "loss": 0.06037339,
+      "step": 1196
+    },
+    {
+      "epoch": 2.394,
+      "grad_norm": 3.1690855026245117,
+      "learning_rate": 2e-05,
+      "loss": 0.08198987,
+      "step": 1197
+    },
+    {
+      "epoch": 2.396,
+      "grad_norm": 1.2297303676605225,
+      "learning_rate": 2e-05,
+      "loss": 0.04745869,
+      "step": 1198
+    },
+    {
+      "epoch": 2.398,
+      "grad_norm": 2.2883074283599854,
+      "learning_rate": 2e-05,
+      "loss": 0.04119899,
+      "step": 1199
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.0673367977142334,
+      "learning_rate": 2e-05,
+      "loss": 0.05730217,
+      "step": 1200
+    },
+    {
+      "epoch": 2.402,
+      "grad_norm": 1.2708475589752197,
+      "learning_rate": 2e-05,
+      "loss": 0.04456803,
+      "step": 1201
+    },
+    {
+      "epoch": 2.404,
+      "grad_norm": 1.3558706045150757,
+      "learning_rate": 2e-05,
+      "loss": 0.05208869,
+      "step": 1202
+    },
+    {
+      "epoch": 2.406,
+      "grad_norm": 1.2289046049118042,
+      "learning_rate": 2e-05,
+      "loss": 0.04607398,
+      "step": 1203
+    },
+    {
+      "epoch": 2.408,
+      "grad_norm": 3.603008985519409,
+      "learning_rate": 2e-05,
+      "loss": 0.08612712,
+      "step": 1204
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.123850345611572,
+      "learning_rate": 2e-05,
+      "loss": 0.06038028,
+      "step": 1205
+    },
+    {
+      "epoch": 2.412,
+      "grad_norm": 2.2089900970458984,
+      "learning_rate": 2e-05,
+      "loss": 0.06190875,
+      "step": 1206
+    },
+    {
+      "epoch": 2.414,
+      "grad_norm": 1.7213412523269653,
+      "learning_rate": 2e-05,
+      "loss": 0.05619193,
+      "step": 1207
+    },
+    {
+      "epoch": 2.416,
+      "grad_norm": 2.1176223754882812,
+      "learning_rate": 2e-05,
+      "loss": 0.06305473,
+      "step": 1208
+    },
+    {
+      "epoch": 2.418,
+      "grad_norm": 2.206848621368408,
+      "learning_rate": 2e-05,
+      "loss": 0.06026538,
+      "step": 1209
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.4014763832092285,
+      "learning_rate": 2e-05,
+      "loss": 0.04447113,
+      "step": 1210
+    },
+    {
+      "epoch": 2.422,
+      "grad_norm": 1.3986562490463257,
+      "learning_rate": 2e-05,
+      "loss": 0.04321969,
+      "step": 1211
+    },
+    {
+      "epoch": 2.424,
+      "grad_norm": 1.840933084487915,
+      "learning_rate": 2e-05,
+      "loss": 0.05479154,
+      "step": 1212
+    },
+    {
+      "epoch": 2.426,
+      "grad_norm": 2.6934752464294434,
+      "learning_rate": 2e-05,
+      "loss": 0.05985066,
+      "step": 1213
+    },
+    {
+      "epoch": 2.428,
+      "grad_norm": 1.4198068380355835,
+      "learning_rate": 2e-05,
+      "loss": 0.04191133,
+      "step": 1214
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.204730272293091,
+      "learning_rate": 2e-05,
+      "loss": 0.04986127,
+      "step": 1215
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 1.7930086851119995,
+      "learning_rate": 2e-05,
+      "loss": 0.03822114,
+      "step": 1216
+    },
+    {
+      "epoch": 2.434,
+      "grad_norm": 1.9965940713882446,
+      "learning_rate": 2e-05,
+      "loss": 0.07922506,
+      "step": 1217
+    },
+    {
+      "epoch": 2.436,
+      "grad_norm": 1.8581851720809937,
+      "learning_rate": 2e-05,
+      "loss": 0.06270727,
+      "step": 1218
+    },
+    {
+      "epoch": 2.438,
+      "grad_norm": 2.940757989883423,
+      "learning_rate": 2e-05,
+      "loss": 0.05123382,
+      "step": 1219
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.8128314018249512,
+      "learning_rate": 2e-05,
+      "loss": 0.05950858,
+      "step": 1220
+    },
+    {
+      "epoch": 2.442,
+      "grad_norm": 2.7984111309051514,
+      "learning_rate": 2e-05,
+      "loss": 0.04888112,
+      "step": 1221
+    },
+    {
+      "epoch": 2.444,
+      "grad_norm": 1.7929303646087646,
+      "learning_rate": 2e-05,
+      "loss": 0.05642758,
+      "step": 1222
+    },
+    {
+      "epoch": 2.446,
+      "grad_norm": 3.5627307891845703,
+      "learning_rate": 2e-05,
+      "loss": 0.06712979,
+      "step": 1223
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 2.1680853366851807,
+      "learning_rate": 2e-05,
+      "loss": 0.04695696,
+      "step": 1224
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.543848991394043,
+      "learning_rate": 2e-05,
+      "loss": 0.0471656,
+      "step": 1225
+    },
+    {
+      "epoch": 2.452,
+      "grad_norm": 1.377380609512329,
+      "learning_rate": 2e-05,
+      "loss": 0.04254293,
+      "step": 1226
+    },
+    {
+      "epoch": 2.454,
+      "grad_norm": 2.116429090499878,
+      "learning_rate": 2e-05,
+      "loss": 0.0888541,
+      "step": 1227
+    },
+    {
+      "epoch": 2.456,
+      "grad_norm": 1.714355707168579,
+      "learning_rate": 2e-05,
+      "loss": 0.04009545,
+      "step": 1228
+    },
+    {
+      "epoch": 2.458,
+      "grad_norm": 2.4635934829711914,
+      "learning_rate": 2e-05,
+      "loss": 0.06196552,
+      "step": 1229
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.6283334493637085,
+      "learning_rate": 2e-05,
+      "loss": 0.05298021,
+      "step": 1230
+    },
+    {
+      "epoch": 2.462,
+      "grad_norm": 2.1310596466064453,
+      "learning_rate": 2e-05,
+      "loss": 0.0582485,
+      "step": 1231
+    },
+    {
+      "epoch": 2.464,
+      "grad_norm": 2.8013837337493896,
+      "learning_rate": 2e-05,
+      "loss": 0.07070208,
+      "step": 1232
+    },
+    {
+      "epoch": 2.466,
+      "grad_norm": 2.706620693206787,
+      "learning_rate": 2e-05,
+      "loss": 0.05937808,
+      "step": 1233
+    },
+    {
+      "epoch": 2.468,
+      "grad_norm": 2.1183664798736572,
+      "learning_rate": 2e-05,
+      "loss": 0.07405435,
+      "step": 1234
+    },
+    {
+      "epoch": 2.4699999999999998,
+      "grad_norm": 1.8695333003997803,
+      "learning_rate": 2e-05,
+      "loss": 0.04990563,
+      "step": 1235
+    },
+    {
+      "epoch": 2.472,
+      "grad_norm": 1.3877243995666504,
+      "learning_rate": 2e-05,
+      "loss": 0.05714301,
+      "step": 1236
+    },
+    {
+      "epoch": 2.474,
+      "grad_norm": 2.160632371902466,
+      "learning_rate": 2e-05,
+      "loss": 0.05262284,
+      "step": 1237
+    },
+    {
+      "epoch": 2.476,
+      "grad_norm": 3.242225408554077,
+      "learning_rate": 2e-05,
+      "loss": 0.07168475,
+      "step": 1238
+    },
+    {
+      "epoch": 2.4779999999999998,
+      "grad_norm": 1.5651336908340454,
+      "learning_rate": 2e-05,
+      "loss": 0.03872868,
+      "step": 1239
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.6729652881622314,
+      "learning_rate": 2e-05,
+      "loss": 0.06602624,
+      "step": 1240
+    },
+    {
+      "epoch": 2.482,
+      "grad_norm": 1.9314351081848145,
+      "learning_rate": 2e-05,
+      "loss": 0.06864213,
+      "step": 1241
+    },
+    {
+      "epoch": 2.484,
+      "grad_norm": 2.9383625984191895,
+      "learning_rate": 2e-05,
+      "loss": 0.07590427,
+      "step": 1242
+    },
+    {
+      "epoch": 2.4859999999999998,
+      "grad_norm": 1.5178899765014648,
+      "learning_rate": 2e-05,
+      "loss": 0.04809076,
+      "step": 1243
+    },
+    {
+      "epoch": 2.488,
+      "grad_norm": 1.3784013986587524,
+      "learning_rate": 2e-05,
+      "loss": 0.04864044,
+      "step": 1244
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.29901123046875,
+      "learning_rate": 2e-05,
+      "loss": 0.03983694,
+      "step": 1245
+    },
+    {
+      "epoch": 2.492,
+      "grad_norm": 2.0041773319244385,
+      "learning_rate": 2e-05,
+      "loss": 0.06119365,
+      "step": 1246
+    },
+    {
+      "epoch": 2.4939999999999998,
+      "grad_norm": 1.760473608970642,
+      "learning_rate": 2e-05,
+      "loss": 0.05547423,
+      "step": 1247
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 1.782887578010559,
+      "learning_rate": 2e-05,
+      "loss": 0.05595577,
+      "step": 1248
+    },
+    {
+      "epoch": 2.498,
+      "grad_norm": 1.431018590927124,
+      "learning_rate": 2e-05,
+      "loss": 0.05510469,
+      "step": 1249
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.7681326866149902,
+      "learning_rate": 2e-05,
+      "loss": 0.07547972,
+      "step": 1250
+    },
+    {
+      "epoch": 2.502,
+      "grad_norm": 3.4799344539642334,
+      "learning_rate": 2e-05,
+      "loss": 0.08792815,
+      "step": 1251
+    },
+    {
+      "epoch": 2.504,
+      "grad_norm": 1.4463987350463867,
+      "learning_rate": 2e-05,
+      "loss": 0.04625604,
+      "step": 1252
+    },
+    {
+      "epoch": 2.5060000000000002,
+      "grad_norm": 7.698814392089844,
+      "learning_rate": 2e-05,
+      "loss": 0.07450949,
+      "step": 1253
+    },
+    {
+      "epoch": 2.508,
+      "grad_norm": 2.0825157165527344,
+      "learning_rate": 2e-05,
+      "loss": 0.06824566,
+      "step": 1254
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.138805389404297,
+      "learning_rate": 2e-05,
+      "loss": 0.06627047,
+      "step": 1255
+    },
+    {
+      "epoch": 2.512,
+      "grad_norm": 2.9669065475463867,
+      "learning_rate": 2e-05,
+      "loss": 0.07500726,
+      "step": 1256
+    },
+    {
+      "epoch": 2.5140000000000002,
+      "grad_norm": 1.8249704837799072,
+      "learning_rate": 2e-05,
+      "loss": 0.05518293,
+      "step": 1257
+    },
+    {
+      "epoch": 2.516,
+      "grad_norm": 1.9127455949783325,
+      "learning_rate": 2e-05,
+      "loss": 0.05548128,
+      "step": 1258
+    },
+    {
+      "epoch": 2.518,
+      "grad_norm": 2.1432528495788574,
+      "learning_rate": 2e-05,
+      "loss": 0.0882006,
+      "step": 1259
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.9188241958618164,
+      "learning_rate": 2e-05,
+      "loss": 0.05358675,
+      "step": 1260
+    },
+    {
+      "epoch": 2.5220000000000002,
+      "grad_norm": 1.2570300102233887,
+      "learning_rate": 2e-05,
+      "loss": 0.04172009,
+      "step": 1261
+    },
+    {
+      "epoch": 2.524,
+      "grad_norm": 1.4593505859375,
+      "learning_rate": 2e-05,
+      "loss": 0.05119102,
+      "step": 1262
+    },
+    {
+      "epoch": 2.526,
+      "grad_norm": 1.6832845211029053,
+      "learning_rate": 2e-05,
+      "loss": 0.04923711,
+      "step": 1263
+    },
+    {
+      "epoch": 2.528,
+      "grad_norm": 2.0682425498962402,
+      "learning_rate": 2e-05,
+      "loss": 0.05748929,
+      "step": 1264
+    },
+    {
+      "epoch": 2.5300000000000002,
+      "grad_norm": 2.118595838546753,
+      "learning_rate": 2e-05,
+      "loss": 0.07105608,
+      "step": 1265
+    },
+    {
+      "epoch": 2.532,
+      "grad_norm": 1.294793725013733,
+      "learning_rate": 2e-05,
+      "loss": 0.05542181,
+      "step": 1266
+    },
+    {
+      "epoch": 2.534,
+      "grad_norm": 2.6284120082855225,
+      "learning_rate": 2e-05,
+      "loss": 0.04738271,
+      "step": 1267
+    },
+    {
+      "epoch": 2.536,
+      "grad_norm": 1.572341799736023,
+      "learning_rate": 2e-05,
+      "loss": 0.04801615,
+      "step": 1268
+    },
+    {
+      "epoch": 2.5380000000000003,
+      "grad_norm": 1.7317556142807007,
+      "learning_rate": 2e-05,
+      "loss": 0.04552308,
+      "step": 1269
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.7648011445999146,
+      "learning_rate": 2e-05,
+      "loss": 0.05321316,
+      "step": 1270
+    },
+    {
+      "epoch": 2.542,
+      "grad_norm": 3.3624472618103027,
+      "learning_rate": 2e-05,
+      "loss": 0.04517394,
+      "step": 1271
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 1.198052167892456,
+      "learning_rate": 2e-05,
+      "loss": 0.03292371,
+      "step": 1272
+    },
+    {
+      "epoch": 2.5460000000000003,
+      "grad_norm": 1.9982932806015015,
+      "learning_rate": 2e-05,
+      "loss": 0.04539865,
+      "step": 1273
+    },
+    {
+      "epoch": 2.548,
+      "grad_norm": 2.1996402740478516,
+      "learning_rate": 2e-05,
+      "loss": 0.06276131,
+      "step": 1274
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.3174142837524414,
+      "learning_rate": 2e-05,
+      "loss": 0.07670379,
+      "step": 1275
+    },
+    {
+      "epoch": 2.552,
+      "grad_norm": 2.5731277465820312,
+      "learning_rate": 2e-05,
+      "loss": 0.05952568,
+      "step": 1276
+    },
+    {
+      "epoch": 2.5540000000000003,
+      "grad_norm": 2.444291591644287,
+      "learning_rate": 2e-05,
+      "loss": 0.06525612,
+      "step": 1277
+    },
+    {
+      "epoch": 2.556,
+      "grad_norm": 2.2760965824127197,
+      "learning_rate": 2e-05,
+      "loss": 0.06015107,
+      "step": 1278
+    },
+    {
+      "epoch": 2.558,
+      "grad_norm": 1.7860162258148193,
+      "learning_rate": 2e-05,
+      "loss": 0.06085454,
+      "step": 1279
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.5104821920394897,
+      "learning_rate": 2e-05,
+      "loss": 0.04777806,
+      "step": 1280
+    },
+    {
+      "epoch": 2.5620000000000003,
+      "grad_norm": 2.0118963718414307,
+      "learning_rate": 2e-05,
+      "loss": 0.06037318,
+      "step": 1281
+    },
+    {
+      "epoch": 2.564,
+      "grad_norm": 1.3859965801239014,
+      "learning_rate": 2e-05,
+      "loss": 0.05690242,
+      "step": 1282
+    },
+    {
+      "epoch": 2.566,
+      "grad_norm": 1.6642067432403564,
+      "learning_rate": 2e-05,
+      "loss": 0.06878574,
+      "step": 1283
+    },
+    {
+      "epoch": 2.568,
+      "grad_norm": 1.7605303525924683,
+      "learning_rate": 2e-05,
+      "loss": 0.04636247,
+      "step": 1284
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.2372325658798218,
+      "learning_rate": 2e-05,
+      "loss": 0.0318442,
+      "step": 1285
+    },
+    {
+      "epoch": 2.572,
+      "grad_norm": 1.015973687171936,
+      "learning_rate": 2e-05,
+      "loss": 0.02155654,
+      "step": 1286
+    },
+    {
+      "epoch": 2.574,
+      "grad_norm": 1.6201320886611938,
+      "learning_rate": 2e-05,
+      "loss": 0.03975923,
+      "step": 1287
+    },
+    {
+      "epoch": 2.576,
+      "grad_norm": 3.8816092014312744,
+      "learning_rate": 2e-05,
+      "loss": 0.08194003,
+      "step": 1288
+    },
+    {
+      "epoch": 2.578,
+      "grad_norm": 1.7358951568603516,
+      "learning_rate": 2e-05,
+      "loss": 0.04974889,
+      "step": 1289
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.392040252685547,
+      "learning_rate": 2e-05,
+      "loss": 0.06020691,
+      "step": 1290
+    },
+    {
+      "epoch": 2.582,
+      "grad_norm": 1.1086469888687134,
+      "learning_rate": 2e-05,
+      "loss": 0.04583845,
+      "step": 1291
+    },
+    {
+      "epoch": 2.584,
+      "grad_norm": 1.5086019039154053,
+      "learning_rate": 2e-05,
+      "loss": 0.04313815,
+      "step": 1292
+    },
+    {
+      "epoch": 2.586,
+      "grad_norm": 1.7809884548187256,
+      "learning_rate": 2e-05,
+      "loss": 0.05329616,
+      "step": 1293
+    },
+    {
+      "epoch": 2.588,
+      "grad_norm": 1.52059006690979,
+      "learning_rate": 2e-05,
+      "loss": 0.04525972,
+      "step": 1294
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.2042572498321533,
+      "learning_rate": 2e-05,
+      "loss": 0.0593795,
+      "step": 1295
+    },
+    {
+      "epoch": 2.592,
+      "grad_norm": 1.675024151802063,
+      "learning_rate": 2e-05,
+      "loss": 0.06052756,
+      "step": 1296
+    },
+    {
+      "epoch": 2.594,
+      "grad_norm": 2.8984549045562744,
+      "learning_rate": 2e-05,
+      "loss": 0.0687076,
+      "step": 1297
+    },
+    {
+      "epoch": 2.596,
+      "grad_norm": 2.2647969722747803,
+      "learning_rate": 2e-05,
+      "loss": 0.05009675,
+      "step": 1298
+    },
+    {
+      "epoch": 2.598,
+      "grad_norm": 1.8025486469268799,
+      "learning_rate": 2e-05,
+      "loss": 0.05575639,
+      "step": 1299
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.8778318166732788,
+      "learning_rate": 2e-05,
+      "loss": 0.04730181,
+      "step": 1300
+    },
+    {
+      "epoch": 2.602,
+      "grad_norm": 1.6435505151748657,
+      "learning_rate": 2e-05,
+      "loss": 0.04583211,
+      "step": 1301
+    },
+    {
+      "epoch": 2.604,
+      "grad_norm": 1.7641148567199707,
+      "learning_rate": 2e-05,
+      "loss": 0.04589385,
+      "step": 1302
+    },
+    {
+      "epoch": 2.606,
+      "grad_norm": 2.1896462440490723,
+      "learning_rate": 2e-05,
+      "loss": 0.05265369,
+      "step": 1303
+    },
+    {
+      "epoch": 2.608,
+      "grad_norm": 2.137312173843384,
+      "learning_rate": 2e-05,
+      "loss": 0.05405066,
+      "step": 1304
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.550561547279358,
+      "learning_rate": 2e-05,
+      "loss": 0.05104889,
+      "step": 1305
+    },
+    {
+      "epoch": 2.612,
+      "grad_norm": 1.8577524423599243,
+      "learning_rate": 2e-05,
+      "loss": 0.06039685,
+      "step": 1306
+    },
+    {
+      "epoch": 2.614,
+      "grad_norm": 2.0149245262145996,
+      "learning_rate": 2e-05,
+      "loss": 0.05752065,
+      "step": 1307
+    },
+    {
+      "epoch": 2.616,
+      "grad_norm": 1.2771774530410767,
+      "learning_rate": 2e-05,
+      "loss": 0.04627162,
+      "step": 1308
+    },
+    {
+      "epoch": 2.618,
+      "grad_norm": 1.2289670705795288,
+      "learning_rate": 2e-05,
+      "loss": 0.04631304,
+      "step": 1309
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.5598198175430298,
+      "learning_rate": 2e-05,
+      "loss": 0.04497156,
+      "step": 1310
+    },
+    {
+      "epoch": 2.622,
+      "grad_norm": 2.00862717628479,
+      "learning_rate": 2e-05,
+      "loss": 0.0671872,
+      "step": 1311
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 1.549716591835022,
+      "learning_rate": 2e-05,
+      "loss": 0.04912385,
+      "step": 1312
+    },
+    {
+      "epoch": 2.626,
+      "grad_norm": 1.7805136442184448,
+      "learning_rate": 2e-05,
+      "loss": 0.04953271,
+      "step": 1313
+    },
+    {
+      "epoch": 2.628,
+      "grad_norm": 1.4479440450668335,
+      "learning_rate": 2e-05,
+      "loss": 0.03930682,
+      "step": 1314
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.026073932647705,
+      "learning_rate": 2e-05,
+      "loss": 0.0562569,
+      "step": 1315
+    },
+    {
+      "epoch": 2.632,
+      "grad_norm": 1.291940450668335,
+      "learning_rate": 2e-05,
+      "loss": 0.05585942,
+      "step": 1316
+    },
+    {
+      "epoch": 2.634,
+      "grad_norm": 1.6085025072097778,
+      "learning_rate": 2e-05,
+      "loss": 0.05520656,
+      "step": 1317
+    },
+    {
+      "epoch": 2.636,
+      "grad_norm": 1.7852100133895874,
+      "learning_rate": 2e-05,
+      "loss": 0.03830192,
+      "step": 1318
+    },
+    {
+      "epoch": 2.638,
+      "grad_norm": 2.1503891944885254,
+      "learning_rate": 2e-05,
+      "loss": 0.05418783,
+      "step": 1319
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.5858957767486572,
+      "learning_rate": 2e-05,
+      "loss": 0.05161075,
+      "step": 1320
+    },
+    {
+      "epoch": 2.642,
+      "grad_norm": 1.517919659614563,
+      "learning_rate": 2e-05,
+      "loss": 0.04599277,
+      "step": 1321
+    },
+    {
+      "epoch": 2.644,
+      "grad_norm": 2.923825979232788,
+      "learning_rate": 2e-05,
+      "loss": 0.05314268,
+      "step": 1322
+    },
+    {
+      "epoch": 2.646,
+      "grad_norm": 1.6294347047805786,
+      "learning_rate": 2e-05,
+      "loss": 0.04354763,
+      "step": 1323
+    },
+    {
+      "epoch": 2.648,
+      "grad_norm": 2.0495193004608154,
+      "learning_rate": 2e-05,
+      "loss": 0.05836939,
+      "step": 1324
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.94076669216156,
+      "learning_rate": 2e-05,
+      "loss": 0.04274346,
+      "step": 1325
+    },
+    {
+      "epoch": 2.652,
+      "grad_norm": 2.360132932662964,
+      "learning_rate": 2e-05,
+      "loss": 0.0647784,
+      "step": 1326
+    },
+    {
+      "epoch": 2.654,
+      "grad_norm": 1.8153984546661377,
+      "learning_rate": 2e-05,
+      "loss": 0.04720091,
+      "step": 1327
+    },
+    {
+      "epoch": 2.656,
+      "grad_norm": 1.725906252861023,
+      "learning_rate": 2e-05,
+      "loss": 0.04768829,
+      "step": 1328
+    },
+    {
+      "epoch": 2.658,
+      "grad_norm": 3.0785069465637207,
+      "learning_rate": 2e-05,
+      "loss": 0.05972423,
+      "step": 1329
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.830115556716919,
+      "learning_rate": 2e-05,
+      "loss": 0.07905363,
+      "step": 1330
+    },
+    {
+      "epoch": 2.662,
+      "grad_norm": 1.4237239360809326,
+      "learning_rate": 2e-05,
+      "loss": 0.05330458,
+      "step": 1331
+    },
+    {
+      "epoch": 2.664,
+      "grad_norm": 1.5219199657440186,
+      "learning_rate": 2e-05,
+      "loss": 0.05649319,
+      "step": 1332
+    },
+    {
+      "epoch": 2.666,
+      "grad_norm": 1.4314558506011963,
+      "learning_rate": 2e-05,
+      "loss": 0.03354269,
+      "step": 1333
+    },
+    {
+      "epoch": 2.668,
+      "grad_norm": 2.2276740074157715,
+      "learning_rate": 2e-05,
+      "loss": 0.07491125,
+      "step": 1334
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.2556209564208984,
+      "learning_rate": 2e-05,
+      "loss": 0.0776452,
+      "step": 1335
+    },
+    {
+      "epoch": 2.672,
+      "grad_norm": 1.8496445417404175,
+      "learning_rate": 2e-05,
+      "loss": 0.04727801,
+      "step": 1336
+    },
+    {
+      "epoch": 2.674,
+      "grad_norm": 2.769913673400879,
+      "learning_rate": 2e-05,
+      "loss": 0.07160389,
+      "step": 1337
+    },
+    {
+      "epoch": 2.676,
+      "grad_norm": 1.8476206064224243,
+      "learning_rate": 2e-05,
+      "loss": 0.04567086,
+      "step": 1338
+    },
+    {
+      "epoch": 2.678,
+      "grad_norm": 1.8353848457336426,
+      "learning_rate": 2e-05,
+      "loss": 0.05308025,
+      "step": 1339
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.6821805238723755,
+      "learning_rate": 2e-05,
+      "loss": 0.04728777,
+      "step": 1340
+    },
+    {
+      "epoch": 2.682,
+      "grad_norm": 2.2069616317749023,
+      "learning_rate": 2e-05,
+      "loss": 0.05139998,
+      "step": 1341
+    },
+    {
+      "epoch": 2.684,
+      "grad_norm": 3.2691590785980225,
+      "learning_rate": 2e-05,
+      "loss": 0.03943311,
+      "step": 1342
+    },
+    {
+      "epoch": 2.686,
+      "grad_norm": 1.286525845527649,
+      "learning_rate": 2e-05,
+      "loss": 0.04956998,
+      "step": 1343
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 1.5744187831878662,
+      "learning_rate": 2e-05,
+      "loss": 0.04305596,
+      "step": 1344
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.1982498168945312,
+      "learning_rate": 2e-05,
+      "loss": 0.06531291,
+      "step": 1345
+    },
+    {
+      "epoch": 2.692,
+      "grad_norm": 1.5973800420761108,
+      "learning_rate": 2e-05,
+      "loss": 0.06064547,
+      "step": 1346
+    },
+    {
+      "epoch": 2.694,
+      "grad_norm": 1.2219719886779785,
+      "learning_rate": 2e-05,
+      "loss": 0.04524098,
+      "step": 1347
+    },
+    {
+      "epoch": 2.6959999999999997,
+      "grad_norm": 2.1111559867858887,
+      "learning_rate": 2e-05,
+      "loss": 0.07144919,
+      "step": 1348
+    },
+    {
+      "epoch": 2.698,
+      "grad_norm": 1.7122925519943237,
+      "learning_rate": 2e-05,
+      "loss": 0.04555336,
+      "step": 1349
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.607475757598877,
+      "learning_rate": 2e-05,
+      "loss": 0.05808245,
+      "step": 1350
+    },
+    {
+      "epoch": 2.702,
+      "grad_norm": 1.6828935146331787,
+      "learning_rate": 2e-05,
+      "loss": 0.05782773,
+      "step": 1351
+    },
+    {
+      "epoch": 2.7039999999999997,
+      "grad_norm": 1.7249581813812256,
+      "learning_rate": 2e-05,
+      "loss": 0.0577189,
+      "step": 1352
+    },
+    {
+      "epoch": 2.706,
+      "grad_norm": 1.476163387298584,
+      "learning_rate": 2e-05,
+      "loss": 0.0563928,
+      "step": 1353
+    },
+    {
+      "epoch": 2.708,
+      "grad_norm": 1.3817315101623535,
+      "learning_rate": 2e-05,
+      "loss": 0.04125504,
+      "step": 1354
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.9194447994232178,
+      "learning_rate": 2e-05,
+      "loss": 0.06330554,
+      "step": 1355
+    },
+    {
+      "epoch": 2.7119999999999997,
+      "grad_norm": 1.357925534248352,
+      "learning_rate": 2e-05,
+      "loss": 0.04602174,
+      "step": 1356
+    },
+    {
+      "epoch": 2.714,
+      "grad_norm": 2.3565990924835205,
+      "learning_rate": 2e-05,
+      "loss": 0.05163093,
+      "step": 1357
+    },
+    {
+      "epoch": 2.716,
+      "grad_norm": 1.6682748794555664,
+      "learning_rate": 2e-05,
+      "loss": 0.03820909,
+      "step": 1358
+    },
+    {
+      "epoch": 2.718,
+      "grad_norm": 2.4361672401428223,
+      "learning_rate": 2e-05,
+      "loss": 0.05794044,
+      "step": 1359
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 2.8087456226348877,
+      "learning_rate": 2e-05,
+      "loss": 0.0474041,
+      "step": 1360
+    },
+    {
+      "epoch": 2.722,
+      "grad_norm": 2.0830671787261963,
+      "learning_rate": 2e-05,
+      "loss": 0.0617595,
+      "step": 1361
+    },
+    {
+      "epoch": 2.724,
+      "grad_norm": 2.5234179496765137,
+      "learning_rate": 2e-05,
+      "loss": 0.07854919,
+      "step": 1362
+    },
+    {
+      "epoch": 2.726,
+      "grad_norm": 2.3360183238983154,
+      "learning_rate": 2e-05,
+      "loss": 0.0680959,
+      "step": 1363
+    },
+    {
+      "epoch": 2.7279999999999998,
+      "grad_norm": 1.46207594871521,
+      "learning_rate": 2e-05,
+      "loss": 0.05441956,
+      "step": 1364
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.2000892162322998,
+      "learning_rate": 2e-05,
+      "loss": 0.03472488,
+      "step": 1365
+    },
+    {
+      "epoch": 2.732,
+      "grad_norm": 1.1089409589767456,
+      "learning_rate": 2e-05,
+      "loss": 0.03588956,
+      "step": 1366
+    },
+    {
+      "epoch": 2.734,
+      "grad_norm": 1.484050989151001,
+      "learning_rate": 2e-05,
+      "loss": 0.0420528,
+      "step": 1367
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 2.1856143474578857,
+      "learning_rate": 2e-05,
+      "loss": 0.07063307,
+      "step": 1368
+    },
+    {
+      "epoch": 2.738,
+      "grad_norm": 1.3538694381713867,
+      "learning_rate": 2e-05,
+      "loss": 0.04442645,
+      "step": 1369
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.1728174686431885,
+      "learning_rate": 2e-05,
+      "loss": 0.03758407,
+      "step": 1370
+    },
+    {
+      "epoch": 2.742,
+      "grad_norm": 1.5417742729187012,
+      "learning_rate": 2e-05,
+      "loss": 0.05347462,
+      "step": 1371
+    },
+    {
+      "epoch": 2.7439999999999998,
+      "grad_norm": 1.2521121501922607,
+      "learning_rate": 2e-05,
+      "loss": 0.03596249,
+      "step": 1372
+    },
+    {
+      "epoch": 2.746,
+      "grad_norm": 1.7495527267456055,
+      "learning_rate": 2e-05,
+      "loss": 0.05753302,
+      "step": 1373
+    },
+    {
+      "epoch": 2.748,
+      "grad_norm": 1.933049201965332,
+      "learning_rate": 2e-05,
+      "loss": 0.04527674,
+      "step": 1374
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.8900275230407715,
+      "learning_rate": 2e-05,
+      "loss": 0.06932513,
+      "step": 1375
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 1.453918695449829,
+      "learning_rate": 2e-05,
+      "loss": 0.04357269,
+      "step": 1376
+    },
+    {
+      "epoch": 2.754,
+      "grad_norm": 1.4563038349151611,
+      "learning_rate": 2e-05,
+      "loss": 0.05056226,
+      "step": 1377
+    },
+    {
+      "epoch": 2.7560000000000002,
+      "grad_norm": 1.504270076751709,
+      "learning_rate": 2e-05,
+      "loss": 0.05600676,
+      "step": 1378
+    },
+    {
+      "epoch": 2.758,
+      "grad_norm": 1.9855608940124512,
+      "learning_rate": 2e-05,
+      "loss": 0.05795199,
+      "step": 1379
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.656665563583374,
+      "learning_rate": 2e-05,
+      "loss": 0.05140248,
+      "step": 1380
+    },
+    {
+      "epoch": 2.762,
+      "grad_norm": 1.7435170412063599,
+      "learning_rate": 2e-05,
+      "loss": 0.04258007,
+      "step": 1381
+    },
+    {
+      "epoch": 2.7640000000000002,
+      "grad_norm": 1.655081033706665,
+      "learning_rate": 2e-05,
+      "loss": 0.0503007,
+      "step": 1382
+    },
+    {
+      "epoch": 2.766,
+      "grad_norm": 1.523956298828125,
+      "learning_rate": 2e-05,
+      "loss": 0.0550363,
+      "step": 1383
+    },
+    {
+      "epoch": 2.768,
+      "grad_norm": 1.4899818897247314,
+      "learning_rate": 2e-05,
+      "loss": 0.04842805,
+      "step": 1384
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.0813798904418945,
+      "learning_rate": 2e-05,
+      "loss": 0.05239079,
+      "step": 1385
+    },
+    {
+      "epoch": 2.7720000000000002,
+      "grad_norm": 1.5431971549987793,
+      "learning_rate": 2e-05,
+      "loss": 0.05558814,
+      "step": 1386
+    },
+    {
+      "epoch": 2.774,
+      "grad_norm": 1.7558424472808838,
+      "learning_rate": 2e-05,
+      "loss": 0.04031524,
+      "step": 1387
+    },
+    {
+      "epoch": 2.776,
+      "grad_norm": 0.9710761308670044,
+      "learning_rate": 2e-05,
+      "loss": 0.02542916,
+      "step": 1388
+    },
+    {
+      "epoch": 2.778,
+      "grad_norm": 1.3783509731292725,
+      "learning_rate": 2e-05,
+      "loss": 0.05585878,
+      "step": 1389
+    },
+    {
+      "epoch": 2.7800000000000002,
+      "grad_norm": 1.7113022804260254,
+      "learning_rate": 2e-05,
+      "loss": 0.06848748,
+      "step": 1390
+    },
+    {
+      "epoch": 2.782,
+      "grad_norm": 1.9026986360549927,
+      "learning_rate": 2e-05,
+      "loss": 0.04627277,
+      "step": 1391
+    },
+    {
+      "epoch": 2.784,
+      "grad_norm": 2.1564736366271973,
+      "learning_rate": 2e-05,
+      "loss": 0.04546306,
+      "step": 1392
+    },
+    {
+      "epoch": 2.786,
+      "grad_norm": 1.8871279954910278,
+      "learning_rate": 2e-05,
+      "loss": 0.05558582,
+      "step": 1393
+    },
+    {
+      "epoch": 2.7880000000000003,
+      "grad_norm": 1.3954287767410278,
+      "learning_rate": 2e-05,
+      "loss": 0.05387101,
+      "step": 1394
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.9269911050796509,
+      "learning_rate": 2e-05,
+      "loss": 0.0572519,
+      "step": 1395
+    },
+    {
+      "epoch": 2.792,
+      "grad_norm": 1.7022475004196167,
+      "learning_rate": 2e-05,
+      "loss": 0.04831417,
+      "step": 1396
+    },
+    {
+      "epoch": 2.794,
+      "grad_norm": 1.7882295846939087,
+      "learning_rate": 2e-05,
+      "loss": 0.03874616,
+      "step": 1397
+    },
+    {
+      "epoch": 2.7960000000000003,
+      "grad_norm": 1.4796777963638306,
+      "learning_rate": 2e-05,
+      "loss": 0.03719701,
+      "step": 1398
+    },
+    {
+      "epoch": 2.798,
+      "grad_norm": 1.5420184135437012,
+      "learning_rate": 2e-05,
+      "loss": 0.05557998,
+      "step": 1399
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.8424732685089111,
+      "learning_rate": 2e-05,
+      "loss": 0.05429619,
+      "step": 1400
+    },
+    {
+      "epoch": 2.802,
+      "grad_norm": 1.4261318445205688,
+      "learning_rate": 2e-05,
+      "loss": 0.05679541,
+      "step": 1401
+    },
+    {
+      "epoch": 2.8040000000000003,
+      "grad_norm": 1.93724524974823,
+      "learning_rate": 2e-05,
+      "loss": 0.04712646,
+      "step": 1402
+    },
+    {
+      "epoch": 2.806,
+      "grad_norm": 1.8298678398132324,
+      "learning_rate": 2e-05,
+      "loss": 0.0460495,
+      "step": 1403
+    },
+    {
+      "epoch": 2.808,
+      "grad_norm": 2.4823086261749268,
+      "learning_rate": 2e-05,
+      "loss": 0.05720939,
+      "step": 1404
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 2.05409574508667,
+      "learning_rate": 2e-05,
+      "loss": 0.06287205,
+      "step": 1405
+    },
+    {
+      "epoch": 2.8120000000000003,
+      "grad_norm": 1.9300044775009155,
+      "learning_rate": 2e-05,
+      "loss": 0.05220521,
+      "step": 1406
+    },
+    {
+      "epoch": 2.814,
+      "grad_norm": 1.6382372379302979,
+      "learning_rate": 2e-05,
+      "loss": 0.04585889,
+      "step": 1407
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 1.6744272708892822,
+      "learning_rate": 2e-05,
+      "loss": 0.05738734,
+      "step": 1408
+    },
+    {
+      "epoch": 2.818,
+      "grad_norm": 1.097986102104187,
+      "learning_rate": 2e-05,
+      "loss": 0.03808478,
+      "step": 1409
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.8194561004638672,
+      "learning_rate": 2e-05,
+      "loss": 0.05379384,
+      "step": 1410
+    },
+    {
+      "epoch": 2.822,
+      "grad_norm": 1.1608667373657227,
+      "learning_rate": 2e-05,
+      "loss": 0.04018103,
+      "step": 1411
+    },
+    {
+      "epoch": 2.824,
+      "grad_norm": 1.5536198616027832,
+      "learning_rate": 2e-05,
+      "loss": 0.0568902,
+      "step": 1412
+    },
+    {
+      "epoch": 2.826,
+      "grad_norm": 1.306771159172058,
+      "learning_rate": 2e-05,
+      "loss": 0.04700731,
+      "step": 1413
+    },
+    {
+      "epoch": 2.828,
+      "grad_norm": 1.4206809997558594,
+      "learning_rate": 2e-05,
+      "loss": 0.03569669,
+      "step": 1414
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.00671124458313,
+      "learning_rate": 2e-05,
+      "loss": 0.0654131,
+      "step": 1415
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 1.3739982843399048,
+      "learning_rate": 2e-05,
+      "loss": 0.05260873,
+      "step": 1416
+    },
+    {
+      "epoch": 2.834,
+      "grad_norm": 1.3874986171722412,
+      "learning_rate": 2e-05,
+      "loss": 0.038274,
+      "step": 1417
+    },
+    {
+      "epoch": 2.836,
+      "grad_norm": 1.056638479232788,
+      "learning_rate": 2e-05,
+      "loss": 0.03164873,
+      "step": 1418
+    },
+    {
+      "epoch": 2.838,
+      "grad_norm": 2.646418809890747,
+      "learning_rate": 2e-05,
+      "loss": 0.0538192,
+      "step": 1419
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.1970181465148926,
+      "learning_rate": 2e-05,
+      "loss": 0.03613362,
+      "step": 1420
+    },
+    {
+      "epoch": 2.842,
+      "grad_norm": 1.7380998134613037,
+      "learning_rate": 2e-05,
+      "loss": 0.05502573,
+      "step": 1421
+    },
+    {
+      "epoch": 2.844,
+      "grad_norm": 1.64398193359375,
+      "learning_rate": 2e-05,
+      "loss": 0.03875776,
+      "step": 1422
+    },
+    {
+      "epoch": 2.846,
+      "grad_norm": 2.1740400791168213,
+      "learning_rate": 2e-05,
+      "loss": 0.05134468,
+      "step": 1423
+    },
+    {
+      "epoch": 2.848,
+      "grad_norm": 2.8908863067626953,
+      "learning_rate": 2e-05,
+      "loss": 0.08046682,
+      "step": 1424
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.89336359500885,
+      "learning_rate": 2e-05,
+      "loss": 0.05308978,
+      "step": 1425
+    },
+    {
+      "epoch": 2.852,
+      "grad_norm": 2.1659324169158936,
+      "learning_rate": 2e-05,
+      "loss": 0.0364816,
+      "step": 1426
+    },
+    {
+      "epoch": 2.854,
+      "grad_norm": 1.402146339416504,
+      "learning_rate": 2e-05,
+      "loss": 0.04808577,
+      "step": 1427
+    },
+    {
+      "epoch": 2.856,
+      "grad_norm": 1.9512687921524048,
+      "learning_rate": 2e-05,
+      "loss": 0.05041477,
+      "step": 1428
+    },
+    {
+      "epoch": 2.858,
+      "grad_norm": 1.3872478008270264,
+      "learning_rate": 2e-05,
+      "loss": 0.04761492,
+      "step": 1429
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.659584403038025,
+      "learning_rate": 2e-05,
+      "loss": 0.06967552,
+      "step": 1430
+    },
+    {
+      "epoch": 2.862,
+      "grad_norm": 1.8425308465957642,
+      "learning_rate": 2e-05,
+      "loss": 0.04844275,
+      "step": 1431
+    },
+    {
+      "epoch": 2.864,
+      "grad_norm": 1.3185906410217285,
+      "learning_rate": 2e-05,
+      "loss": 0.04250592,
+      "step": 1432
+    },
+    {
+      "epoch": 2.866,
+      "grad_norm": 1.1444284915924072,
+      "learning_rate": 2e-05,
+      "loss": 0.03224308,
+      "step": 1433
+    },
+    {
+      "epoch": 2.868,
+      "grad_norm": 1.4065706729888916,
+      "learning_rate": 2e-05,
+      "loss": 0.05243749,
+      "step": 1434
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.1064388751983643,
+      "learning_rate": 2e-05,
+      "loss": 0.02927379,
+      "step": 1435
+    },
+    {
+      "epoch": 2.872,
+      "grad_norm": 1.6056631803512573,
+      "learning_rate": 2e-05,
+      "loss": 0.03909209,
+      "step": 1436
+    },
+    {
+      "epoch": 2.874,
+      "grad_norm": 1.3546314239501953,
+      "learning_rate": 2e-05,
+      "loss": 0.0485357,
+      "step": 1437
+    },
+    {
+      "epoch": 2.876,
+      "grad_norm": 1.5769225358963013,
+      "learning_rate": 2e-05,
+      "loss": 0.05911487,
+      "step": 1438
+    },
+    {
+      "epoch": 2.878,
+      "grad_norm": 1.589120864868164,
+      "learning_rate": 2e-05,
+      "loss": 0.04771549,
+      "step": 1439
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 2.1113979816436768,
+      "learning_rate": 2e-05,
+      "loss": 0.07277437,
+      "step": 1440
+    },
+    {
+      "epoch": 2.882,
+      "grad_norm": 2.021388530731201,
+      "learning_rate": 2e-05,
+      "loss": 0.04649474,
+      "step": 1441
+    },
+    {
+      "epoch": 2.884,
+      "grad_norm": 1.8449045419692993,
+      "learning_rate": 2e-05,
+      "loss": 0.04096931,
+      "step": 1442
+    },
+    {
+      "epoch": 2.886,
+      "grad_norm": 2.4654793739318848,
+      "learning_rate": 2e-05,
+      "loss": 0.06707292,
+      "step": 1443
+    },
+    {
+      "epoch": 2.888,
+      "grad_norm": 1.8038156032562256,
+      "learning_rate": 2e-05,
+      "loss": 0.05822627,
+      "step": 1444
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.6336687803268433,
+      "learning_rate": 2e-05,
+      "loss": 0.04999952,
+      "step": 1445
+    },
+    {
+      "epoch": 2.892,
+      "grad_norm": 1.4008294343948364,
+      "learning_rate": 2e-05,
+      "loss": 0.04774325,
+      "step": 1446
+    },
+    {
+      "epoch": 2.894,
+      "grad_norm": 1.8084869384765625,
+      "learning_rate": 2e-05,
+      "loss": 0.07069649,
+      "step": 1447
+    },
+    {
+      "epoch": 2.896,
+      "grad_norm": 3.3638651371002197,
+      "learning_rate": 2e-05,
+      "loss": 0.09050111,
+      "step": 1448
+    },
+    {
+      "epoch": 2.898,
+      "grad_norm": 1.1725375652313232,
+      "learning_rate": 2e-05,
+      "loss": 0.04985755,
+      "step": 1449
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.0440391302108765,
+      "learning_rate": 2e-05,
+      "loss": 0.05144562,
+      "step": 1450
+    },
+    {
+      "epoch": 2.902,
+      "grad_norm": 2.280170440673828,
+      "learning_rate": 2e-05,
+      "loss": 0.04932537,
+      "step": 1451
+    },
+    {
+      "epoch": 2.904,
+      "grad_norm": 1.461732268333435,
+      "learning_rate": 2e-05,
+      "loss": 0.04785193,
+      "step": 1452
+    },
+    {
+      "epoch": 2.906,
+      "grad_norm": 1.9254556894302368,
+      "learning_rate": 2e-05,
+      "loss": 0.05030049,
+      "step": 1453
+    },
+    {
+      "epoch": 2.908,
+      "grad_norm": 1.182039737701416,
+      "learning_rate": 2e-05,
+      "loss": 0.03006881,
+      "step": 1454
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 2.5947253704071045,
+      "learning_rate": 2e-05,
+      "loss": 0.04492339,
+      "step": 1455
+    },
+    {
+      "epoch": 2.912,
+      "grad_norm": 1.543723225593567,
+      "learning_rate": 2e-05,
+      "loss": 0.05002248,
+      "step": 1456
+    },
+    {
+      "epoch": 2.914,
+      "grad_norm": 1.1024924516677856,
+      "learning_rate": 2e-05,
+      "loss": 0.041133,
+      "step": 1457
+    },
+    {
+      "epoch": 2.916,
+      "grad_norm": 1.7310285568237305,
+      "learning_rate": 2e-05,
+      "loss": 0.05967389,
+      "step": 1458
+    },
+    {
+      "epoch": 2.918,
+      "grad_norm": 2.281721830368042,
+      "learning_rate": 2e-05,
+      "loss": 0.06781383,
+      "step": 1459
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.6971452236175537,
+      "learning_rate": 2e-05,
+      "loss": 0.05646857,
+      "step": 1460
+    },
+    {
+      "epoch": 2.922,
+      "grad_norm": 1.741514801979065,
+      "learning_rate": 2e-05,
+      "loss": 0.07292978,
+      "step": 1461
+    },
+    {
+      "epoch": 2.924,
+      "grad_norm": 1.4831644296646118,
+      "learning_rate": 2e-05,
+      "loss": 0.03878976,
+      "step": 1462
+    },
+    {
+      "epoch": 2.926,
+      "grad_norm": 1.676571011543274,
+      "learning_rate": 2e-05,
+      "loss": 0.05792604,
+      "step": 1463
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 1.8864375352859497,
+      "learning_rate": 2e-05,
+      "loss": 0.06544764,
+      "step": 1464
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 2.198659658432007,
+      "learning_rate": 2e-05,
+      "loss": 0.06713847,
+      "step": 1465
+    },
+    {
+      "epoch": 2.932,
+      "grad_norm": 1.498012900352478,
+      "learning_rate": 2e-05,
+      "loss": 0.03821286,
+      "step": 1466
+    },
+    {
+      "epoch": 2.934,
+      "grad_norm": 1.4330593347549438,
+      "learning_rate": 2e-05,
+      "loss": 0.06282357,
+      "step": 1467
+    },
+    {
+      "epoch": 2.936,
+      "grad_norm": 1.3292213678359985,
+      "learning_rate": 2e-05,
+      "loss": 0.06121788,
+      "step": 1468
+    },
+    {
+      "epoch": 2.9379999999999997,
+      "grad_norm": 1.5773099660873413,
+      "learning_rate": 2e-05,
+      "loss": 0.03123096,
+      "step": 1469
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.3884388208389282,
+      "learning_rate": 2e-05,
+      "loss": 0.04133505,
+      "step": 1470
+    },
+    {
+      "epoch": 2.942,
+      "grad_norm": 1.1826488971710205,
+      "learning_rate": 2e-05,
+      "loss": 0.03776342,
+      "step": 1471
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 1.5211957693099976,
+      "learning_rate": 2e-05,
+      "loss": 0.04243449,
+      "step": 1472
+    },
+    {
+      "epoch": 2.9459999999999997,
+      "grad_norm": 2.148106336593628,
+      "learning_rate": 2e-05,
+      "loss": 0.06043831,
+      "step": 1473
+    },
+    {
+      "epoch": 2.948,
+      "grad_norm": 1.1765345335006714,
+      "learning_rate": 2e-05,
+      "loss": 0.04386732,
+      "step": 1474
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.502057433128357,
+      "learning_rate": 2e-05,
+      "loss": 0.0411534,
+      "step": 1475
+    },
+    {
+      "epoch": 2.952,
+      "grad_norm": 1.6430046558380127,
+      "learning_rate": 2e-05,
+      "loss": 0.037657,
+      "step": 1476
+    },
+    {
+      "epoch": 2.9539999999999997,
+      "grad_norm": 1.535294771194458,
+      "learning_rate": 2e-05,
+      "loss": 0.05130748,
+      "step": 1477
+    },
+    {
+      "epoch": 2.956,
+      "grad_norm": 1.9392225742340088,
+      "learning_rate": 2e-05,
+      "loss": 0.05215064,
+      "step": 1478
+    },
+    {
+      "epoch": 2.958,
+      "grad_norm": 1.1696999073028564,
+      "learning_rate": 2e-05,
+      "loss": 0.02946492,
+      "step": 1479
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.414081335067749,
+      "learning_rate": 2e-05,
+      "loss": 0.04422535,
+      "step": 1480
+    },
+    {
+      "epoch": 2.9619999999999997,
+      "grad_norm": 1.719420075416565,
+      "learning_rate": 2e-05,
+      "loss": 0.05473565,
+      "step": 1481
+    },
+    {
+      "epoch": 2.964,
+      "grad_norm": 2.0351169109344482,
+      "learning_rate": 2e-05,
+      "loss": 0.04414459,
+      "step": 1482
+    },
+    {
+      "epoch": 2.966,
+      "grad_norm": 1.012802004814148,
+      "learning_rate": 2e-05,
+      "loss": 0.03308199,
+      "step": 1483
+    },
+    {
+      "epoch": 2.968,
+      "grad_norm": 2.026688814163208,
+      "learning_rate": 2e-05,
+      "loss": 0.04603078,
+      "step": 1484
+    },
+    {
+      "epoch": 2.9699999999999998,
+      "grad_norm": 2.1403253078460693,
+      "learning_rate": 2e-05,
+      "loss": 0.04239823,
+      "step": 1485
+    },
+    {
+      "epoch": 2.972,
+      "grad_norm": 2.087897539138794,
+      "learning_rate": 2e-05,
+      "loss": 0.05140383,
+      "step": 1486
+    },
+    {
+      "epoch": 2.974,
+      "grad_norm": 3.251877546310425,
+      "learning_rate": 2e-05,
+      "loss": 0.05295721,
+      "step": 1487
+    },
+    {
+      "epoch": 2.976,
+      "grad_norm": 2.2770659923553467,
+      "learning_rate": 2e-05,
+      "loss": 0.06836365,
+      "step": 1488
+    },
+    {
+      "epoch": 2.9779999999999998,
+      "grad_norm": 1.4098083972930908,
+      "learning_rate": 2e-05,
+      "loss": 0.03099969,
+      "step": 1489
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.4418541193008423,
+      "learning_rate": 2e-05,
+      "loss": 0.0420429,
+      "step": 1490
+    },
+    {
+      "epoch": 2.982,
+      "grad_norm": 1.7272592782974243,
+      "learning_rate": 2e-05,
+      "loss": 0.05751514,
+      "step": 1491
+    },
+    {
+      "epoch": 2.984,
+      "grad_norm": 1.1021864414215088,
+      "learning_rate": 2e-05,
+      "loss": 0.03865705,
+      "step": 1492
+    },
+    {
+      "epoch": 2.9859999999999998,
+      "grad_norm": 1.4109399318695068,
+      "learning_rate": 2e-05,
+      "loss": 0.05035442,
+      "step": 1493
+    },
+    {
+      "epoch": 2.988,
+      "grad_norm": 1.5310124158859253,
+      "learning_rate": 2e-05,
+      "loss": 0.04429567,
+      "step": 1494
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.1835428476333618,
+      "learning_rate": 2e-05,
+      "loss": 0.04300354,
+      "step": 1495
+    },
+    {
+      "epoch": 2.992,
+      "grad_norm": 1.6832672357559204,
+      "learning_rate": 2e-05,
+      "loss": 0.06267562,
+      "step": 1496
+    },
+    {
+      "epoch": 2.9939999999999998,
+      "grad_norm": 1.7960203886032104,
+      "learning_rate": 2e-05,
+      "loss": 0.04719928,
+      "step": 1497
+    },
+    {
+      "epoch": 2.996,
+      "grad_norm": 1.2733722925186157,
+      "learning_rate": 2e-05,
+      "loss": 0.04313634,
+      "step": 1498
+    },
+    {
+      "epoch": 2.998,
+      "grad_norm": 2.289060592651367,
+      "learning_rate": 2e-05,
+      "loss": 0.04968451,
+      "step": 1499
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.4316084384918213,
+      "learning_rate": 2e-05,
+      "loss": 0.04869864,
+      "step": 1500
+    },
+    {
+      "epoch": 3.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.992,
+        "AngleClassification_2": 0.52,
+        "AngleClassification_3": 0.49101796407185627,
+        "Equal_1": 0.97,
+        "Equal_2": 0.8323353293413174,
+        "Equal_3": 0.720558882235529,
+        "LineComparison_1": 0.992,
+        "LineComparison_2": 0.9740518962075848,
+        "LineComparison_3": 0.8582834331337326,
+        "Parallel_1": 0.9458917835671342,
+        "Parallel_2": 0.9438877755511023,
+        "Parallel_3": 0.744,
+        "Perpendicular_1": 0.934,
+        "Perpendicular_2": 0.41,
+        "Perpendicular_3": 0.19939879759519039,
+        "PointLiesOnCircle_1": 0.9859719438877755,
+        "PointLiesOnCircle_2": 0.9889333333333334,
+        "PointLiesOnCircle_3": 0.7892666666666667,
+        "PointLiesOnLine_1": 0.9799599198396793,
+        "PointLiesOnLine_2": 0.9238476953907816,
+        "PointLiesOnLine_3": 0.49101796407185627
+      },
+      "eval_runtime": 321.7153,
+      "eval_samples_per_second": 32.638,
+      "eval_steps_per_second": 0.653,
+      "step": 1500
+    },
+    {
+      "epoch": 3.002,
+      "grad_norm": 2.2160494327545166,
+      "learning_rate": 2e-05,
+      "loss": 0.04842204,
+      "step": 1501
+    },
+    {
+      "epoch": 3.004,
+      "grad_norm": 2.940279006958008,
+      "learning_rate": 2e-05,
+      "loss": 0.06389765,
+      "step": 1502
+    },
+    {
+      "epoch": 3.006,
+      "grad_norm": 1.4351305961608887,
+      "learning_rate": 2e-05,
+      "loss": 0.03972311,
+      "step": 1503
+    },
+    {
+      "epoch": 3.008,
+      "grad_norm": 1.8173421621322632,
+      "learning_rate": 2e-05,
+      "loss": 0.06714956,
+      "step": 1504
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 1.6168220043182373,
+      "learning_rate": 2e-05,
+      "loss": 0.05728213,
+      "step": 1505
+    },
+    {
+      "epoch": 3.012,
+      "grad_norm": 2.713430166244507,
+      "learning_rate": 2e-05,
+      "loss": 0.08517258,
+      "step": 1506
+    },
+    {
+      "epoch": 3.014,
+      "grad_norm": 1.6535557508468628,
+      "learning_rate": 2e-05,
+      "loss": 0.05091209,
+      "step": 1507
+    },
+    {
+      "epoch": 3.016,
+      "grad_norm": 1.5731172561645508,
+      "learning_rate": 2e-05,
+      "loss": 0.04936452,
+      "step": 1508
+    },
+    {
+      "epoch": 3.018,
+      "grad_norm": 2.121694326400757,
+      "learning_rate": 2e-05,
+      "loss": 0.05214013,
+      "step": 1509
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 1.7850323915481567,
+      "learning_rate": 2e-05,
+      "loss": 0.04803599,
+      "step": 1510
+    },
+    {
+      "epoch": 3.022,
+      "grad_norm": 1.0655592679977417,
+      "learning_rate": 2e-05,
+      "loss": 0.04061773,
+      "step": 1511
+    },
+    {
+      "epoch": 3.024,
+      "grad_norm": 1.7327584028244019,
+      "learning_rate": 2e-05,
+      "loss": 0.05965272,
+      "step": 1512
+    },
+    {
+      "epoch": 3.026,
+      "grad_norm": 1.4035505056381226,
+      "learning_rate": 2e-05,
+      "loss": 0.04205831,
+      "step": 1513
+    },
+    {
+      "epoch": 3.028,
+      "grad_norm": 1.2730896472930908,
+      "learning_rate": 2e-05,
+      "loss": 0.04025223,
+      "step": 1514
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 1.6347522735595703,
+      "learning_rate": 2e-05,
+      "loss": 0.06551866,
+      "step": 1515
+    },
+    {
+      "epoch": 3.032,
+      "grad_norm": 1.24717116355896,
+      "learning_rate": 2e-05,
+      "loss": 0.04352552,
+      "step": 1516
+    },
+    {
+      "epoch": 3.034,
+      "grad_norm": 2.378849744796753,
+      "learning_rate": 2e-05,
+      "loss": 0.05578585,
+      "step": 1517
+    },
+    {
+      "epoch": 3.036,
+      "grad_norm": 1.5153645277023315,
+      "learning_rate": 2e-05,
+      "loss": 0.06007598,
+      "step": 1518
+    },
+    {
+      "epoch": 3.038,
+      "grad_norm": 1.4317395687103271,
+      "learning_rate": 2e-05,
+      "loss": 0.04686758,
+      "step": 1519
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 1.150730013847351,
+      "learning_rate": 2e-05,
+      "loss": 0.03575693,
+      "step": 1520
+    },
+    {
+      "epoch": 3.042,
+      "grad_norm": 1.221483588218689,
+      "learning_rate": 2e-05,
+      "loss": 0.03753293,
+      "step": 1521
+    },
+    {
+      "epoch": 3.044,
+      "grad_norm": 1.595436453819275,
+      "learning_rate": 2e-05,
+      "loss": 0.04876781,
+      "step": 1522
+    },
+    {
+      "epoch": 3.046,
+      "grad_norm": 1.5910844802856445,
+      "learning_rate": 2e-05,
+      "loss": 0.04243572,
+      "step": 1523
+    },
+    {
+      "epoch": 3.048,
+      "grad_norm": 1.333764672279358,
+      "learning_rate": 2e-05,
+      "loss": 0.04689807,
+      "step": 1524
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.244879961013794,
+      "learning_rate": 2e-05,
+      "loss": 0.05209486,
+      "step": 1525
+    },
+    {
+      "epoch": 3.052,
+      "grad_norm": 2.523118734359741,
+      "learning_rate": 2e-05,
+      "loss": 0.05441052,
+      "step": 1526
+    },
+    {
+      "epoch": 3.054,
+      "grad_norm": 1.1830792427062988,
+      "learning_rate": 2e-05,
+      "loss": 0.03513453,
+      "step": 1527
+    },
+    {
+      "epoch": 3.056,
+      "grad_norm": 1.4500566720962524,
+      "learning_rate": 2e-05,
+      "loss": 0.03788603,
+      "step": 1528
+    },
+    {
+      "epoch": 3.058,
+      "grad_norm": 1.9590797424316406,
+      "learning_rate": 2e-05,
+      "loss": 0.07833979,
+      "step": 1529
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 2.1240177154541016,
+      "learning_rate": 2e-05,
+      "loss": 0.06781566,
+      "step": 1530
+    },
+    {
+      "epoch": 3.062,
+      "grad_norm": 2.877790689468384,
+      "learning_rate": 2e-05,
+      "loss": 0.05031575,
+      "step": 1531
+    },
+    {
+      "epoch": 3.064,
+      "grad_norm": 1.827868938446045,
+      "learning_rate": 2e-05,
+      "loss": 0.05153899,
+      "step": 1532
+    },
+    {
+      "epoch": 3.066,
+      "grad_norm": 1.4657886028289795,
+      "learning_rate": 2e-05,
+      "loss": 0.05448882,
+      "step": 1533
+    },
+    {
+      "epoch": 3.068,
+      "grad_norm": 2.052285671234131,
+      "learning_rate": 2e-05,
+      "loss": 0.06243248,
+      "step": 1534
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.7502951622009277,
+      "learning_rate": 2e-05,
+      "loss": 0.02465053,
+      "step": 1535
+    },
+    {
+      "epoch": 3.072,
+      "grad_norm": 1.1672446727752686,
+      "learning_rate": 2e-05,
+      "loss": 0.03131329,
+      "step": 1536
+    },
+    {
+      "epoch": 3.074,
+      "grad_norm": 6.168121337890625,
+      "learning_rate": 2e-05,
+      "loss": 0.05717981,
+      "step": 1537
+    },
+    {
+      "epoch": 3.076,
+      "grad_norm": 1.6621938943862915,
+      "learning_rate": 2e-05,
+      "loss": 0.04681021,
+      "step": 1538
+    },
+    {
+      "epoch": 3.078,
+      "grad_norm": 1.9949926137924194,
+      "learning_rate": 2e-05,
+      "loss": 0.05344428,
+      "step": 1539
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 1.7297910451889038,
+      "learning_rate": 2e-05,
+      "loss": 0.06000186,
+      "step": 1540
+    },
+    {
+      "epoch": 3.082,
+      "grad_norm": 1.9841177463531494,
+      "learning_rate": 2e-05,
+      "loss": 0.05356345,
+      "step": 1541
+    },
+    {
+      "epoch": 3.084,
+      "grad_norm": 1.619390606880188,
+      "learning_rate": 2e-05,
+      "loss": 0.05409247,
+      "step": 1542
+    },
+    {
+      "epoch": 3.086,
+      "grad_norm": 2.220137357711792,
+      "learning_rate": 2e-05,
+      "loss": 0.06488518,
+      "step": 1543
+    },
+    {
+      "epoch": 3.088,
+      "grad_norm": 1.2586517333984375,
+      "learning_rate": 2e-05,
+      "loss": 0.0334148,
+      "step": 1544
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 2.6507513523101807,
+      "learning_rate": 2e-05,
+      "loss": 0.05396133,
+      "step": 1545
+    },
+    {
+      "epoch": 3.092,
+      "grad_norm": 2.484825372695923,
+      "learning_rate": 2e-05,
+      "loss": 0.05893756,
+      "step": 1546
+    },
+    {
+      "epoch": 3.094,
+      "grad_norm": 1.530316710472107,
+      "learning_rate": 2e-05,
+      "loss": 0.04786216,
+      "step": 1547
+    },
+    {
+      "epoch": 3.096,
+      "grad_norm": 1.5662510395050049,
+      "learning_rate": 2e-05,
+      "loss": 0.05811158,
+      "step": 1548
+    },
+    {
+      "epoch": 3.098,
+      "grad_norm": 1.7030950784683228,
+      "learning_rate": 2e-05,
+      "loss": 0.06066509,
+      "step": 1549
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 1.5851967334747314,
+      "learning_rate": 2e-05,
+      "loss": 0.06345069,
+      "step": 1550
+    },
+    {
+      "epoch": 3.102,
+      "grad_norm": 2.7735044956207275,
+      "learning_rate": 2e-05,
+      "loss": 0.0733242,
+      "step": 1551
+    },
+    {
+      "epoch": 3.104,
+      "grad_norm": 1.4052716493606567,
+      "learning_rate": 2e-05,
+      "loss": 0.03768426,
+      "step": 1552
+    },
+    {
+      "epoch": 3.106,
+      "grad_norm": 2.072314977645874,
+      "learning_rate": 2e-05,
+      "loss": 0.06333002,
+      "step": 1553
+    },
+    {
+      "epoch": 3.108,
+      "grad_norm": 1.3217315673828125,
+      "learning_rate": 2e-05,
+      "loss": 0.04815859,
+      "step": 1554
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 1.2543447017669678,
+      "learning_rate": 2e-05,
+      "loss": 0.04736142,
+      "step": 1555
+    },
+    {
+      "epoch": 3.112,
+      "grad_norm": 1.8522039651870728,
+      "learning_rate": 2e-05,
+      "loss": 0.04406268,
+      "step": 1556
+    },
+    {
+      "epoch": 3.114,
+      "grad_norm": 1.7204567193984985,
+      "learning_rate": 2e-05,
+      "loss": 0.04844861,
+      "step": 1557
+    },
+    {
+      "epoch": 3.116,
+      "grad_norm": 1.9749001264572144,
+      "learning_rate": 2e-05,
+      "loss": 0.05746056,
+      "step": 1558
+    },
+    {
+      "epoch": 3.118,
+      "grad_norm": 1.9995803833007812,
+      "learning_rate": 2e-05,
+      "loss": 0.05737771,
+      "step": 1559
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 2.1211681365966797,
+      "learning_rate": 2e-05,
+      "loss": 0.04288347,
+      "step": 1560
+    },
+    {
+      "epoch": 3.122,
+      "grad_norm": 1.4146476984024048,
+      "learning_rate": 2e-05,
+      "loss": 0.05071546,
+      "step": 1561
+    },
+    {
+      "epoch": 3.124,
+      "grad_norm": 1.660503625869751,
+      "learning_rate": 2e-05,
+      "loss": 0.05359786,
+      "step": 1562
+    },
+    {
+      "epoch": 3.126,
+      "grad_norm": 2.073803186416626,
+      "learning_rate": 2e-05,
+      "loss": 0.0541298,
+      "step": 1563
+    },
+    {
+      "epoch": 3.128,
+      "grad_norm": 2.1681854724884033,
+      "learning_rate": 2e-05,
+      "loss": 0.03845184,
+      "step": 1564
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 1.6317102909088135,
+      "learning_rate": 2e-05,
+      "loss": 0.04459539,
+      "step": 1565
+    },
+    {
+      "epoch": 3.132,
+      "grad_norm": 1.2800451517105103,
+      "learning_rate": 2e-05,
+      "loss": 0.05038659,
+      "step": 1566
+    },
+    {
+      "epoch": 3.134,
+      "grad_norm": 1.627054214477539,
+      "learning_rate": 2e-05,
+      "loss": 0.05846292,
+      "step": 1567
+    },
+    {
+      "epoch": 3.136,
+      "grad_norm": 1.5014915466308594,
+      "learning_rate": 2e-05,
+      "loss": 0.0461411,
+      "step": 1568
+    },
+    {
+      "epoch": 3.138,
+      "grad_norm": 1.4676257371902466,
+      "learning_rate": 2e-05,
+      "loss": 0.04794473,
+      "step": 1569
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.8870722055435181,
+      "learning_rate": 2e-05,
+      "loss": 0.0262022,
+      "step": 1570
+    },
+    {
+      "epoch": 3.142,
+      "grad_norm": 1.5116006135940552,
+      "learning_rate": 2e-05,
+      "loss": 0.04896776,
+      "step": 1571
+    },
+    {
+      "epoch": 3.144,
+      "grad_norm": 2.2803430557250977,
+      "learning_rate": 2e-05,
+      "loss": 0.03939671,
+      "step": 1572
+    },
+    {
+      "epoch": 3.146,
+      "grad_norm": 1.6452845335006714,
+      "learning_rate": 2e-05,
+      "loss": 0.06173242,
+      "step": 1573
+    },
+    {
+      "epoch": 3.148,
+      "grad_norm": 1.3423179388046265,
+      "learning_rate": 2e-05,
+      "loss": 0.03313474,
+      "step": 1574
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 1.491223692893982,
+      "learning_rate": 2e-05,
+      "loss": 0.04076514,
+      "step": 1575
+    },
+    {
+      "epoch": 3.152,
+      "grad_norm": 1.7405002117156982,
+      "learning_rate": 2e-05,
+      "loss": 0.06292631,
+      "step": 1576
+    },
+    {
+      "epoch": 3.154,
+      "grad_norm": 2.301616907119751,
+      "learning_rate": 2e-05,
+      "loss": 0.04174115,
+      "step": 1577
+    },
+    {
+      "epoch": 3.156,
+      "grad_norm": 1.4608789682388306,
+      "learning_rate": 2e-05,
+      "loss": 0.03542221,
+      "step": 1578
+    },
+    {
+      "epoch": 3.158,
+      "grad_norm": 1.6150435209274292,
+      "learning_rate": 2e-05,
+      "loss": 0.0463053,
+      "step": 1579
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 1.156436562538147,
+      "learning_rate": 2e-05,
+      "loss": 0.03821928,
+      "step": 1580
+    },
+    {
+      "epoch": 3.162,
+      "grad_norm": 1.483613133430481,
+      "learning_rate": 2e-05,
+      "loss": 0.06052083,
+      "step": 1581
+    },
+    {
+      "epoch": 3.164,
+      "grad_norm": 3.3407390117645264,
+      "learning_rate": 2e-05,
+      "loss": 0.05487797,
+      "step": 1582
+    },
+    {
+      "epoch": 3.166,
+      "grad_norm": 1.3785268068313599,
+      "learning_rate": 2e-05,
+      "loss": 0.04208247,
+      "step": 1583
+    },
+    {
+      "epoch": 3.168,
+      "grad_norm": 2.0314083099365234,
+      "learning_rate": 2e-05,
+      "loss": 0.03377452,
+      "step": 1584
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 2.2946577072143555,
+      "learning_rate": 2e-05,
+      "loss": 0.03864856,
+      "step": 1585
+    },
+    {
+      "epoch": 3.172,
+      "grad_norm": 1.5391055345535278,
+      "learning_rate": 2e-05,
+      "loss": 0.04968462,
+      "step": 1586
+    },
+    {
+      "epoch": 3.174,
+      "grad_norm": 2.3214282989501953,
+      "learning_rate": 2e-05,
+      "loss": 0.0490016,
+      "step": 1587
+    },
+    {
+      "epoch": 3.176,
+      "grad_norm": 2.729959487915039,
+      "learning_rate": 2e-05,
+      "loss": 0.06228638,
+      "step": 1588
+    },
+    {
+      "epoch": 3.178,
+      "grad_norm": 1.3159915208816528,
+      "learning_rate": 2e-05,
+      "loss": 0.04129592,
+      "step": 1589
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 2.102053642272949,
+      "learning_rate": 2e-05,
+      "loss": 0.06420384,
+      "step": 1590
+    },
+    {
+      "epoch": 3.182,
+      "grad_norm": 1.6806936264038086,
+      "learning_rate": 2e-05,
+      "loss": 0.04690688,
+      "step": 1591
+    },
+    {
+      "epoch": 3.184,
+      "grad_norm": 1.5850759744644165,
+      "learning_rate": 2e-05,
+      "loss": 0.05270191,
+      "step": 1592
+    },
+    {
+      "epoch": 3.186,
+      "grad_norm": 1.6031419038772583,
+      "learning_rate": 2e-05,
+      "loss": 0.04544835,
+      "step": 1593
+    },
+    {
+      "epoch": 3.188,
+      "grad_norm": 1.542733907699585,
+      "learning_rate": 2e-05,
+      "loss": 0.05052012,
+      "step": 1594
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 1.3358275890350342,
+      "learning_rate": 2e-05,
+      "loss": 0.04404087,
+      "step": 1595
+    },
+    {
+      "epoch": 3.192,
+      "grad_norm": 3.1650984287261963,
+      "learning_rate": 2e-05,
+      "loss": 0.0582697,
+      "step": 1596
+    },
+    {
+      "epoch": 3.194,
+      "grad_norm": 2.426821231842041,
+      "learning_rate": 2e-05,
+      "loss": 0.06109107,
+      "step": 1597
+    },
+    {
+      "epoch": 3.196,
+      "grad_norm": 1.221462607383728,
+      "learning_rate": 2e-05,
+      "loss": 0.03468029,
+      "step": 1598
+    },
+    {
+      "epoch": 3.198,
+      "grad_norm": 1.832042932510376,
+      "learning_rate": 2e-05,
+      "loss": 0.04503892,
+      "step": 1599
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.6637980937957764,
+      "learning_rate": 2e-05,
+      "loss": 0.05590947,
+      "step": 1600
+    },
+    {
+      "epoch": 3.202,
+      "grad_norm": 1.707131266593933,
+      "learning_rate": 2e-05,
+      "loss": 0.04391931,
+      "step": 1601
+    },
+    {
+      "epoch": 3.204,
+      "grad_norm": 1.2049839496612549,
+      "learning_rate": 2e-05,
+      "loss": 0.03543879,
+      "step": 1602
+    },
+    {
+      "epoch": 3.206,
+      "grad_norm": 1.2202122211456299,
+      "learning_rate": 2e-05,
+      "loss": 0.0452034,
+      "step": 1603
+    },
+    {
+      "epoch": 3.208,
+      "grad_norm": 1.4739149808883667,
+      "learning_rate": 2e-05,
+      "loss": 0.05101965,
+      "step": 1604
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 1.7410937547683716,
+      "learning_rate": 2e-05,
+      "loss": 0.05315409,
+      "step": 1605
+    },
+    {
+      "epoch": 3.212,
+      "grad_norm": 1.943514108657837,
+      "learning_rate": 2e-05,
+      "loss": 0.05091185,
+      "step": 1606
+    },
+    {
+      "epoch": 3.214,
+      "grad_norm": 2.420563220977783,
+      "learning_rate": 2e-05,
+      "loss": 0.04633863,
+      "step": 1607
+    },
+    {
+      "epoch": 3.216,
+      "grad_norm": 1.5087945461273193,
+      "learning_rate": 2e-05,
+      "loss": 0.0458011,
+      "step": 1608
+    },
+    {
+      "epoch": 3.218,
+      "grad_norm": 1.2308595180511475,
+      "learning_rate": 2e-05,
+      "loss": 0.04334211,
+      "step": 1609
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 1.7495505809783936,
+      "learning_rate": 2e-05,
+      "loss": 0.05809302,
+      "step": 1610
+    },
+    {
+      "epoch": 3.222,
+      "grad_norm": 1.3433486223220825,
+      "learning_rate": 2e-05,
+      "loss": 0.0264309,
+      "step": 1611
+    },
+    {
+      "epoch": 3.224,
+      "grad_norm": 2.1560001373291016,
+      "learning_rate": 2e-05,
+      "loss": 0.05610461,
+      "step": 1612
+    },
+    {
+      "epoch": 3.226,
+      "grad_norm": 1.4553375244140625,
+      "learning_rate": 2e-05,
+      "loss": 0.06240574,
+      "step": 1613
+    },
+    {
+      "epoch": 3.228,
+      "grad_norm": 1.6326318979263306,
+      "learning_rate": 2e-05,
+      "loss": 0.04741699,
+      "step": 1614
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 2.0145437717437744,
+      "learning_rate": 2e-05,
+      "loss": 0.06130683,
+      "step": 1615
+    },
+    {
+      "epoch": 3.232,
+      "grad_norm": 2.395770311355591,
+      "learning_rate": 2e-05,
+      "loss": 0.06174326,
+      "step": 1616
+    },
+    {
+      "epoch": 3.234,
+      "grad_norm": 1.4244645833969116,
+      "learning_rate": 2e-05,
+      "loss": 0.04702844,
+      "step": 1617
+    },
+    {
+      "epoch": 3.2359999999999998,
+      "grad_norm": 1.611910104751587,
+      "learning_rate": 2e-05,
+      "loss": 0.05492343,
+      "step": 1618
+    },
+    {
+      "epoch": 3.238,
+      "grad_norm": 1.3452534675598145,
+      "learning_rate": 2e-05,
+      "loss": 0.04002935,
+      "step": 1619
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 1.4256281852722168,
+      "learning_rate": 2e-05,
+      "loss": 0.06138282,
+      "step": 1620
+    },
+    {
+      "epoch": 3.242,
+      "grad_norm": 2.094010353088379,
+      "learning_rate": 2e-05,
+      "loss": 0.05636217,
+      "step": 1621
+    },
+    {
+      "epoch": 3.2439999999999998,
+      "grad_norm": 1.8228808641433716,
+      "learning_rate": 2e-05,
+      "loss": 0.05640348,
+      "step": 1622
+    },
+    {
+      "epoch": 3.246,
+      "grad_norm": 0.991633415222168,
+      "learning_rate": 2e-05,
+      "loss": 0.03676727,
+      "step": 1623
+    },
+    {
+      "epoch": 3.248,
+      "grad_norm": 1.5511411428451538,
+      "learning_rate": 2e-05,
+      "loss": 0.03195164,
+      "step": 1624
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 1.3886339664459229,
+      "learning_rate": 2e-05,
+      "loss": 0.06085005,
+      "step": 1625
+    },
+    {
+      "epoch": 3.252,
+      "grad_norm": 1.6809405088424683,
+      "learning_rate": 2e-05,
+      "loss": 0.06114616,
+      "step": 1626
+    },
+    {
+      "epoch": 3.254,
+      "grad_norm": 1.2058086395263672,
+      "learning_rate": 2e-05,
+      "loss": 0.03140325,
+      "step": 1627
+    },
+    {
+      "epoch": 3.2560000000000002,
+      "grad_norm": 1.2712700366973877,
+      "learning_rate": 2e-05,
+      "loss": 0.05181075,
+      "step": 1628
+    },
+    {
+      "epoch": 3.258,
+      "grad_norm": 1.6511715650558472,
+      "learning_rate": 2e-05,
+      "loss": 0.03143429,
+      "step": 1629
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 1.5783659219741821,
+      "learning_rate": 2e-05,
+      "loss": 0.04343471,
+      "step": 1630
+    },
+    {
+      "epoch": 3.262,
+      "grad_norm": 0.9034122824668884,
+      "learning_rate": 2e-05,
+      "loss": 0.02218392,
+      "step": 1631
+    },
+    {
+      "epoch": 3.2640000000000002,
+      "grad_norm": 1.5698509216308594,
+      "learning_rate": 2e-05,
+      "loss": 0.05552916,
+      "step": 1632
+    },
+    {
+      "epoch": 3.266,
+      "grad_norm": 1.4005275964736938,
+      "learning_rate": 2e-05,
+      "loss": 0.03984913,
+      "step": 1633
+    },
+    {
+      "epoch": 3.268,
+      "grad_norm": 1.3338640928268433,
+      "learning_rate": 2e-05,
+      "loss": 0.03644662,
+      "step": 1634
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 1.7735025882720947,
+      "learning_rate": 2e-05,
+      "loss": 0.03694466,
+      "step": 1635
+    },
+    {
+      "epoch": 3.2720000000000002,
+      "grad_norm": 2.160663604736328,
+      "learning_rate": 2e-05,
+      "loss": 0.05129207,
+      "step": 1636
+    },
+    {
+      "epoch": 3.274,
+      "grad_norm": 2.0547683238983154,
+      "learning_rate": 2e-05,
+      "loss": 0.03293342,
+      "step": 1637
+    },
+    {
+      "epoch": 3.276,
+      "grad_norm": 1.309456467628479,
+      "learning_rate": 2e-05,
+      "loss": 0.04493797,
+      "step": 1638
+    },
+    {
+      "epoch": 3.278,
+      "grad_norm": 2.039910316467285,
+      "learning_rate": 2e-05,
+      "loss": 0.06677075,
+      "step": 1639
+    },
+    {
+      "epoch": 3.2800000000000002,
+      "grad_norm": 2.196495771408081,
+      "learning_rate": 2e-05,
+      "loss": 0.05029328,
+      "step": 1640
+    },
+    {
+      "epoch": 3.282,
+      "grad_norm": 2.049243927001953,
+      "learning_rate": 2e-05,
+      "loss": 0.05796874,
+      "step": 1641
+    },
+    {
+      "epoch": 3.284,
+      "grad_norm": 2.323168992996216,
+      "learning_rate": 2e-05,
+      "loss": 0.04754086,
+      "step": 1642
+    },
+    {
+      "epoch": 3.286,
+      "grad_norm": 1.9049650430679321,
+      "learning_rate": 2e-05,
+      "loss": 0.05514568,
+      "step": 1643
+    },
+    {
+      "epoch": 3.288,
+      "grad_norm": 1.898848295211792,
+      "learning_rate": 2e-05,
+      "loss": 0.03569252,
+      "step": 1644
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 2.345487356185913,
+      "learning_rate": 2e-05,
+      "loss": 0.0490138,
+      "step": 1645
+    },
+    {
+      "epoch": 3.292,
+      "grad_norm": 2.001127004623413,
+      "learning_rate": 2e-05,
+      "loss": 0.05215742,
+      "step": 1646
+    },
+    {
+      "epoch": 3.294,
+      "grad_norm": 1.4057964086532593,
+      "learning_rate": 2e-05,
+      "loss": 0.03418482,
+      "step": 1647
+    },
+    {
+      "epoch": 3.296,
+      "grad_norm": 1.412276029586792,
+      "learning_rate": 2e-05,
+      "loss": 0.03291162,
+      "step": 1648
+    },
+    {
+      "epoch": 3.298,
+      "grad_norm": 1.5926165580749512,
+      "learning_rate": 2e-05,
+      "loss": 0.04363141,
+      "step": 1649
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 1.9674831628799438,
+      "learning_rate": 2e-05,
+      "loss": 0.06002123,
+      "step": 1650
+    },
+    {
+      "epoch": 3.302,
+      "grad_norm": 1.581549048423767,
+      "learning_rate": 2e-05,
+      "loss": 0.07400438,
+      "step": 1651
+    },
+    {
+      "epoch": 3.304,
+      "grad_norm": 1.1746406555175781,
+      "learning_rate": 2e-05,
+      "loss": 0.03233917,
+      "step": 1652
+    },
+    {
+      "epoch": 3.306,
+      "grad_norm": 1.3594640493392944,
+      "learning_rate": 2e-05,
+      "loss": 0.06445151,
+      "step": 1653
+    },
+    {
+      "epoch": 3.308,
+      "grad_norm": 1.3005573749542236,
+      "learning_rate": 2e-05,
+      "loss": 0.03622051,
+      "step": 1654
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.8649536371231079,
+      "learning_rate": 2e-05,
+      "loss": 0.03161732,
+      "step": 1655
+    },
+    {
+      "epoch": 3.312,
+      "grad_norm": 1.1120051145553589,
+      "learning_rate": 2e-05,
+      "loss": 0.03670903,
+      "step": 1656
+    },
+    {
+      "epoch": 3.314,
+      "grad_norm": 2.0635063648223877,
+      "learning_rate": 2e-05,
+      "loss": 0.03296005,
+      "step": 1657
+    },
+    {
+      "epoch": 3.316,
+      "grad_norm": 1.942264437675476,
+      "learning_rate": 2e-05,
+      "loss": 0.04086999,
+      "step": 1658
+    },
+    {
+      "epoch": 3.318,
+      "grad_norm": 1.2889881134033203,
+      "learning_rate": 2e-05,
+      "loss": 0.03653413,
+      "step": 1659
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.5247739553451538,
+      "learning_rate": 2e-05,
+      "loss": 0.05022556,
+      "step": 1660
+    },
+    {
+      "epoch": 3.322,
+      "grad_norm": 1.208240270614624,
+      "learning_rate": 2e-05,
+      "loss": 0.04346113,
+      "step": 1661
+    },
+    {
+      "epoch": 3.324,
+      "grad_norm": 1.5146716833114624,
+      "learning_rate": 2e-05,
+      "loss": 0.04634988,
+      "step": 1662
+    },
+    {
+      "epoch": 3.326,
+      "grad_norm": 1.9538823366165161,
+      "learning_rate": 2e-05,
+      "loss": 0.04254812,
+      "step": 1663
+    },
+    {
+      "epoch": 3.328,
+      "grad_norm": 2.021803379058838,
+      "learning_rate": 2e-05,
+      "loss": 0.05505613,
+      "step": 1664
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 1.4671275615692139,
+      "learning_rate": 2e-05,
+      "loss": 0.04185315,
+      "step": 1665
+    },
+    {
+      "epoch": 3.332,
+      "grad_norm": 1.3730452060699463,
+      "learning_rate": 2e-05,
+      "loss": 0.04696137,
+      "step": 1666
+    },
+    {
+      "epoch": 3.334,
+      "grad_norm": 1.9206724166870117,
+      "learning_rate": 2e-05,
+      "loss": 0.05052917,
+      "step": 1667
+    },
+    {
+      "epoch": 3.336,
+      "grad_norm": 1.4539846181869507,
+      "learning_rate": 2e-05,
+      "loss": 0.04785593,
+      "step": 1668
+    },
+    {
+      "epoch": 3.338,
+      "grad_norm": 1.1836252212524414,
+      "learning_rate": 2e-05,
+      "loss": 0.03441792,
+      "step": 1669
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 1.3856287002563477,
+      "learning_rate": 2e-05,
+      "loss": 0.03965211,
+      "step": 1670
+    },
+    {
+      "epoch": 3.342,
+      "grad_norm": 1.7527414560317993,
+      "learning_rate": 2e-05,
+      "loss": 0.05044956,
+      "step": 1671
+    },
+    {
+      "epoch": 3.344,
+      "grad_norm": 1.460963249206543,
+      "learning_rate": 2e-05,
+      "loss": 0.0464677,
+      "step": 1672
+    },
+    {
+      "epoch": 3.346,
+      "grad_norm": 2.3444771766662598,
+      "learning_rate": 2e-05,
+      "loss": 0.05217274,
+      "step": 1673
+    },
+    {
+      "epoch": 3.348,
+      "grad_norm": 1.5559501647949219,
+      "learning_rate": 2e-05,
+      "loss": 0.04852623,
+      "step": 1674
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 1.9685094356536865,
+      "learning_rate": 2e-05,
+      "loss": 0.07367245,
+      "step": 1675
+    },
+    {
+      "epoch": 3.352,
+      "grad_norm": 1.3340469598770142,
+      "learning_rate": 2e-05,
+      "loss": 0.03031318,
+      "step": 1676
+    },
+    {
+      "epoch": 3.354,
+      "grad_norm": 1.5422812700271606,
+      "learning_rate": 2e-05,
+      "loss": 0.04720259,
+      "step": 1677
+    },
+    {
+      "epoch": 3.356,
+      "grad_norm": 1.6444870233535767,
+      "learning_rate": 2e-05,
+      "loss": 0.05606821,
+      "step": 1678
+    },
+    {
+      "epoch": 3.358,
+      "grad_norm": 2.2023096084594727,
+      "learning_rate": 2e-05,
+      "loss": 0.06454101,
+      "step": 1679
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 1.5785565376281738,
+      "learning_rate": 2e-05,
+      "loss": 0.04989371,
+      "step": 1680
+    },
+    {
+      "epoch": 3.362,
+      "grad_norm": 1.6977547407150269,
+      "learning_rate": 2e-05,
+      "loss": 0.06437075,
+      "step": 1681
+    },
+    {
+      "epoch": 3.364,
+      "grad_norm": 2.207683801651001,
+      "learning_rate": 2e-05,
+      "loss": 0.05562439,
+      "step": 1682
+    },
+    {
+      "epoch": 3.366,
+      "grad_norm": 1.5128238201141357,
+      "learning_rate": 2e-05,
+      "loss": 0.05253971,
+      "step": 1683
+    },
+    {
+      "epoch": 3.368,
+      "grad_norm": 1.5503803491592407,
+      "learning_rate": 2e-05,
+      "loss": 0.03294653,
+      "step": 1684
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 2.744611978530884,
+      "learning_rate": 2e-05,
+      "loss": 0.0665991,
+      "step": 1685
+    },
+    {
+      "epoch": 3.372,
+      "grad_norm": 1.9426320791244507,
+      "learning_rate": 2e-05,
+      "loss": 0.03777114,
+      "step": 1686
+    },
+    {
+      "epoch": 3.374,
+      "grad_norm": 1.7858504056930542,
+      "learning_rate": 2e-05,
+      "loss": 0.06296871,
+      "step": 1687
+    },
+    {
+      "epoch": 3.376,
+      "grad_norm": 1.6751807928085327,
+      "learning_rate": 2e-05,
+      "loss": 0.0362752,
+      "step": 1688
+    },
+    {
+      "epoch": 3.378,
+      "grad_norm": 1.8033543825149536,
+      "learning_rate": 2e-05,
+      "loss": 0.05158364,
+      "step": 1689
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 1.0224000215530396,
+      "learning_rate": 2e-05,
+      "loss": 0.03424178,
+      "step": 1690
+    },
+    {
+      "epoch": 3.382,
+      "grad_norm": 1.2834762334823608,
+      "learning_rate": 2e-05,
+      "loss": 0.04011747,
+      "step": 1691
+    },
+    {
+      "epoch": 3.384,
+      "grad_norm": 1.573851466178894,
+      "learning_rate": 2e-05,
+      "loss": 0.04472176,
+      "step": 1692
+    },
+    {
+      "epoch": 3.386,
+      "grad_norm": 1.0216726064682007,
+      "learning_rate": 2e-05,
+      "loss": 0.03067708,
+      "step": 1693
+    },
+    {
+      "epoch": 3.388,
+      "grad_norm": 1.6406019926071167,
+      "learning_rate": 2e-05,
+      "loss": 0.06715198,
+      "step": 1694
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 1.9565666913986206,
+      "learning_rate": 2e-05,
+      "loss": 0.04441306,
+      "step": 1695
+    },
+    {
+      "epoch": 3.392,
+      "grad_norm": 1.312374234199524,
+      "learning_rate": 2e-05,
+      "loss": 0.04021595,
+      "step": 1696
+    },
+    {
+      "epoch": 3.394,
+      "grad_norm": 1.5732117891311646,
+      "learning_rate": 2e-05,
+      "loss": 0.05825553,
+      "step": 1697
+    },
+    {
+      "epoch": 3.396,
+      "grad_norm": 1.7435842752456665,
+      "learning_rate": 2e-05,
+      "loss": 0.03737545,
+      "step": 1698
+    },
+    {
+      "epoch": 3.398,
+      "grad_norm": 1.3936766386032104,
+      "learning_rate": 2e-05,
+      "loss": 0.04649488,
+      "step": 1699
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.222420573234558,
+      "learning_rate": 2e-05,
+      "loss": 0.04838939,
+      "step": 1700
+    },
+    {
+      "epoch": 3.402,
+      "grad_norm": 1.7847157716751099,
+      "learning_rate": 2e-05,
+      "loss": 0.04798005,
+      "step": 1701
+    },
+    {
+      "epoch": 3.404,
+      "grad_norm": 1.8903214931488037,
+      "learning_rate": 2e-05,
+      "loss": 0.04042132,
+      "step": 1702
+    },
+    {
+      "epoch": 3.406,
+      "grad_norm": 1.9824771881103516,
+      "learning_rate": 2e-05,
+      "loss": 0.04405872,
+      "step": 1703
+    },
+    {
+      "epoch": 3.408,
+      "grad_norm": 2.978909730911255,
+      "learning_rate": 2e-05,
+      "loss": 0.06235117,
+      "step": 1704
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 2.1281321048736572,
+      "learning_rate": 2e-05,
+      "loss": 0.0631174,
+      "step": 1705
+    },
+    {
+      "epoch": 3.412,
+      "grad_norm": 1.207214117050171,
+      "learning_rate": 2e-05,
+      "loss": 0.04360072,
+      "step": 1706
+    },
+    {
+      "epoch": 3.414,
+      "grad_norm": 2.4614744186401367,
+      "learning_rate": 2e-05,
+      "loss": 0.05545459,
+      "step": 1707
+    },
+    {
+      "epoch": 3.416,
+      "grad_norm": 2.2295455932617188,
+      "learning_rate": 2e-05,
+      "loss": 0.05138841,
+      "step": 1708
+    },
+    {
+      "epoch": 3.418,
+      "grad_norm": 1.6426289081573486,
+      "learning_rate": 2e-05,
+      "loss": 0.03976878,
+      "step": 1709
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 2.6790168285369873,
+      "learning_rate": 2e-05,
+      "loss": 0.06980205,
+      "step": 1710
+    },
+    {
+      "epoch": 3.422,
+      "grad_norm": 1.8094228506088257,
+      "learning_rate": 2e-05,
+      "loss": 0.05368996,
+      "step": 1711
+    },
+    {
+      "epoch": 3.424,
+      "grad_norm": 2.0319266319274902,
+      "learning_rate": 2e-05,
+      "loss": 0.0543413,
+      "step": 1712
+    },
+    {
+      "epoch": 3.426,
+      "grad_norm": 1.3606593608856201,
+      "learning_rate": 2e-05,
+      "loss": 0.03617115,
+      "step": 1713
+    },
+    {
+      "epoch": 3.428,
+      "grad_norm": 4.077547550201416,
+      "learning_rate": 2e-05,
+      "loss": 0.0491549,
+      "step": 1714
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 2.273555040359497,
+      "learning_rate": 2e-05,
+      "loss": 0.06512973,
+      "step": 1715
+    },
+    {
+      "epoch": 3.432,
+      "grad_norm": 1.533727765083313,
+      "learning_rate": 2e-05,
+      "loss": 0.05353583,
+      "step": 1716
+    },
+    {
+      "epoch": 3.434,
+      "grad_norm": 1.5889472961425781,
+      "learning_rate": 2e-05,
+      "loss": 0.05391718,
+      "step": 1717
+    },
+    {
+      "epoch": 3.436,
+      "grad_norm": 1.513043761253357,
+      "learning_rate": 2e-05,
+      "loss": 0.05352363,
+      "step": 1718
+    },
+    {
+      "epoch": 3.438,
+      "grad_norm": 1.708634853363037,
+      "learning_rate": 2e-05,
+      "loss": 0.042217,
+      "step": 1719
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 1.5580065250396729,
+      "learning_rate": 2e-05,
+      "loss": 0.05138306,
+      "step": 1720
+    },
+    {
+      "epoch": 3.442,
+      "grad_norm": 1.3362562656402588,
+      "learning_rate": 2e-05,
+      "loss": 0.04820231,
+      "step": 1721
+    },
+    {
+      "epoch": 3.444,
+      "grad_norm": 1.8119944334030151,
+      "learning_rate": 2e-05,
+      "loss": 0.03760799,
+      "step": 1722
+    },
+    {
+      "epoch": 3.446,
+      "grad_norm": 1.7421449422836304,
+      "learning_rate": 2e-05,
+      "loss": 0.04001708,
+      "step": 1723
+    },
+    {
+      "epoch": 3.448,
+      "grad_norm": 1.6809686422348022,
+      "learning_rate": 2e-05,
+      "loss": 0.05122007,
+      "step": 1724
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 3.9546453952789307,
+      "learning_rate": 2e-05,
+      "loss": 0.04566801,
+      "step": 1725
+    },
+    {
+      "epoch": 3.452,
+      "grad_norm": 1.462601661682129,
+      "learning_rate": 2e-05,
+      "loss": 0.04439667,
+      "step": 1726
+    },
+    {
+      "epoch": 3.454,
+      "grad_norm": 2.2390847206115723,
+      "learning_rate": 2e-05,
+      "loss": 0.03910905,
+      "step": 1727
+    },
+    {
+      "epoch": 3.456,
+      "grad_norm": 1.7394695281982422,
+      "learning_rate": 2e-05,
+      "loss": 0.05194352,
+      "step": 1728
+    },
+    {
+      "epoch": 3.458,
+      "grad_norm": 2.4958033561706543,
+      "learning_rate": 2e-05,
+      "loss": 0.05337561,
+      "step": 1729
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 2.3025898933410645,
+      "learning_rate": 2e-05,
+      "loss": 0.04093119,
+      "step": 1730
+    },
+    {
+      "epoch": 3.462,
+      "grad_norm": 1.6952155828475952,
+      "learning_rate": 2e-05,
+      "loss": 0.0470874,
+      "step": 1731
+    },
+    {
+      "epoch": 3.464,
+      "grad_norm": 1.3717252016067505,
+      "learning_rate": 2e-05,
+      "loss": 0.02857566,
+      "step": 1732
+    },
+    {
+      "epoch": 3.466,
+      "grad_norm": 2.161407232284546,
+      "learning_rate": 2e-05,
+      "loss": 0.06686375,
+      "step": 1733
+    },
+    {
+      "epoch": 3.468,
+      "grad_norm": 1.595219373703003,
+      "learning_rate": 2e-05,
+      "loss": 0.05054267,
+      "step": 1734
+    },
+    {
+      "epoch": 3.4699999999999998,
+      "grad_norm": 1.613440752029419,
+      "learning_rate": 2e-05,
+      "loss": 0.02639636,
+      "step": 1735
+    },
+    {
+      "epoch": 3.472,
+      "grad_norm": 2.263852596282959,
+      "learning_rate": 2e-05,
+      "loss": 0.05445278,
+      "step": 1736
+    },
+    {
+      "epoch": 3.474,
+      "grad_norm": 2.566851854324341,
+      "learning_rate": 2e-05,
+      "loss": 0.05587868,
+      "step": 1737
+    },
+    {
+      "epoch": 3.476,
+      "grad_norm": 2.8093175888061523,
+      "learning_rate": 2e-05,
+      "loss": 0.05327387,
+      "step": 1738
+    },
+    {
+      "epoch": 3.4779999999999998,
+      "grad_norm": 1.6908395290374756,
+      "learning_rate": 2e-05,
+      "loss": 0.04662281,
+      "step": 1739
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 3.4398069381713867,
+      "learning_rate": 2e-05,
+      "loss": 0.06926405,
+      "step": 1740
+    },
+    {
+      "epoch": 3.482,
+      "grad_norm": 1.7976182699203491,
+      "learning_rate": 2e-05,
+      "loss": 0.05067788,
+      "step": 1741
+    },
+    {
+      "epoch": 3.484,
+      "grad_norm": 1.8646904230117798,
+      "learning_rate": 2e-05,
+      "loss": 0.04594427,
+      "step": 1742
+    },
+    {
+      "epoch": 3.4859999999999998,
+      "grad_norm": 1.8577203750610352,
+      "learning_rate": 2e-05,
+      "loss": 0.05025358,
+      "step": 1743
+    },
+    {
+      "epoch": 3.488,
+      "grad_norm": 1.3906282186508179,
+      "learning_rate": 2e-05,
+      "loss": 0.04262529,
+      "step": 1744
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 1.8370797634124756,
+      "learning_rate": 2e-05,
+      "loss": 0.05269311,
+      "step": 1745
+    },
+    {
+      "epoch": 3.492,
+      "grad_norm": 2.2525057792663574,
+      "learning_rate": 2e-05,
+      "loss": 0.04886598,
+      "step": 1746
+    },
+    {
+      "epoch": 3.4939999999999998,
+      "grad_norm": 1.8896589279174805,
+      "learning_rate": 2e-05,
+      "loss": 0.05545288,
+      "step": 1747
+    },
+    {
+      "epoch": 3.496,
+      "grad_norm": 1.3138065338134766,
+      "learning_rate": 2e-05,
+      "loss": 0.04557213,
+      "step": 1748
+    },
+    {
+      "epoch": 3.498,
+      "grad_norm": 2.4323270320892334,
+      "learning_rate": 2e-05,
+      "loss": 0.05506615,
+      "step": 1749
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 1.2671258449554443,
+      "learning_rate": 2e-05,
+      "loss": 0.03962918,
+      "step": 1750
+    },
+    {
+      "epoch": 3.502,
+      "grad_norm": 1.2083110809326172,
+      "learning_rate": 2e-05,
+      "loss": 0.04052875,
+      "step": 1751
+    },
+    {
+      "epoch": 3.504,
+      "grad_norm": 2.204509735107422,
+      "learning_rate": 2e-05,
+      "loss": 0.05543607,
+      "step": 1752
+    },
+    {
+      "epoch": 3.5060000000000002,
+      "grad_norm": 2.0916733741760254,
+      "learning_rate": 2e-05,
+      "loss": 0.05030297,
+      "step": 1753
+    },
+    {
+      "epoch": 3.508,
+      "grad_norm": 1.2749862670898438,
+      "learning_rate": 2e-05,
+      "loss": 0.05265682,
+      "step": 1754
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.2697100639343262,
+      "learning_rate": 2e-05,
+      "loss": 0.04578009,
+      "step": 1755
+    },
+    {
+      "epoch": 3.512,
+      "grad_norm": 2.575673818588257,
+      "learning_rate": 2e-05,
+      "loss": 0.05128983,
+      "step": 1756
+    },
+    {
+      "epoch": 3.5140000000000002,
+      "grad_norm": 1.548751711845398,
+      "learning_rate": 2e-05,
+      "loss": 0.04093613,
+      "step": 1757
+    },
+    {
+      "epoch": 3.516,
+      "grad_norm": 1.305117130279541,
+      "learning_rate": 2e-05,
+      "loss": 0.04004899,
+      "step": 1758
+    },
+    {
+      "epoch": 3.518,
+      "grad_norm": 1.4659463167190552,
+      "learning_rate": 2e-05,
+      "loss": 0.04068562,
+      "step": 1759
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 2.3224215507507324,
+      "learning_rate": 2e-05,
+      "loss": 0.03831296,
+      "step": 1760
+    },
+    {
+      "epoch": 3.5220000000000002,
+      "grad_norm": 1.6149941682815552,
+      "learning_rate": 2e-05,
+      "loss": 0.0297474,
+      "step": 1761
+    },
+    {
+      "epoch": 3.524,
+      "grad_norm": 1.7875992059707642,
+      "learning_rate": 2e-05,
+      "loss": 0.04237871,
+      "step": 1762
+    },
+    {
+      "epoch": 3.526,
+      "grad_norm": 1.6973800659179688,
+      "learning_rate": 2e-05,
+      "loss": 0.0459117,
+      "step": 1763
+    },
+    {
+      "epoch": 3.528,
+      "grad_norm": 2.5080761909484863,
+      "learning_rate": 2e-05,
+      "loss": 0.05171615,
+      "step": 1764
+    },
+    {
+      "epoch": 3.5300000000000002,
+      "grad_norm": 1.5514436960220337,
+      "learning_rate": 2e-05,
+      "loss": 0.03801546,
+      "step": 1765
+    },
+    {
+      "epoch": 3.532,
+      "grad_norm": 1.5856919288635254,
+      "learning_rate": 2e-05,
+      "loss": 0.03361356,
+      "step": 1766
+    },
+    {
+      "epoch": 3.534,
+      "grad_norm": 1.5322203636169434,
+      "learning_rate": 2e-05,
+      "loss": 0.04025748,
+      "step": 1767
+    },
+    {
+      "epoch": 3.536,
+      "grad_norm": 2.3193933963775635,
+      "learning_rate": 2e-05,
+      "loss": 0.06852245,
+      "step": 1768
+    },
+    {
+      "epoch": 3.5380000000000003,
+      "grad_norm": 2.101360559463501,
+      "learning_rate": 2e-05,
+      "loss": 0.06704597,
+      "step": 1769
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 2.1348562240600586,
+      "learning_rate": 2e-05,
+      "loss": 0.03175708,
+      "step": 1770
+    },
+    {
+      "epoch": 3.542,
+      "grad_norm": 1.613877296447754,
+      "learning_rate": 2e-05,
+      "loss": 0.05038162,
+      "step": 1771
+    },
+    {
+      "epoch": 3.544,
+      "grad_norm": 1.5137550830841064,
+      "learning_rate": 2e-05,
+      "loss": 0.04188209,
+      "step": 1772
+    },
+    {
+      "epoch": 3.5460000000000003,
+      "grad_norm": 1.4603201150894165,
+      "learning_rate": 2e-05,
+      "loss": 0.03608978,
+      "step": 1773
+    },
+    {
+      "epoch": 3.548,
+      "grad_norm": 1.6376720666885376,
+      "learning_rate": 2e-05,
+      "loss": 0.05980166,
+      "step": 1774
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 2.329392671585083,
+      "learning_rate": 2e-05,
+      "loss": 0.04924781,
+      "step": 1775
+    },
+    {
+      "epoch": 3.552,
+      "grad_norm": 2.361161470413208,
+      "learning_rate": 2e-05,
+      "loss": 0.05286585,
+      "step": 1776
+    },
+    {
+      "epoch": 3.5540000000000003,
+      "grad_norm": 1.530290126800537,
+      "learning_rate": 2e-05,
+      "loss": 0.03231184,
+      "step": 1777
+    },
+    {
+      "epoch": 3.556,
+      "grad_norm": 1.1076228618621826,
+      "learning_rate": 2e-05,
+      "loss": 0.02891509,
+      "step": 1778
+    },
+    {
+      "epoch": 3.558,
+      "grad_norm": 1.950429916381836,
+      "learning_rate": 2e-05,
+      "loss": 0.06025671,
+      "step": 1779
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 1.4713891744613647,
+      "learning_rate": 2e-05,
+      "loss": 0.04263363,
+      "step": 1780
+    },
+    {
+      "epoch": 3.5620000000000003,
+      "grad_norm": 1.9598925113677979,
+      "learning_rate": 2e-05,
+      "loss": 0.0472781,
+      "step": 1781
+    },
+    {
+      "epoch": 3.564,
+      "grad_norm": 1.9756190776824951,
+      "learning_rate": 2e-05,
+      "loss": 0.05379218,
+      "step": 1782
+    },
+    {
+      "epoch": 3.566,
+      "grad_norm": 1.9871381521224976,
+      "learning_rate": 2e-05,
+      "loss": 0.04629632,
+      "step": 1783
+    },
+    {
+      "epoch": 3.568,
+      "grad_norm": 2.374814510345459,
+      "learning_rate": 2e-05,
+      "loss": 0.05010781,
+      "step": 1784
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 1.6160145998001099,
+      "learning_rate": 2e-05,
+      "loss": 0.04810391,
+      "step": 1785
+    },
+    {
+      "epoch": 3.572,
+      "grad_norm": 1.333565354347229,
+      "learning_rate": 2e-05,
+      "loss": 0.04096656,
+      "step": 1786
+    },
+    {
+      "epoch": 3.574,
+      "grad_norm": 1.2630254030227661,
+      "learning_rate": 2e-05,
+      "loss": 0.04580157,
+      "step": 1787
+    },
+    {
+      "epoch": 3.576,
+      "grad_norm": 2.068432569503784,
+      "learning_rate": 2e-05,
+      "loss": 0.06533036,
+      "step": 1788
+    },
+    {
+      "epoch": 3.578,
+      "grad_norm": 1.422972321510315,
+      "learning_rate": 2e-05,
+      "loss": 0.04293206,
+      "step": 1789
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 1.5883013010025024,
+      "learning_rate": 2e-05,
+      "loss": 0.03554819,
+      "step": 1790
+    },
+    {
+      "epoch": 3.582,
+      "grad_norm": 1.2352162599563599,
+      "learning_rate": 2e-05,
+      "loss": 0.03722744,
+      "step": 1791
+    },
+    {
+      "epoch": 3.584,
+      "grad_norm": 1.440484642982483,
+      "learning_rate": 2e-05,
+      "loss": 0.03654459,
+      "step": 1792
+    },
+    {
+      "epoch": 3.586,
+      "grad_norm": 2.2191922664642334,
+      "learning_rate": 2e-05,
+      "loss": 0.05253022,
+      "step": 1793
+    },
+    {
+      "epoch": 3.588,
+      "grad_norm": 1.7797856330871582,
+      "learning_rate": 2e-05,
+      "loss": 0.04316117,
+      "step": 1794
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 1.476678729057312,
+      "learning_rate": 2e-05,
+      "loss": 0.04034651,
+      "step": 1795
+    },
+    {
+      "epoch": 3.592,
+      "grad_norm": 1.2704788446426392,
+      "learning_rate": 2e-05,
+      "loss": 0.03967745,
+      "step": 1796
+    },
+    {
+      "epoch": 3.594,
+      "grad_norm": 2.31673002243042,
+      "learning_rate": 2e-05,
+      "loss": 0.0395114,
+      "step": 1797
+    },
+    {
+      "epoch": 3.596,
+      "grad_norm": 1.2823116779327393,
+      "learning_rate": 2e-05,
+      "loss": 0.04466548,
+      "step": 1798
+    },
+    {
+      "epoch": 3.598,
+      "grad_norm": 1.1849486827850342,
+      "learning_rate": 2e-05,
+      "loss": 0.04506912,
+      "step": 1799
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.558552622795105,
+      "learning_rate": 2e-05,
+      "loss": 0.05311072,
+      "step": 1800
+    },
+    {
+      "epoch": 3.602,
+      "grad_norm": 1.5442121028900146,
+      "learning_rate": 2e-05,
+      "loss": 0.0526191,
+      "step": 1801
+    },
+    {
+      "epoch": 3.604,
+      "grad_norm": 1.8791892528533936,
+      "learning_rate": 2e-05,
+      "loss": 0.06307591,
+      "step": 1802
+    },
+    {
+      "epoch": 3.606,
+      "grad_norm": 1.6081328392028809,
+      "learning_rate": 2e-05,
+      "loss": 0.03651448,
+      "step": 1803
+    },
+    {
+      "epoch": 3.608,
+      "grad_norm": 1.4104983806610107,
+      "learning_rate": 2e-05,
+      "loss": 0.04336455,
+      "step": 1804
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 2.381561756134033,
+      "learning_rate": 2e-05,
+      "loss": 0.05898841,
+      "step": 1805
+    },
+    {
+      "epoch": 3.612,
+      "grad_norm": 1.3324648141860962,
+      "learning_rate": 2e-05,
+      "loss": 0.03454057,
+      "step": 1806
+    },
+    {
+      "epoch": 3.614,
+      "grad_norm": 1.5297226905822754,
+      "learning_rate": 2e-05,
+      "loss": 0.05038363,
+      "step": 1807
+    },
+    {
+      "epoch": 3.616,
+      "grad_norm": 1.5506995916366577,
+      "learning_rate": 2e-05,
+      "loss": 0.05092872,
+      "step": 1808
+    },
+    {
+      "epoch": 3.618,
+      "grad_norm": 1.181553602218628,
+      "learning_rate": 2e-05,
+      "loss": 0.03173397,
+      "step": 1809
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 1.6461181640625,
+      "learning_rate": 2e-05,
+      "loss": 0.05175463,
+      "step": 1810
+    },
+    {
+      "epoch": 3.622,
+      "grad_norm": 1.9184272289276123,
+      "learning_rate": 2e-05,
+      "loss": 0.06124686,
+      "step": 1811
+    },
+    {
+      "epoch": 3.624,
+      "grad_norm": 1.6065723896026611,
+      "learning_rate": 2e-05,
+      "loss": 0.04484671,
+      "step": 1812
+    },
+    {
+      "epoch": 3.626,
+      "grad_norm": 1.443337082862854,
+      "learning_rate": 2e-05,
+      "loss": 0.04267926,
+      "step": 1813
+    },
+    {
+      "epoch": 3.628,
+      "grad_norm": 1.178184151649475,
+      "learning_rate": 2e-05,
+      "loss": 0.02920187,
+      "step": 1814
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 1.146735429763794,
+      "learning_rate": 2e-05,
+      "loss": 0.02786646,
+      "step": 1815
+    },
+    {
+      "epoch": 3.632,
+      "grad_norm": 1.1989175081253052,
+      "learning_rate": 2e-05,
+      "loss": 0.03249002,
+      "step": 1816
+    },
+    {
+      "epoch": 3.634,
+      "grad_norm": 1.540666103363037,
+      "learning_rate": 2e-05,
+      "loss": 0.03965627,
+      "step": 1817
+    },
+    {
+      "epoch": 3.636,
+      "grad_norm": 1.4439377784729004,
+      "learning_rate": 2e-05,
+      "loss": 0.04156953,
+      "step": 1818
+    },
+    {
+      "epoch": 3.638,
+      "grad_norm": 1.2657922506332397,
+      "learning_rate": 2e-05,
+      "loss": 0.03509516,
+      "step": 1819
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 2.9393463134765625,
+      "learning_rate": 2e-05,
+      "loss": 0.04102836,
+      "step": 1820
+    },
+    {
+      "epoch": 3.642,
+      "grad_norm": 1.263395071029663,
+      "learning_rate": 2e-05,
+      "loss": 0.03908949,
+      "step": 1821
+    },
+    {
+      "epoch": 3.644,
+      "grad_norm": 1.1321371793746948,
+      "learning_rate": 2e-05,
+      "loss": 0.03896394,
+      "step": 1822
+    },
+    {
+      "epoch": 3.646,
+      "grad_norm": 1.1515556573867798,
+      "learning_rate": 2e-05,
+      "loss": 0.0317642,
+      "step": 1823
+    },
+    {
+      "epoch": 3.648,
+      "grad_norm": 1.4203664064407349,
+      "learning_rate": 2e-05,
+      "loss": 0.05269869,
+      "step": 1824
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.6567198038101196,
+      "learning_rate": 2e-05,
+      "loss": 0.03793368,
+      "step": 1825
+    },
+    {
+      "epoch": 3.652,
+      "grad_norm": 2.2994821071624756,
+      "learning_rate": 2e-05,
+      "loss": 0.0429872,
+      "step": 1826
+    },
+    {
+      "epoch": 3.654,
+      "grad_norm": 1.6607061624526978,
+      "learning_rate": 2e-05,
+      "loss": 0.03379823,
+      "step": 1827
+    },
+    {
+      "epoch": 3.656,
+      "grad_norm": 1.7107027769088745,
+      "learning_rate": 2e-05,
+      "loss": 0.04675277,
+      "step": 1828
+    },
+    {
+      "epoch": 3.658,
+      "grad_norm": 1.8303852081298828,
+      "learning_rate": 2e-05,
+      "loss": 0.04841805,
+      "step": 1829
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 1.8825572729110718,
+      "learning_rate": 2e-05,
+      "loss": 0.04874373,
+      "step": 1830
+    },
+    {
+      "epoch": 3.662,
+      "grad_norm": 1.7029788494110107,
+      "learning_rate": 2e-05,
+      "loss": 0.05050893,
+      "step": 1831
+    },
+    {
+      "epoch": 3.664,
+      "grad_norm": 1.1050704717636108,
+      "learning_rate": 2e-05,
+      "loss": 0.03281447,
+      "step": 1832
+    },
+    {
+      "epoch": 3.666,
+      "grad_norm": 1.4161999225616455,
+      "learning_rate": 2e-05,
+      "loss": 0.0544338,
+      "step": 1833
+    },
+    {
+      "epoch": 3.668,
+      "grad_norm": 1.964059829711914,
+      "learning_rate": 2e-05,
+      "loss": 0.03872099,
+      "step": 1834
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 2.016484022140503,
+      "learning_rate": 2e-05,
+      "loss": 0.05875476,
+      "step": 1835
+    },
+    {
+      "epoch": 3.672,
+      "grad_norm": 1.473386287689209,
+      "learning_rate": 2e-05,
+      "loss": 0.03022585,
+      "step": 1836
+    },
+    {
+      "epoch": 3.674,
+      "grad_norm": 1.348415493965149,
+      "learning_rate": 2e-05,
+      "loss": 0.04690105,
+      "step": 1837
+    },
+    {
+      "epoch": 3.676,
+      "grad_norm": 0.9426001310348511,
+      "learning_rate": 2e-05,
+      "loss": 0.02384089,
+      "step": 1838
+    },
+    {
+      "epoch": 3.678,
+      "grad_norm": 3.1298086643218994,
+      "learning_rate": 2e-05,
+      "loss": 0.06468902,
+      "step": 1839
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 1.0675461292266846,
+      "learning_rate": 2e-05,
+      "loss": 0.03523882,
+      "step": 1840
+    },
+    {
+      "epoch": 3.682,
+      "grad_norm": 0.8646126985549927,
+      "learning_rate": 2e-05,
+      "loss": 0.01838133,
+      "step": 1841
+    },
+    {
+      "epoch": 3.684,
+      "grad_norm": 1.141176700592041,
+      "learning_rate": 2e-05,
+      "loss": 0.03865944,
+      "step": 1842
+    },
+    {
+      "epoch": 3.686,
+      "grad_norm": 2.4700798988342285,
+      "learning_rate": 2e-05,
+      "loss": 0.04170835,
+      "step": 1843
+    },
+    {
+      "epoch": 3.6879999999999997,
+      "grad_norm": 1.2326807975769043,
+      "learning_rate": 2e-05,
+      "loss": 0.03476126,
+      "step": 1844
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 1.714464545249939,
+      "learning_rate": 2e-05,
+      "loss": 0.03549024,
+      "step": 1845
+    },
+    {
+      "epoch": 3.692,
+      "grad_norm": 1.038811206817627,
+      "learning_rate": 2e-05,
+      "loss": 0.02916805,
+      "step": 1846
+    },
+    {
+      "epoch": 3.694,
+      "grad_norm": 2.0106890201568604,
+      "learning_rate": 2e-05,
+      "loss": 0.04282944,
+      "step": 1847
+    },
+    {
+      "epoch": 3.6959999999999997,
+      "grad_norm": 1.4869582653045654,
+      "learning_rate": 2e-05,
+      "loss": 0.0293393,
+      "step": 1848
+    },
+    {
+      "epoch": 3.698,
+      "grad_norm": 1.699388861656189,
+      "learning_rate": 2e-05,
+      "loss": 0.03086464,
+      "step": 1849
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.3598302602767944,
+      "learning_rate": 2e-05,
+      "loss": 0.03127223,
+      "step": 1850
+    },
+    {
+      "epoch": 3.702,
+      "grad_norm": 1.3136847019195557,
+      "learning_rate": 2e-05,
+      "loss": 0.03167934,
+      "step": 1851
+    },
+    {
+      "epoch": 3.7039999999999997,
+      "grad_norm": 1.9143792390823364,
+      "learning_rate": 2e-05,
+      "loss": 0.03092597,
+      "step": 1852
+    },
+    {
+      "epoch": 3.706,
+      "grad_norm": 1.9676464796066284,
+      "learning_rate": 2e-05,
+      "loss": 0.04451537,
+      "step": 1853
+    },
+    {
+      "epoch": 3.708,
+      "grad_norm": 0.9787660241127014,
+      "learning_rate": 2e-05,
+      "loss": 0.02394712,
+      "step": 1854
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 2.257038116455078,
+      "learning_rate": 2e-05,
+      "loss": 0.04062683,
+      "step": 1855
+    },
+    {
+      "epoch": 3.7119999999999997,
+      "grad_norm": 1.5607311725616455,
+      "learning_rate": 2e-05,
+      "loss": 0.02945242,
+      "step": 1856
+    },
+    {
+      "epoch": 3.714,
+      "grad_norm": 1.4506772756576538,
+      "learning_rate": 2e-05,
+      "loss": 0.04430741,
+      "step": 1857
+    },
+    {
+      "epoch": 3.716,
+      "grad_norm": 2.7266697883605957,
+      "learning_rate": 2e-05,
+      "loss": 0.03880604,
+      "step": 1858
+    },
+    {
+      "epoch": 3.718,
+      "grad_norm": 1.774509072303772,
+      "learning_rate": 2e-05,
+      "loss": 0.04546557,
+      "step": 1859
+    },
+    {
+      "epoch": 3.7199999999999998,
+      "grad_norm": 2.044501543045044,
+      "learning_rate": 2e-05,
+      "loss": 0.05731916,
+      "step": 1860
+    },
+    {
+      "epoch": 3.722,
+      "grad_norm": 1.1676855087280273,
+      "learning_rate": 2e-05,
+      "loss": 0.03595718,
+      "step": 1861
+    },
+    {
+      "epoch": 3.724,
+      "grad_norm": 1.1329604387283325,
+      "learning_rate": 2e-05,
+      "loss": 0.03459671,
+      "step": 1862
+    },
+    {
+      "epoch": 3.726,
+      "grad_norm": 1.5749149322509766,
+      "learning_rate": 2e-05,
+      "loss": 0.03534314,
+      "step": 1863
+    },
+    {
+      "epoch": 3.7279999999999998,
+      "grad_norm": 2.428170680999756,
+      "learning_rate": 2e-05,
+      "loss": 0.08179246,
+      "step": 1864
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 1.7956899404525757,
+      "learning_rate": 2e-05,
+      "loss": 0.06140782,
+      "step": 1865
+    },
+    {
+      "epoch": 3.732,
+      "grad_norm": 1.7499301433563232,
+      "learning_rate": 2e-05,
+      "loss": 0.04886729,
+      "step": 1866
+    },
+    {
+      "epoch": 3.734,
+      "grad_norm": 1.4074490070343018,
+      "learning_rate": 2e-05,
+      "loss": 0.04775254,
+      "step": 1867
+    },
+    {
+      "epoch": 3.7359999999999998,
+      "grad_norm": 1.8201322555541992,
+      "learning_rate": 2e-05,
+      "loss": 0.04804891,
+      "step": 1868
+    },
+    {
+      "epoch": 3.738,
+      "grad_norm": 1.9848655462265015,
+      "learning_rate": 2e-05,
+      "loss": 0.05379272,
+      "step": 1869
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 1.2629104852676392,
+      "learning_rate": 2e-05,
+      "loss": 0.03760821,
+      "step": 1870
+    },
+    {
+      "epoch": 3.742,
+      "grad_norm": 1.1420437097549438,
+      "learning_rate": 2e-05,
+      "loss": 0.02983224,
+      "step": 1871
+    },
+    {
+      "epoch": 3.7439999999999998,
+      "grad_norm": 1.4850512742996216,
+      "learning_rate": 2e-05,
+      "loss": 0.04077002,
+      "step": 1872
+    },
+    {
+      "epoch": 3.746,
+      "grad_norm": 1.3481850624084473,
+      "learning_rate": 2e-05,
+      "loss": 0.04353747,
+      "step": 1873
+    },
+    {
+      "epoch": 3.748,
+      "grad_norm": 1.613179326057434,
+      "learning_rate": 2e-05,
+      "loss": 0.04657446,
+      "step": 1874
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 1.436995029449463,
+      "learning_rate": 2e-05,
+      "loss": 0.04443494,
+      "step": 1875
+    },
+    {
+      "epoch": 3.752,
+      "grad_norm": 2.0734074115753174,
+      "learning_rate": 2e-05,
+      "loss": 0.05218915,
+      "step": 1876
+    },
+    {
+      "epoch": 3.754,
+      "grad_norm": 1.7502717971801758,
+      "learning_rate": 2e-05,
+      "loss": 0.0336446,
+      "step": 1877
+    },
+    {
+      "epoch": 3.7560000000000002,
+      "grad_norm": 1.8787667751312256,
+      "learning_rate": 2e-05,
+      "loss": 0.04104609,
+      "step": 1878
+    },
+    {
+      "epoch": 3.758,
+      "grad_norm": 2.121859550476074,
+      "learning_rate": 2e-05,
+      "loss": 0.04827151,
+      "step": 1879
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 2.4621684551239014,
+      "learning_rate": 2e-05,
+      "loss": 0.04723193,
+      "step": 1880
+    },
+    {
+      "epoch": 3.762,
+      "grad_norm": 2.063389778137207,
+      "learning_rate": 2e-05,
+      "loss": 0.04516068,
+      "step": 1881
+    },
+    {
+      "epoch": 3.7640000000000002,
+      "grad_norm": 1.6342523097991943,
+      "learning_rate": 2e-05,
+      "loss": 0.05464017,
+      "step": 1882
+    },
+    {
+      "epoch": 3.766,
+      "grad_norm": 1.649129867553711,
+      "learning_rate": 2e-05,
+      "loss": 0.06015805,
+      "step": 1883
+    },
+    {
+      "epoch": 3.768,
+      "grad_norm": 1.4318526983261108,
+      "learning_rate": 2e-05,
+      "loss": 0.030782,
+      "step": 1884
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 1.4619206190109253,
+      "learning_rate": 2e-05,
+      "loss": 0.04468732,
+      "step": 1885
+    },
+    {
+      "epoch": 3.7720000000000002,
+      "grad_norm": 2.3498082160949707,
+      "learning_rate": 2e-05,
+      "loss": 0.03591853,
+      "step": 1886
+    },
+    {
+      "epoch": 3.774,
+      "grad_norm": 2.933544397354126,
+      "learning_rate": 2e-05,
+      "loss": 0.04159092,
+      "step": 1887
+    },
+    {
+      "epoch": 3.776,
+      "grad_norm": 3.0242841243743896,
+      "learning_rate": 2e-05,
+      "loss": 0.07077288,
+      "step": 1888
+    },
+    {
+      "epoch": 3.778,
+      "grad_norm": 0.9896936416625977,
+      "learning_rate": 2e-05,
+      "loss": 0.0333928,
+      "step": 1889
+    },
+    {
+      "epoch": 3.7800000000000002,
+      "grad_norm": 2.85379958152771,
+      "learning_rate": 2e-05,
+      "loss": 0.04417305,
+      "step": 1890
+    },
+    {
+      "epoch": 3.782,
+      "grad_norm": 1.0257110595703125,
+      "learning_rate": 2e-05,
+      "loss": 0.03142691,
+      "step": 1891
+    },
+    {
+      "epoch": 3.784,
+      "grad_norm": 1.1791033744812012,
+      "learning_rate": 2e-05,
+      "loss": 0.03500428,
+      "step": 1892
+    },
+    {
+      "epoch": 3.786,
+      "grad_norm": 1.6889530420303345,
+      "learning_rate": 2e-05,
+      "loss": 0.05240297,
+      "step": 1893
+    },
+    {
+      "epoch": 3.7880000000000003,
+      "grad_norm": 1.7314518690109253,
+      "learning_rate": 2e-05,
+      "loss": 0.04307431,
+      "step": 1894
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.1842002868652344,
+      "learning_rate": 2e-05,
+      "loss": 0.04247685,
+      "step": 1895
+    },
+    {
+      "epoch": 3.792,
+      "grad_norm": 2.100853681564331,
+      "learning_rate": 2e-05,
+      "loss": 0.05021702,
+      "step": 1896
+    },
+    {
+      "epoch": 3.794,
+      "grad_norm": 1.5216481685638428,
+      "learning_rate": 2e-05,
+      "loss": 0.03309655,
+      "step": 1897
+    },
+    {
+      "epoch": 3.7960000000000003,
+      "grad_norm": 2.276388168334961,
+      "learning_rate": 2e-05,
+      "loss": 0.05299261,
+      "step": 1898
+    },
+    {
+      "epoch": 3.798,
+      "grad_norm": 2.0013959407806396,
+      "learning_rate": 2e-05,
+      "loss": 0.04387742,
+      "step": 1899
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.5426726341247559,
+      "learning_rate": 2e-05,
+      "loss": 0.04475781,
+      "step": 1900
+    },
+    {
+      "epoch": 3.802,
+      "grad_norm": 1.730944037437439,
+      "learning_rate": 2e-05,
+      "loss": 0.04797804,
+      "step": 1901
+    },
+    {
+      "epoch": 3.8040000000000003,
+      "grad_norm": 3.512629747390747,
+      "learning_rate": 2e-05,
+      "loss": 0.06400887,
+      "step": 1902
+    },
+    {
+      "epoch": 3.806,
+      "grad_norm": 1.5255076885223389,
+      "learning_rate": 2e-05,
+      "loss": 0.03512551,
+      "step": 1903
+    },
+    {
+      "epoch": 3.808,
+      "grad_norm": 1.6803914308547974,
+      "learning_rate": 2e-05,
+      "loss": 0.04707751,
+      "step": 1904
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 1.6464742422103882,
+      "learning_rate": 2e-05,
+      "loss": 0.04364287,
+      "step": 1905
+    },
+    {
+      "epoch": 3.8120000000000003,
+      "grad_norm": 1.5913269519805908,
+      "learning_rate": 2e-05,
+      "loss": 0.05464795,
+      "step": 1906
+    },
+    {
+      "epoch": 3.814,
+      "grad_norm": 1.8374617099761963,
+      "learning_rate": 2e-05,
+      "loss": 0.04795897,
+      "step": 1907
+    },
+    {
+      "epoch": 3.816,
+      "grad_norm": 2.12955641746521,
+      "learning_rate": 2e-05,
+      "loss": 0.06177934,
+      "step": 1908
+    },
+    {
+      "epoch": 3.818,
+      "grad_norm": 1.9309619665145874,
+      "learning_rate": 2e-05,
+      "loss": 0.02248725,
+      "step": 1909
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 1.7270091772079468,
+      "learning_rate": 2e-05,
+      "loss": 0.03895885,
+      "step": 1910
+    },
+    {
+      "epoch": 3.822,
+      "grad_norm": 1.6948938369750977,
+      "learning_rate": 2e-05,
+      "loss": 0.06202386,
+      "step": 1911
+    },
+    {
+      "epoch": 3.824,
+      "grad_norm": 1.287054181098938,
+      "learning_rate": 2e-05,
+      "loss": 0.03878157,
+      "step": 1912
+    },
+    {
+      "epoch": 3.826,
+      "grad_norm": 1.2202260494232178,
+      "learning_rate": 2e-05,
+      "loss": 0.03554882,
+      "step": 1913
+    },
+    {
+      "epoch": 3.828,
+      "grad_norm": 0.9808116555213928,
+      "learning_rate": 2e-05,
+      "loss": 0.02788639,
+      "step": 1914
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.3717739582061768,
+      "learning_rate": 2e-05,
+      "loss": 0.04405915,
+      "step": 1915
+    },
+    {
+      "epoch": 3.832,
+      "grad_norm": 1.2487013339996338,
+      "learning_rate": 2e-05,
+      "loss": 0.03764036,
+      "step": 1916
+    },
+    {
+      "epoch": 3.834,
+      "grad_norm": 1.122071623802185,
+      "learning_rate": 2e-05,
+      "loss": 0.0333156,
+      "step": 1917
+    },
+    {
+      "epoch": 3.836,
+      "grad_norm": 1.3125051259994507,
+      "learning_rate": 2e-05,
+      "loss": 0.04306629,
+      "step": 1918
+    },
+    {
+      "epoch": 3.838,
+      "grad_norm": 1.7904739379882812,
+      "learning_rate": 2e-05,
+      "loss": 0.0361197,
+      "step": 1919
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.1583505868911743,
+      "learning_rate": 2e-05,
+      "loss": 0.03510105,
+      "step": 1920
+    },
+    {
+      "epoch": 3.842,
+      "grad_norm": 1.6849479675292969,
+      "learning_rate": 2e-05,
+      "loss": 0.04817548,
+      "step": 1921
+    },
+    {
+      "epoch": 3.844,
+      "grad_norm": 2.617076873779297,
+      "learning_rate": 2e-05,
+      "loss": 0.06620353,
+      "step": 1922
+    },
+    {
+      "epoch": 3.846,
+      "grad_norm": 2.478874683380127,
+      "learning_rate": 2e-05,
+      "loss": 0.04455588,
+      "step": 1923
+    },
+    {
+      "epoch": 3.848,
+      "grad_norm": 1.146331548690796,
+      "learning_rate": 2e-05,
+      "loss": 0.04490707,
+      "step": 1924
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 1.0759429931640625,
+      "learning_rate": 2e-05,
+      "loss": 0.03621227,
+      "step": 1925
+    },
+    {
+      "epoch": 3.852,
+      "grad_norm": 1.6324303150177002,
+      "learning_rate": 2e-05,
+      "loss": 0.04010741,
+      "step": 1926
+    },
+    {
+      "epoch": 3.854,
+      "grad_norm": 2.0158345699310303,
+      "learning_rate": 2e-05,
+      "loss": 0.05399124,
+      "step": 1927
+    },
+    {
+      "epoch": 3.856,
+      "grad_norm": 1.5364727973937988,
+      "learning_rate": 2e-05,
+      "loss": 0.04785627,
+      "step": 1928
+    },
+    {
+      "epoch": 3.858,
+      "grad_norm": 1.6611193418502808,
+      "learning_rate": 2e-05,
+      "loss": 0.03403072,
+      "step": 1929
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 2.6205263137817383,
+      "learning_rate": 2e-05,
+      "loss": 0.04760019,
+      "step": 1930
+    },
+    {
+      "epoch": 3.862,
+      "grad_norm": 1.1166229248046875,
+      "learning_rate": 2e-05,
+      "loss": 0.03396893,
+      "step": 1931
+    },
+    {
+      "epoch": 3.864,
+      "grad_norm": 2.4609854221343994,
+      "learning_rate": 2e-05,
+      "loss": 0.05028281,
+      "step": 1932
+    },
+    {
+      "epoch": 3.866,
+      "grad_norm": 1.2173758745193481,
+      "learning_rate": 2e-05,
+      "loss": 0.03039585,
+      "step": 1933
+    },
+    {
+      "epoch": 3.868,
+      "grad_norm": 1.3882139921188354,
+      "learning_rate": 2e-05,
+      "loss": 0.04112496,
+      "step": 1934
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 1.486579418182373,
+      "learning_rate": 2e-05,
+      "loss": 0.03768492,
+      "step": 1935
+    },
+    {
+      "epoch": 3.872,
+      "grad_norm": 1.4759503602981567,
+      "learning_rate": 2e-05,
+      "loss": 0.03338876,
+      "step": 1936
+    },
+    {
+      "epoch": 3.874,
+      "grad_norm": 1.290807843208313,
+      "learning_rate": 2e-05,
+      "loss": 0.03643284,
+      "step": 1937
+    },
+    {
+      "epoch": 3.876,
+      "grad_norm": 1.520533800125122,
+      "learning_rate": 2e-05,
+      "loss": 0.04213966,
+      "step": 1938
+    },
+    {
+      "epoch": 3.878,
+      "grad_norm": 1.262356162071228,
+      "learning_rate": 2e-05,
+      "loss": 0.04538347,
+      "step": 1939
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 2.381612777709961,
+      "learning_rate": 2e-05,
+      "loss": 0.03353626,
+      "step": 1940
+    },
+    {
+      "epoch": 3.882,
+      "grad_norm": 1.6339051723480225,
+      "learning_rate": 2e-05,
+      "loss": 0.04302477,
+      "step": 1941
+    },
+    {
+      "epoch": 3.884,
+      "grad_norm": 1.4285898208618164,
+      "learning_rate": 2e-05,
+      "loss": 0.05180392,
+      "step": 1942
+    },
+    {
+      "epoch": 3.886,
+      "grad_norm": 1.1873326301574707,
+      "learning_rate": 2e-05,
+      "loss": 0.03653756,
+      "step": 1943
+    },
+    {
+      "epoch": 3.888,
+      "grad_norm": 1.3497165441513062,
+      "learning_rate": 2e-05,
+      "loss": 0.03856751,
+      "step": 1944
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 1.258852481842041,
+      "learning_rate": 2e-05,
+      "loss": 0.04039215,
+      "step": 1945
+    },
+    {
+      "epoch": 3.892,
+      "grad_norm": 1.983088731765747,
+      "learning_rate": 2e-05,
+      "loss": 0.04817604,
+      "step": 1946
+    },
+    {
+      "epoch": 3.894,
+      "grad_norm": 2.92964768409729,
+      "learning_rate": 2e-05,
+      "loss": 0.053841,
+      "step": 1947
+    },
+    {
+      "epoch": 3.896,
+      "grad_norm": 1.1992443799972534,
+      "learning_rate": 2e-05,
+      "loss": 0.04172537,
+      "step": 1948
+    },
+    {
+      "epoch": 3.898,
+      "grad_norm": 1.1795976161956787,
+      "learning_rate": 2e-05,
+      "loss": 0.03664827,
+      "step": 1949
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 1.2616045475006104,
+      "learning_rate": 2e-05,
+      "loss": 0.04682916,
+      "step": 1950
+    },
+    {
+      "epoch": 3.902,
+      "grad_norm": 3.0420405864715576,
+      "learning_rate": 2e-05,
+      "loss": 0.0575822,
+      "step": 1951
+    },
+    {
+      "epoch": 3.904,
+      "grad_norm": 1.0600016117095947,
+      "learning_rate": 2e-05,
+      "loss": 0.03584382,
+      "step": 1952
+    },
+    {
+      "epoch": 3.906,
+      "grad_norm": 1.396815299987793,
+      "learning_rate": 2e-05,
+      "loss": 0.03886579,
+      "step": 1953
+    },
+    {
+      "epoch": 3.908,
+      "grad_norm": 2.188819408416748,
+      "learning_rate": 2e-05,
+      "loss": 0.06812514,
+      "step": 1954
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 1.5482600927352905,
+      "learning_rate": 2e-05,
+      "loss": 0.03375176,
+      "step": 1955
+    },
+    {
+      "epoch": 3.912,
+      "grad_norm": 1.212693691253662,
+      "learning_rate": 2e-05,
+      "loss": 0.0395262,
+      "step": 1956
+    },
+    {
+      "epoch": 3.914,
+      "grad_norm": 2.2088680267333984,
+      "learning_rate": 2e-05,
+      "loss": 0.03797888,
+      "step": 1957
+    },
+    {
+      "epoch": 3.916,
+      "grad_norm": 1.1913862228393555,
+      "learning_rate": 2e-05,
+      "loss": 0.02954198,
+      "step": 1958
+    },
+    {
+      "epoch": 3.918,
+      "grad_norm": 1.874579906463623,
+      "learning_rate": 2e-05,
+      "loss": 0.05844161,
+      "step": 1959
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 1.4673166275024414,
+      "learning_rate": 2e-05,
+      "loss": 0.05150342,
+      "step": 1960
+    },
+    {
+      "epoch": 3.922,
+      "grad_norm": 2.861666679382324,
+      "learning_rate": 2e-05,
+      "loss": 0.05179626,
+      "step": 1961
+    },
+    {
+      "epoch": 3.924,
+      "grad_norm": 1.771929144859314,
+      "learning_rate": 2e-05,
+      "loss": 0.04365551,
+      "step": 1962
+    },
+    {
+      "epoch": 3.926,
+      "grad_norm": 1.3273606300354004,
+      "learning_rate": 2e-05,
+      "loss": 0.05284184,
+      "step": 1963
+    },
+    {
+      "epoch": 3.928,
+      "grad_norm": 1.9596598148345947,
+      "learning_rate": 2e-05,
+      "loss": 0.05003749,
+      "step": 1964
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 2.103850841522217,
+      "learning_rate": 2e-05,
+      "loss": 0.04942531,
+      "step": 1965
+    },
+    {
+      "epoch": 3.932,
+      "grad_norm": 1.1391785144805908,
+      "learning_rate": 2e-05,
+      "loss": 0.02786987,
+      "step": 1966
+    },
+    {
+      "epoch": 3.934,
+      "grad_norm": 1.7148336172103882,
+      "learning_rate": 2e-05,
+      "loss": 0.04135071,
+      "step": 1967
+    },
+    {
+      "epoch": 3.936,
+      "grad_norm": 1.2991632223129272,
+      "learning_rate": 2e-05,
+      "loss": 0.06019251,
+      "step": 1968
+    },
+    {
+      "epoch": 3.9379999999999997,
+      "grad_norm": 1.5509945154190063,
+      "learning_rate": 2e-05,
+      "loss": 0.03836875,
+      "step": 1969
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 1.69589364528656,
+      "learning_rate": 2e-05,
+      "loss": 0.05219426,
+      "step": 1970
+    },
+    {
+      "epoch": 3.942,
+      "grad_norm": 1.6247663497924805,
+      "learning_rate": 2e-05,
+      "loss": 0.05570246,
+      "step": 1971
+    },
+    {
+      "epoch": 3.944,
+      "grad_norm": 1.2264633178710938,
+      "learning_rate": 2e-05,
+      "loss": 0.04570191,
+      "step": 1972
+    },
+    {
+      "epoch": 3.9459999999999997,
+      "grad_norm": 2.1913163661956787,
+      "learning_rate": 2e-05,
+      "loss": 0.05165469,
+      "step": 1973
+    },
+    {
+      "epoch": 3.948,
+      "grad_norm": 1.0560282468795776,
+      "learning_rate": 2e-05,
+      "loss": 0.02876104,
+      "step": 1974
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 1.5731909275054932,
+      "learning_rate": 2e-05,
+      "loss": 0.04167511,
+      "step": 1975
+    },
+    {
+      "epoch": 3.952,
+      "grad_norm": 1.399192452430725,
+      "learning_rate": 2e-05,
+      "loss": 0.04265817,
+      "step": 1976
+    },
+    {
+      "epoch": 3.9539999999999997,
+      "grad_norm": 1.8346514701843262,
+      "learning_rate": 2e-05,
+      "loss": 0.04521923,
+      "step": 1977
+    },
+    {
+      "epoch": 3.956,
+      "grad_norm": 1.6072944402694702,
+      "learning_rate": 2e-05,
+      "loss": 0.04234926,
+      "step": 1978
+    },
+    {
+      "epoch": 3.958,
+      "grad_norm": 1.2165111303329468,
+      "learning_rate": 2e-05,
+      "loss": 0.03402507,
+      "step": 1979
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 1.2799257040023804,
+      "learning_rate": 2e-05,
+      "loss": 0.0419311,
+      "step": 1980
+    },
+    {
+      "epoch": 3.9619999999999997,
+      "grad_norm": 1.89944589138031,
+      "learning_rate": 2e-05,
+      "loss": 0.06260209,
+      "step": 1981
+    },
+    {
+      "epoch": 3.964,
+      "grad_norm": 1.6802691221237183,
+      "learning_rate": 2e-05,
+      "loss": 0.06025416,
+      "step": 1982
+    },
+    {
+      "epoch": 3.966,
+      "grad_norm": 1.2507396936416626,
+      "learning_rate": 2e-05,
+      "loss": 0.03038237,
+      "step": 1983
+    },
+    {
+      "epoch": 3.968,
+      "grad_norm": 1.3977298736572266,
+      "learning_rate": 2e-05,
+      "loss": 0.04241063,
+      "step": 1984
+    },
+    {
+      "epoch": 3.9699999999999998,
+      "grad_norm": 1.399631381034851,
+      "learning_rate": 2e-05,
+      "loss": 0.03736674,
+      "step": 1985
+    },
+    {
+      "epoch": 3.972,
+      "grad_norm": 1.1037839651107788,
+      "learning_rate": 2e-05,
+      "loss": 0.0263209,
+      "step": 1986
+    },
+    {
+      "epoch": 3.974,
+      "grad_norm": 1.7092723846435547,
+      "learning_rate": 2e-05,
+      "loss": 0.04478872,
+      "step": 1987
+    },
+    {
+      "epoch": 3.976,
+      "grad_norm": 2.23612117767334,
+      "learning_rate": 2e-05,
+      "loss": 0.03863956,
+      "step": 1988
+    },
+    {
+      "epoch": 3.9779999999999998,
+      "grad_norm": 1.2527269124984741,
+      "learning_rate": 2e-05,
+      "loss": 0.04319048,
+      "step": 1989
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 1.2724004983901978,
+      "learning_rate": 2e-05,
+      "loss": 0.04712515,
+      "step": 1990
+    },
+    {
+      "epoch": 3.982,
+      "grad_norm": 1.1430199146270752,
+      "learning_rate": 2e-05,
+      "loss": 0.03822499,
+      "step": 1991
+    },
+    {
+      "epoch": 3.984,
+      "grad_norm": 1.8308535814285278,
+      "learning_rate": 2e-05,
+      "loss": 0.0425014,
+      "step": 1992
+    },
+    {
+      "epoch": 3.9859999999999998,
+      "grad_norm": 1.5834559202194214,
+      "learning_rate": 2e-05,
+      "loss": 0.04654464,
+      "step": 1993
+    },
+    {
+      "epoch": 3.988,
+      "grad_norm": 1.4839645624160767,
+      "learning_rate": 2e-05,
+      "loss": 0.06130801,
+      "step": 1994
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 1.290731430053711,
+      "learning_rate": 2e-05,
+      "loss": 0.03394639,
+      "step": 1995
+    },
+    {
+      "epoch": 3.992,
+      "grad_norm": 1.468553066253662,
+      "learning_rate": 2e-05,
+      "loss": 0.05200931,
+      "step": 1996
+    },
+    {
+      "epoch": 3.9939999999999998,
+      "grad_norm": 2.330862522125244,
+      "learning_rate": 2e-05,
+      "loss": 0.04680311,
+      "step": 1997
+    },
+    {
+      "epoch": 3.996,
+      "grad_norm": 1.8820016384124756,
+      "learning_rate": 2e-05,
+      "loss": 0.06257437,
+      "step": 1998
+    },
+    {
+      "epoch": 3.998,
+      "grad_norm": 1.4825798273086548,
+      "learning_rate": 2e-05,
+      "loss": 0.03755412,
+      "step": 1999
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.8379441499710083,
+      "learning_rate": 2e-05,
+      "loss": 0.04851647,
+      "step": 2000
+    },
+    {
+      "epoch": 4.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.986,
+        "AngleClassification_2": 0.978,
+        "AngleClassification_3": 0.7784431137724551,
+        "Equal_1": 0.972,
+        "Equal_2": 0.874251497005988,
+        "Equal_3": 0.7445109780439122,
+        "LineComparison_1": 0.988,
+        "LineComparison_2": 0.9860279441117764,
+        "LineComparison_3": 0.9481037924151696,
+        "Parallel_1": 0.9919839679358717,
+        "Parallel_2": 0.9779559118236473,
+        "Parallel_3": 0.752,
+        "Perpendicular_1": 0.974,
+        "Perpendicular_2": 0.512,
+        "Perpendicular_3": 0.2565130260521042,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9976666666666667,
+        "PointLiesOnCircle_3": 0.9385333333333333,
+        "PointLiesOnLine_1": 0.9819639278557114,
+        "PointLiesOnLine_2": 0.9579158316633266,
+        "PointLiesOnLine_3": 0.5249500998003992
+      },
+      "eval_runtime": 321.6565,
+      "eval_samples_per_second": 32.644,
+      "eval_steps_per_second": 0.653,
+      "step": 2000
+    },
+    {
+      "epoch": 4.002,
+      "grad_norm": 2.689896583557129,
+      "learning_rate": 2e-05,
+      "loss": 0.06689534,
+      "step": 2001
+    },
+    {
+      "epoch": 4.004,
+      "grad_norm": 2.3726253509521484,
+      "learning_rate": 2e-05,
+      "loss": 0.06310491,
+      "step": 2002
+    },
+    {
+      "epoch": 4.006,
+      "grad_norm": 1.7566965818405151,
+      "learning_rate": 2e-05,
+      "loss": 0.05225249,
+      "step": 2003
+    },
+    {
+      "epoch": 4.008,
+      "grad_norm": 1.3011726140975952,
+      "learning_rate": 2e-05,
+      "loss": 0.03929686,
+      "step": 2004
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 1.2295416593551636,
+      "learning_rate": 2e-05,
+      "loss": 0.03298599,
+      "step": 2005
+    },
+    {
+      "epoch": 4.012,
+      "grad_norm": 1.9119962453842163,
+      "learning_rate": 2e-05,
+      "loss": 0.04884803,
+      "step": 2006
+    },
+    {
+      "epoch": 4.014,
+      "grad_norm": 1.5606404542922974,
+      "learning_rate": 2e-05,
+      "loss": 0.04282958,
+      "step": 2007
+    },
+    {
+      "epoch": 4.016,
+      "grad_norm": 2.755392074584961,
+      "learning_rate": 2e-05,
+      "loss": 0.05961449,
+      "step": 2008
+    },
+    {
+      "epoch": 4.018,
+      "grad_norm": 1.6373087167739868,
+      "learning_rate": 2e-05,
+      "loss": 0.05367161,
+      "step": 2009
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 2.529672861099243,
+      "learning_rate": 2e-05,
+      "loss": 0.04300452,
+      "step": 2010
+    },
+    {
+      "epoch": 4.022,
+      "grad_norm": 1.8200832605361938,
+      "learning_rate": 2e-05,
+      "loss": 0.05623111,
+      "step": 2011
+    },
+    {
+      "epoch": 4.024,
+      "grad_norm": 1.3841335773468018,
+      "learning_rate": 2e-05,
+      "loss": 0.03216402,
+      "step": 2012
+    },
+    {
+      "epoch": 4.026,
+      "grad_norm": 2.039104700088501,
+      "learning_rate": 2e-05,
+      "loss": 0.0455704,
+      "step": 2013
+    },
+    {
+      "epoch": 4.028,
+      "grad_norm": 2.0491440296173096,
+      "learning_rate": 2e-05,
+      "loss": 0.03618041,
+      "step": 2014
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 1.554465889930725,
+      "learning_rate": 2e-05,
+      "loss": 0.02181048,
+      "step": 2015
+    },
+    {
+      "epoch": 4.032,
+      "grad_norm": 1.2771095037460327,
+      "learning_rate": 2e-05,
+      "loss": 0.03514237,
+      "step": 2016
+    },
+    {
+      "epoch": 4.034,
+      "grad_norm": 1.587876558303833,
+      "learning_rate": 2e-05,
+      "loss": 0.04905714,
+      "step": 2017
+    },
+    {
+      "epoch": 4.036,
+      "grad_norm": 2.0158097743988037,
+      "learning_rate": 2e-05,
+      "loss": 0.05387116,
+      "step": 2018
+    },
+    {
+      "epoch": 4.038,
+      "grad_norm": 2.1845550537109375,
+      "learning_rate": 2e-05,
+      "loss": 0.06457246,
+      "step": 2019
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 1.155949354171753,
+      "learning_rate": 2e-05,
+      "loss": 0.02948808,
+      "step": 2020
+    },
+    {
+      "epoch": 4.042,
+      "grad_norm": 1.5860084295272827,
+      "learning_rate": 2e-05,
+      "loss": 0.05309289,
+      "step": 2021
+    },
+    {
+      "epoch": 4.044,
+      "grad_norm": 1.1510523557662964,
+      "learning_rate": 2e-05,
+      "loss": 0.04007091,
+      "step": 2022
+    },
+    {
+      "epoch": 4.046,
+      "grad_norm": 2.096463680267334,
+      "learning_rate": 2e-05,
+      "loss": 0.05365406,
+      "step": 2023
+    },
+    {
+      "epoch": 4.048,
+      "grad_norm": 1.4027256965637207,
+      "learning_rate": 2e-05,
+      "loss": 0.03679424,
+      "step": 2024
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 1.503724455833435,
+      "learning_rate": 2e-05,
+      "loss": 0.03995445,
+      "step": 2025
+    },
+    {
+      "epoch": 4.052,
+      "grad_norm": 3.1506829261779785,
+      "learning_rate": 2e-05,
+      "loss": 0.07259705,
+      "step": 2026
+    },
+    {
+      "epoch": 4.054,
+      "grad_norm": 1.8490493297576904,
+      "learning_rate": 2e-05,
+      "loss": 0.04446945,
+      "step": 2027
+    },
+    {
+      "epoch": 4.056,
+      "grad_norm": 1.203890323638916,
+      "learning_rate": 2e-05,
+      "loss": 0.04504851,
+      "step": 2028
+    },
+    {
+      "epoch": 4.058,
+      "grad_norm": 2.1894469261169434,
+      "learning_rate": 2e-05,
+      "loss": 0.03918433,
+      "step": 2029
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 1.990956425666809,
+      "learning_rate": 2e-05,
+      "loss": 0.03764991,
+      "step": 2030
+    },
+    {
+      "epoch": 4.062,
+      "grad_norm": 1.7570072412490845,
+      "learning_rate": 2e-05,
+      "loss": 0.04656107,
+      "step": 2031
+    },
+    {
+      "epoch": 4.064,
+      "grad_norm": 1.1472625732421875,
+      "learning_rate": 2e-05,
+      "loss": 0.02630271,
+      "step": 2032
+    },
+    {
+      "epoch": 4.066,
+      "grad_norm": 2.1167778968811035,
+      "learning_rate": 2e-05,
+      "loss": 0.03036687,
+      "step": 2033
+    },
+    {
+      "epoch": 4.068,
+      "grad_norm": 1.1994069814682007,
+      "learning_rate": 2e-05,
+      "loss": 0.03812218,
+      "step": 2034
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 2.758052349090576,
+      "learning_rate": 2e-05,
+      "loss": 0.0450132,
+      "step": 2035
+    },
+    {
+      "epoch": 4.072,
+      "grad_norm": 1.4564862251281738,
+      "learning_rate": 2e-05,
+      "loss": 0.03845803,
+      "step": 2036
+    },
+    {
+      "epoch": 4.074,
+      "grad_norm": 1.1046653985977173,
+      "learning_rate": 2e-05,
+      "loss": 0.0331301,
+      "step": 2037
+    },
+    {
+      "epoch": 4.076,
+      "grad_norm": 2.1612513065338135,
+      "learning_rate": 2e-05,
+      "loss": 0.05518108,
+      "step": 2038
+    },
+    {
+      "epoch": 4.078,
+      "grad_norm": 1.5514698028564453,
+      "learning_rate": 2e-05,
+      "loss": 0.03647344,
+      "step": 2039
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 2.8072474002838135,
+      "learning_rate": 2e-05,
+      "loss": 0.04705288,
+      "step": 2040
+    },
+    {
+      "epoch": 4.082,
+      "grad_norm": 1.4394696950912476,
+      "learning_rate": 2e-05,
+      "loss": 0.03800115,
+      "step": 2041
+    },
+    {
+      "epoch": 4.084,
+      "grad_norm": 1.580580711364746,
+      "learning_rate": 2e-05,
+      "loss": 0.03440983,
+      "step": 2042
+    },
+    {
+      "epoch": 4.086,
+      "grad_norm": 2.7896625995635986,
+      "learning_rate": 2e-05,
+      "loss": 0.0448301,
+      "step": 2043
+    },
+    {
+      "epoch": 4.088,
+      "grad_norm": 1.8632712364196777,
+      "learning_rate": 2e-05,
+      "loss": 0.03617346,
+      "step": 2044
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 1.8190380334854126,
+      "learning_rate": 2e-05,
+      "loss": 0.0343765,
+      "step": 2045
+    },
+    {
+      "epoch": 4.092,
+      "grad_norm": 1.6877806186676025,
+      "learning_rate": 2e-05,
+      "loss": 0.04055114,
+      "step": 2046
+    },
+    {
+      "epoch": 4.094,
+      "grad_norm": 2.117882013320923,
+      "learning_rate": 2e-05,
+      "loss": 0.048363,
+      "step": 2047
+    },
+    {
+      "epoch": 4.096,
+      "grad_norm": 2.1963632106781006,
+      "learning_rate": 2e-05,
+      "loss": 0.05769034,
+      "step": 2048
+    },
+    {
+      "epoch": 4.098,
+      "grad_norm": 2.3755226135253906,
+      "learning_rate": 2e-05,
+      "loss": 0.07200582,
+      "step": 2049
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 1.4905271530151367,
+      "learning_rate": 2e-05,
+      "loss": 0.0467202,
+      "step": 2050
+    },
+    {
+      "epoch": 4.102,
+      "grad_norm": 1.6176154613494873,
+      "learning_rate": 2e-05,
+      "loss": 0.05290875,
+      "step": 2051
+    },
+    {
+      "epoch": 4.104,
+      "grad_norm": 1.9842934608459473,
+      "learning_rate": 2e-05,
+      "loss": 0.0542838,
+      "step": 2052
+    },
+    {
+      "epoch": 4.106,
+      "grad_norm": 1.1524161100387573,
+      "learning_rate": 2e-05,
+      "loss": 0.03312632,
+      "step": 2053
+    },
+    {
+      "epoch": 4.108,
+      "grad_norm": 1.7908684015274048,
+      "learning_rate": 2e-05,
+      "loss": 0.03798482,
+      "step": 2054
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 1.8829431533813477,
+      "learning_rate": 2e-05,
+      "loss": 0.03899117,
+      "step": 2055
+    },
+    {
+      "epoch": 4.112,
+      "grad_norm": 1.7515336275100708,
+      "learning_rate": 2e-05,
+      "loss": 0.04303076,
+      "step": 2056
+    },
+    {
+      "epoch": 4.114,
+      "grad_norm": 1.6121231317520142,
+      "learning_rate": 2e-05,
+      "loss": 0.04086188,
+      "step": 2057
+    },
+    {
+      "epoch": 4.116,
+      "grad_norm": 1.3123854398727417,
+      "learning_rate": 2e-05,
+      "loss": 0.04525847,
+      "step": 2058
+    },
+    {
+      "epoch": 4.118,
+      "grad_norm": 2.75618314743042,
+      "learning_rate": 2e-05,
+      "loss": 0.05385438,
+      "step": 2059
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 1.0119373798370361,
+      "learning_rate": 2e-05,
+      "loss": 0.02559848,
+      "step": 2060
+    },
+    {
+      "epoch": 4.122,
+      "grad_norm": 1.6642717123031616,
+      "learning_rate": 2e-05,
+      "loss": 0.0500881,
+      "step": 2061
+    },
+    {
+      "epoch": 4.124,
+      "grad_norm": 1.3954622745513916,
+      "learning_rate": 2e-05,
+      "loss": 0.03746237,
+      "step": 2062
+    },
+    {
+      "epoch": 4.126,
+      "grad_norm": 1.5439870357513428,
+      "learning_rate": 2e-05,
+      "loss": 0.05281479,
+      "step": 2063
+    },
+    {
+      "epoch": 4.128,
+      "grad_norm": 1.3099309206008911,
+      "learning_rate": 2e-05,
+      "loss": 0.03846513,
+      "step": 2064
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 1.7383625507354736,
+      "learning_rate": 2e-05,
+      "loss": 0.04843453,
+      "step": 2065
+    },
+    {
+      "epoch": 4.132,
+      "grad_norm": 1.8208372592926025,
+      "learning_rate": 2e-05,
+      "loss": 0.05107617,
+      "step": 2066
+    },
+    {
+      "epoch": 4.134,
+      "grad_norm": 2.1992876529693604,
+      "learning_rate": 2e-05,
+      "loss": 0.05615732,
+      "step": 2067
+    },
+    {
+      "epoch": 4.136,
+      "grad_norm": 1.8214548826217651,
+      "learning_rate": 2e-05,
+      "loss": 0.05642605,
+      "step": 2068
+    },
+    {
+      "epoch": 4.138,
+      "grad_norm": 1.3719658851623535,
+      "learning_rate": 2e-05,
+      "loss": 0.04263125,
+      "step": 2069
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 1.8242491483688354,
+      "learning_rate": 2e-05,
+      "loss": 0.0747356,
+      "step": 2070
+    },
+    {
+      "epoch": 4.142,
+      "grad_norm": 2.162917375564575,
+      "learning_rate": 2e-05,
+      "loss": 0.05902673,
+      "step": 2071
+    },
+    {
+      "epoch": 4.144,
+      "grad_norm": 1.3159345388412476,
+      "learning_rate": 2e-05,
+      "loss": 0.04596569,
+      "step": 2072
+    },
+    {
+      "epoch": 4.146,
+      "grad_norm": 1.8809359073638916,
+      "learning_rate": 2e-05,
+      "loss": 0.05340003,
+      "step": 2073
+    },
+    {
+      "epoch": 4.148,
+      "grad_norm": 1.67131769657135,
+      "learning_rate": 2e-05,
+      "loss": 0.03571043,
+      "step": 2074
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 1.7250995635986328,
+      "learning_rate": 2e-05,
+      "loss": 0.0564668,
+      "step": 2075
+    },
+    {
+      "epoch": 4.152,
+      "grad_norm": 1.3855060338974,
+      "learning_rate": 2e-05,
+      "loss": 0.04370652,
+      "step": 2076
+    },
+    {
+      "epoch": 4.154,
+      "grad_norm": 1.1010857820510864,
+      "learning_rate": 2e-05,
+      "loss": 0.03384936,
+      "step": 2077
+    },
+    {
+      "epoch": 4.156,
+      "grad_norm": 2.0639870166778564,
+      "learning_rate": 2e-05,
+      "loss": 0.03384199,
+      "step": 2078
+    },
+    {
+      "epoch": 4.158,
+      "grad_norm": 1.4212632179260254,
+      "learning_rate": 2e-05,
+      "loss": 0.03396594,
+      "step": 2079
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 2.1525704860687256,
+      "learning_rate": 2e-05,
+      "loss": 0.05491355,
+      "step": 2080
+    },
+    {
+      "epoch": 4.162,
+      "grad_norm": 1.7113659381866455,
+      "learning_rate": 2e-05,
+      "loss": 0.04376032,
+      "step": 2081
+    },
+    {
+      "epoch": 4.164,
+      "grad_norm": 1.223374366760254,
+      "learning_rate": 2e-05,
+      "loss": 0.03837118,
+      "step": 2082
+    },
+    {
+      "epoch": 4.166,
+      "grad_norm": 1.7024589776992798,
+      "learning_rate": 2e-05,
+      "loss": 0.04713647,
+      "step": 2083
+    },
+    {
+      "epoch": 4.168,
+      "grad_norm": 1.7157230377197266,
+      "learning_rate": 2e-05,
+      "loss": 0.05556125,
+      "step": 2084
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 1.5272091627120972,
+      "learning_rate": 2e-05,
+      "loss": 0.04492594,
+      "step": 2085
+    },
+    {
+      "epoch": 4.172,
+      "grad_norm": 1.2956258058547974,
+      "learning_rate": 2e-05,
+      "loss": 0.04288995,
+      "step": 2086
+    },
+    {
+      "epoch": 4.174,
+      "grad_norm": 1.3768620491027832,
+      "learning_rate": 2e-05,
+      "loss": 0.02881315,
+      "step": 2087
+    },
+    {
+      "epoch": 4.176,
+      "grad_norm": 1.5189354419708252,
+      "learning_rate": 2e-05,
+      "loss": 0.05045771,
+      "step": 2088
+    },
+    {
+      "epoch": 4.178,
+      "grad_norm": 1.8950650691986084,
+      "learning_rate": 2e-05,
+      "loss": 0.05032203,
+      "step": 2089
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 1.9955432415008545,
+      "learning_rate": 2e-05,
+      "loss": 0.02519334,
+      "step": 2090
+    },
+    {
+      "epoch": 4.182,
+      "grad_norm": 1.4326856136322021,
+      "learning_rate": 2e-05,
+      "loss": 0.03771212,
+      "step": 2091
+    },
+    {
+      "epoch": 4.184,
+      "grad_norm": 1.1596758365631104,
+      "learning_rate": 2e-05,
+      "loss": 0.02768318,
+      "step": 2092
+    },
+    {
+      "epoch": 4.186,
+      "grad_norm": 1.1794545650482178,
+      "learning_rate": 2e-05,
+      "loss": 0.03789561,
+      "step": 2093
+    },
+    {
+      "epoch": 4.188,
+      "grad_norm": 0.9308399558067322,
+      "learning_rate": 2e-05,
+      "loss": 0.02805153,
+      "step": 2094
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 1.1102815866470337,
+      "learning_rate": 2e-05,
+      "loss": 0.02839583,
+      "step": 2095
+    },
+    {
+      "epoch": 4.192,
+      "grad_norm": 1.020172357559204,
+      "learning_rate": 2e-05,
+      "loss": 0.03357933,
+      "step": 2096
+    },
+    {
+      "epoch": 4.194,
+      "grad_norm": 1.7360633611679077,
+      "learning_rate": 2e-05,
+      "loss": 0.05581947,
+      "step": 2097
+    },
+    {
+      "epoch": 4.196,
+      "grad_norm": 1.2498151063919067,
+      "learning_rate": 2e-05,
+      "loss": 0.03497965,
+      "step": 2098
+    },
+    {
+      "epoch": 4.198,
+      "grad_norm": 1.4616661071777344,
+      "learning_rate": 2e-05,
+      "loss": 0.04537179,
+      "step": 2099
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 1.6234368085861206,
+      "learning_rate": 2e-05,
+      "loss": 0.0405514,
+      "step": 2100
+    },
+    {
+      "epoch": 4.202,
+      "grad_norm": 1.50863516330719,
+      "learning_rate": 2e-05,
+      "loss": 0.03994677,
+      "step": 2101
+    },
+    {
+      "epoch": 4.204,
+      "grad_norm": 2.2599105834960938,
+      "learning_rate": 2e-05,
+      "loss": 0.04103535,
+      "step": 2102
+    },
+    {
+      "epoch": 4.206,
+      "grad_norm": 1.642195224761963,
+      "learning_rate": 2e-05,
+      "loss": 0.02920062,
+      "step": 2103
+    },
+    {
+      "epoch": 4.208,
+      "grad_norm": 2.0362472534179688,
+      "learning_rate": 2e-05,
+      "loss": 0.03564665,
+      "step": 2104
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 2.113086223602295,
+      "learning_rate": 2e-05,
+      "loss": 0.04320297,
+      "step": 2105
+    },
+    {
+      "epoch": 4.212,
+      "grad_norm": 1.1269563436508179,
+      "learning_rate": 2e-05,
+      "loss": 0.02520046,
+      "step": 2106
+    },
+    {
+      "epoch": 4.214,
+      "grad_norm": 2.32094144821167,
+      "learning_rate": 2e-05,
+      "loss": 0.06330884,
+      "step": 2107
+    },
+    {
+      "epoch": 4.216,
+      "grad_norm": 1.5615673065185547,
+      "learning_rate": 2e-05,
+      "loss": 0.0331628,
+      "step": 2108
+    },
+    {
+      "epoch": 4.218,
+      "grad_norm": 0.9397850632667542,
+      "learning_rate": 2e-05,
+      "loss": 0.02459784,
+      "step": 2109
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 1.9889795780181885,
+      "learning_rate": 2e-05,
+      "loss": 0.03781548,
+      "step": 2110
+    },
+    {
+      "epoch": 4.222,
+      "grad_norm": 1.8364139795303345,
+      "learning_rate": 2e-05,
+      "loss": 0.03902372,
+      "step": 2111
+    },
+    {
+      "epoch": 4.224,
+      "grad_norm": 1.2772095203399658,
+      "learning_rate": 2e-05,
+      "loss": 0.0316204,
+      "step": 2112
+    },
+    {
+      "epoch": 4.226,
+      "grad_norm": 2.7459285259246826,
+      "learning_rate": 2e-05,
+      "loss": 0.06582908,
+      "step": 2113
+    },
+    {
+      "epoch": 4.228,
+      "grad_norm": 1.5086251497268677,
+      "learning_rate": 2e-05,
+      "loss": 0.04670456,
+      "step": 2114
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 2.2771124839782715,
+      "learning_rate": 2e-05,
+      "loss": 0.07231095,
+      "step": 2115
+    },
+    {
+      "epoch": 4.232,
+      "grad_norm": 1.7897669076919556,
+      "learning_rate": 2e-05,
+      "loss": 0.04256113,
+      "step": 2116
+    },
+    {
+      "epoch": 4.234,
+      "grad_norm": 1.2214994430541992,
+      "learning_rate": 2e-05,
+      "loss": 0.03983855,
+      "step": 2117
+    },
+    {
+      "epoch": 4.236,
+      "grad_norm": 1.4263696670532227,
+      "learning_rate": 2e-05,
+      "loss": 0.03172107,
+      "step": 2118
+    },
+    {
+      "epoch": 4.2379999999999995,
+      "grad_norm": 2.4489498138427734,
+      "learning_rate": 2e-05,
+      "loss": 0.05162663,
+      "step": 2119
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 1.7401384115219116,
+      "learning_rate": 2e-05,
+      "loss": 0.0567259,
+      "step": 2120
+    },
+    {
+      "epoch": 4.242,
+      "grad_norm": 2.355661153793335,
+      "learning_rate": 2e-05,
+      "loss": 0.04312699,
+      "step": 2121
+    },
+    {
+      "epoch": 4.244,
+      "grad_norm": 0.8899902701377869,
+      "learning_rate": 2e-05,
+      "loss": 0.02231044,
+      "step": 2122
+    },
+    {
+      "epoch": 4.246,
+      "grad_norm": 1.3299999237060547,
+      "learning_rate": 2e-05,
+      "loss": 0.03867707,
+      "step": 2123
+    },
+    {
+      "epoch": 4.248,
+      "grad_norm": 2.1780574321746826,
+      "learning_rate": 2e-05,
+      "loss": 0.05338889,
+      "step": 2124
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.6817718744277954,
+      "learning_rate": 2e-05,
+      "loss": 0.04446897,
+      "step": 2125
+    },
+    {
+      "epoch": 4.252,
+      "grad_norm": 1.7029000520706177,
+      "learning_rate": 2e-05,
+      "loss": 0.03678191,
+      "step": 2126
+    },
+    {
+      "epoch": 4.254,
+      "grad_norm": 1.601068139076233,
+      "learning_rate": 2e-05,
+      "loss": 0.04554681,
+      "step": 2127
+    },
+    {
+      "epoch": 4.256,
+      "grad_norm": 2.417534828186035,
+      "learning_rate": 2e-05,
+      "loss": 0.06136102,
+      "step": 2128
+    },
+    {
+      "epoch": 4.258,
+      "grad_norm": 2.018188953399658,
+      "learning_rate": 2e-05,
+      "loss": 0.06119014,
+      "step": 2129
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 2.963576078414917,
+      "learning_rate": 2e-05,
+      "loss": 0.06493448,
+      "step": 2130
+    },
+    {
+      "epoch": 4.2620000000000005,
+      "grad_norm": 1.6151946783065796,
+      "learning_rate": 2e-05,
+      "loss": 0.05280596,
+      "step": 2131
+    },
+    {
+      "epoch": 4.264,
+      "grad_norm": 1.7862824201583862,
+      "learning_rate": 2e-05,
+      "loss": 0.03816576,
+      "step": 2132
+    },
+    {
+      "epoch": 4.266,
+      "grad_norm": 2.704598903656006,
+      "learning_rate": 2e-05,
+      "loss": 0.06200828,
+      "step": 2133
+    },
+    {
+      "epoch": 4.268,
+      "grad_norm": 1.2484264373779297,
+      "learning_rate": 2e-05,
+      "loss": 0.02782817,
+      "step": 2134
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 1.6894198656082153,
+      "learning_rate": 2e-05,
+      "loss": 0.0363794,
+      "step": 2135
+    },
+    {
+      "epoch": 4.272,
+      "grad_norm": 1.383814811706543,
+      "learning_rate": 2e-05,
+      "loss": 0.0340181,
+      "step": 2136
+    },
+    {
+      "epoch": 4.274,
+      "grad_norm": 1.5126454830169678,
+      "learning_rate": 2e-05,
+      "loss": 0.04073911,
+      "step": 2137
+    },
+    {
+      "epoch": 4.276,
+      "grad_norm": 1.4518946409225464,
+      "learning_rate": 2e-05,
+      "loss": 0.04448681,
+      "step": 2138
+    },
+    {
+      "epoch": 4.2780000000000005,
+      "grad_norm": 1.3890987634658813,
+      "learning_rate": 2e-05,
+      "loss": 0.02615726,
+      "step": 2139
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 1.6702042818069458,
+      "learning_rate": 2e-05,
+      "loss": 0.05457231,
+      "step": 2140
+    },
+    {
+      "epoch": 4.282,
+      "grad_norm": 2.1039538383483887,
+      "learning_rate": 2e-05,
+      "loss": 0.04261757,
+      "step": 2141
+    },
+    {
+      "epoch": 4.284,
+      "grad_norm": 1.6950783729553223,
+      "learning_rate": 2e-05,
+      "loss": 0.05137327,
+      "step": 2142
+    },
+    {
+      "epoch": 4.286,
+      "grad_norm": 1.2793657779693604,
+      "learning_rate": 2e-05,
+      "loss": 0.03890596,
+      "step": 2143
+    },
+    {
+      "epoch": 4.288,
+      "grad_norm": 1.0241049528121948,
+      "learning_rate": 2e-05,
+      "loss": 0.02962255,
+      "step": 2144
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 1.3482152223587036,
+      "learning_rate": 2e-05,
+      "loss": 0.04371033,
+      "step": 2145
+    },
+    {
+      "epoch": 4.292,
+      "grad_norm": 1.825649619102478,
+      "learning_rate": 2e-05,
+      "loss": 0.05308119,
+      "step": 2146
+    },
+    {
+      "epoch": 4.294,
+      "grad_norm": 1.4800101518630981,
+      "learning_rate": 2e-05,
+      "loss": 0.0398606,
+      "step": 2147
+    },
+    {
+      "epoch": 4.296,
+      "grad_norm": 1.812041163444519,
+      "learning_rate": 2e-05,
+      "loss": 0.04485359,
+      "step": 2148
+    },
+    {
+      "epoch": 4.298,
+      "grad_norm": 1.0440144538879395,
+      "learning_rate": 2e-05,
+      "loss": 0.04142414,
+      "step": 2149
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 1.4665697813034058,
+      "learning_rate": 2e-05,
+      "loss": 0.03760947,
+      "step": 2150
+    },
+    {
+      "epoch": 4.302,
+      "grad_norm": 1.5210576057434082,
+      "learning_rate": 2e-05,
+      "loss": 0.03952883,
+      "step": 2151
+    },
+    {
+      "epoch": 4.304,
+      "grad_norm": 1.5270603895187378,
+      "learning_rate": 2e-05,
+      "loss": 0.04734674,
+      "step": 2152
+    },
+    {
+      "epoch": 4.306,
+      "grad_norm": 2.6761133670806885,
+      "learning_rate": 2e-05,
+      "loss": 0.05432082,
+      "step": 2153
+    },
+    {
+      "epoch": 4.308,
+      "grad_norm": 1.2086443901062012,
+      "learning_rate": 2e-05,
+      "loss": 0.03031913,
+      "step": 2154
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 1.5387802124023438,
+      "learning_rate": 2e-05,
+      "loss": 0.0282026,
+      "step": 2155
+    },
+    {
+      "epoch": 4.312,
+      "grad_norm": 1.694658875465393,
+      "learning_rate": 2e-05,
+      "loss": 0.05619892,
+      "step": 2156
+    },
+    {
+      "epoch": 4.314,
+      "grad_norm": 1.5027580261230469,
+      "learning_rate": 2e-05,
+      "loss": 0.04229368,
+      "step": 2157
+    },
+    {
+      "epoch": 4.316,
+      "grad_norm": 1.3247681856155396,
+      "learning_rate": 2e-05,
+      "loss": 0.0459209,
+      "step": 2158
+    },
+    {
+      "epoch": 4.318,
+      "grad_norm": 0.9879952669143677,
+      "learning_rate": 2e-05,
+      "loss": 0.03440651,
+      "step": 2159
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.2517679929733276,
+      "learning_rate": 2e-05,
+      "loss": 0.0451726,
+      "step": 2160
+    },
+    {
+      "epoch": 4.322,
+      "grad_norm": 1.853196620941162,
+      "learning_rate": 2e-05,
+      "loss": 0.03265027,
+      "step": 2161
+    },
+    {
+      "epoch": 4.324,
+      "grad_norm": 1.710205316543579,
+      "learning_rate": 2e-05,
+      "loss": 0.04459624,
+      "step": 2162
+    },
+    {
+      "epoch": 4.326,
+      "grad_norm": 1.6105256080627441,
+      "learning_rate": 2e-05,
+      "loss": 0.06459425,
+      "step": 2163
+    },
+    {
+      "epoch": 4.328,
+      "grad_norm": 1.3017997741699219,
+      "learning_rate": 2e-05,
+      "loss": 0.05446584,
+      "step": 2164
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 1.122078776359558,
+      "learning_rate": 2e-05,
+      "loss": 0.0316535,
+      "step": 2165
+    },
+    {
+      "epoch": 4.332,
+      "grad_norm": 1.4984065294265747,
+      "learning_rate": 2e-05,
+      "loss": 0.03722906,
+      "step": 2166
+    },
+    {
+      "epoch": 4.334,
+      "grad_norm": 1.2994331121444702,
+      "learning_rate": 2e-05,
+      "loss": 0.0416957,
+      "step": 2167
+    },
+    {
+      "epoch": 4.336,
+      "grad_norm": 2.6297388076782227,
+      "learning_rate": 2e-05,
+      "loss": 0.05040634,
+      "step": 2168
+    },
+    {
+      "epoch": 4.338,
+      "grad_norm": 1.6323825120925903,
+      "learning_rate": 2e-05,
+      "loss": 0.03700216,
+      "step": 2169
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 1.054328203201294,
+      "learning_rate": 2e-05,
+      "loss": 0.03690439,
+      "step": 2170
+    },
+    {
+      "epoch": 4.342,
+      "grad_norm": 1.3024274110794067,
+      "learning_rate": 2e-05,
+      "loss": 0.04472798,
+      "step": 2171
+    },
+    {
+      "epoch": 4.344,
+      "grad_norm": 1.3375612497329712,
+      "learning_rate": 2e-05,
+      "loss": 0.03631008,
+      "step": 2172
+    },
+    {
+      "epoch": 4.346,
+      "grad_norm": 1.0233983993530273,
+      "learning_rate": 2e-05,
+      "loss": 0.03837822,
+      "step": 2173
+    },
+    {
+      "epoch": 4.348,
+      "grad_norm": 2.863868474960327,
+      "learning_rate": 2e-05,
+      "loss": 0.04723253,
+      "step": 2174
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.1577509641647339,
+      "learning_rate": 2e-05,
+      "loss": 0.04589339,
+      "step": 2175
+    },
+    {
+      "epoch": 4.352,
+      "grad_norm": 1.5174366235733032,
+      "learning_rate": 2e-05,
+      "loss": 0.03971744,
+      "step": 2176
+    },
+    {
+      "epoch": 4.354,
+      "grad_norm": 1.822798490524292,
+      "learning_rate": 2e-05,
+      "loss": 0.04629911,
+      "step": 2177
+    },
+    {
+      "epoch": 4.356,
+      "grad_norm": 2.334824562072754,
+      "learning_rate": 2e-05,
+      "loss": 0.04560362,
+      "step": 2178
+    },
+    {
+      "epoch": 4.358,
+      "grad_norm": 1.554010033607483,
+      "learning_rate": 2e-05,
+      "loss": 0.04555508,
+      "step": 2179
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 1.4686816930770874,
+      "learning_rate": 2e-05,
+      "loss": 0.0364374,
+      "step": 2180
+    },
+    {
+      "epoch": 4.362,
+      "grad_norm": 1.435239315032959,
+      "learning_rate": 2e-05,
+      "loss": 0.03441803,
+      "step": 2181
+    },
+    {
+      "epoch": 4.364,
+      "grad_norm": 1.2505661249160767,
+      "learning_rate": 2e-05,
+      "loss": 0.03555541,
+      "step": 2182
+    },
+    {
+      "epoch": 4.366,
+      "grad_norm": 2.002845048904419,
+      "learning_rate": 2e-05,
+      "loss": 0.07516761,
+      "step": 2183
+    },
+    {
+      "epoch": 4.368,
+      "grad_norm": 2.682558059692383,
+      "learning_rate": 2e-05,
+      "loss": 0.0619045,
+      "step": 2184
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 1.4598617553710938,
+      "learning_rate": 2e-05,
+      "loss": 0.02849967,
+      "step": 2185
+    },
+    {
+      "epoch": 4.372,
+      "grad_norm": 2.2034389972686768,
+      "learning_rate": 2e-05,
+      "loss": 0.05034188,
+      "step": 2186
+    },
+    {
+      "epoch": 4.374,
+      "grad_norm": 1.2549974918365479,
+      "learning_rate": 2e-05,
+      "loss": 0.03955576,
+      "step": 2187
+    },
+    {
+      "epoch": 4.376,
+      "grad_norm": 1.4328988790512085,
+      "learning_rate": 2e-05,
+      "loss": 0.04043427,
+      "step": 2188
+    },
+    {
+      "epoch": 4.378,
+      "grad_norm": 1.1671024560928345,
+      "learning_rate": 2e-05,
+      "loss": 0.03745596,
+      "step": 2189
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 1.1555709838867188,
+      "learning_rate": 2e-05,
+      "loss": 0.03046387,
+      "step": 2190
+    },
+    {
+      "epoch": 4.382,
+      "grad_norm": 2.62123703956604,
+      "learning_rate": 2e-05,
+      "loss": 0.03433674,
+      "step": 2191
+    },
+    {
+      "epoch": 4.384,
+      "grad_norm": 1.5227280855178833,
+      "learning_rate": 2e-05,
+      "loss": 0.02560518,
+      "step": 2192
+    },
+    {
+      "epoch": 4.386,
+      "grad_norm": 1.8068288564682007,
+      "learning_rate": 2e-05,
+      "loss": 0.0400132,
+      "step": 2193
+    },
+    {
+      "epoch": 4.388,
+      "grad_norm": 1.201744794845581,
+      "learning_rate": 2e-05,
+      "loss": 0.04389308,
+      "step": 2194
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 1.1598666906356812,
+      "learning_rate": 2e-05,
+      "loss": 0.04563789,
+      "step": 2195
+    },
+    {
+      "epoch": 4.392,
+      "grad_norm": 1.885544776916504,
+      "learning_rate": 2e-05,
+      "loss": 0.03769413,
+      "step": 2196
+    },
+    {
+      "epoch": 4.394,
+      "grad_norm": 1.3363451957702637,
+      "learning_rate": 2e-05,
+      "loss": 0.03677461,
+      "step": 2197
+    },
+    {
+      "epoch": 4.396,
+      "grad_norm": 1.197109341621399,
+      "learning_rate": 2e-05,
+      "loss": 0.03512957,
+      "step": 2198
+    },
+    {
+      "epoch": 4.398,
+      "grad_norm": 1.189475417137146,
+      "learning_rate": 2e-05,
+      "loss": 0.05237271,
+      "step": 2199
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.9987121224403381,
+      "learning_rate": 2e-05,
+      "loss": 0.01947338,
+      "step": 2200
+    },
+    {
+      "epoch": 4.402,
+      "grad_norm": 1.0477992296218872,
+      "learning_rate": 2e-05,
+      "loss": 0.0338387,
+      "step": 2201
+    },
+    {
+      "epoch": 4.404,
+      "grad_norm": 2.6130008697509766,
+      "learning_rate": 2e-05,
+      "loss": 0.04995614,
+      "step": 2202
+    },
+    {
+      "epoch": 4.406,
+      "grad_norm": 1.40412437915802,
+      "learning_rate": 2e-05,
+      "loss": 0.04779441,
+      "step": 2203
+    },
+    {
+      "epoch": 4.408,
+      "grad_norm": 2.32883358001709,
+      "learning_rate": 2e-05,
+      "loss": 0.04897148,
+      "step": 2204
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 3.4168405532836914,
+      "learning_rate": 2e-05,
+      "loss": 0.05897108,
+      "step": 2205
+    },
+    {
+      "epoch": 4.412,
+      "grad_norm": 1.6130071878433228,
+      "learning_rate": 2e-05,
+      "loss": 0.03696927,
+      "step": 2206
+    },
+    {
+      "epoch": 4.414,
+      "grad_norm": 1.5120912790298462,
+      "learning_rate": 2e-05,
+      "loss": 0.036088,
+      "step": 2207
+    },
+    {
+      "epoch": 4.416,
+      "grad_norm": 1.9834599494934082,
+      "learning_rate": 2e-05,
+      "loss": 0.06459089,
+      "step": 2208
+    },
+    {
+      "epoch": 4.418,
+      "grad_norm": 2.0346577167510986,
+      "learning_rate": 2e-05,
+      "loss": 0.05516463,
+      "step": 2209
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 1.2212589979171753,
+      "learning_rate": 2e-05,
+      "loss": 0.0262365,
+      "step": 2210
+    },
+    {
+      "epoch": 4.422,
+      "grad_norm": 1.5008301734924316,
+      "learning_rate": 2e-05,
+      "loss": 0.04188941,
+      "step": 2211
+    },
+    {
+      "epoch": 4.424,
+      "grad_norm": 0.9288143515586853,
+      "learning_rate": 2e-05,
+      "loss": 0.02780385,
+      "step": 2212
+    },
+    {
+      "epoch": 4.426,
+      "grad_norm": 1.2318435907363892,
+      "learning_rate": 2e-05,
+      "loss": 0.02683347,
+      "step": 2213
+    },
+    {
+      "epoch": 4.428,
+      "grad_norm": 1.711323618888855,
+      "learning_rate": 2e-05,
+      "loss": 0.04659836,
+      "step": 2214
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 1.3409011363983154,
+      "learning_rate": 2e-05,
+      "loss": 0.04741023,
+      "step": 2215
+    },
+    {
+      "epoch": 4.432,
+      "grad_norm": 1.6755443811416626,
+      "learning_rate": 2e-05,
+      "loss": 0.05947503,
+      "step": 2216
+    },
+    {
+      "epoch": 4.434,
+      "grad_norm": 2.3621582984924316,
+      "learning_rate": 2e-05,
+      "loss": 0.06439509,
+      "step": 2217
+    },
+    {
+      "epoch": 4.436,
+      "grad_norm": 1.113885760307312,
+      "learning_rate": 2e-05,
+      "loss": 0.03710642,
+      "step": 2218
+    },
+    {
+      "epoch": 4.438,
+      "grad_norm": 0.9438255429267883,
+      "learning_rate": 2e-05,
+      "loss": 0.02726935,
+      "step": 2219
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 1.442383050918579,
+      "learning_rate": 2e-05,
+      "loss": 0.04566254,
+      "step": 2220
+    },
+    {
+      "epoch": 4.442,
+      "grad_norm": 0.9339209794998169,
+      "learning_rate": 2e-05,
+      "loss": 0.02968401,
+      "step": 2221
+    },
+    {
+      "epoch": 4.444,
+      "grad_norm": 1.852805256843567,
+      "learning_rate": 2e-05,
+      "loss": 0.05942626,
+      "step": 2222
+    },
+    {
+      "epoch": 4.446,
+      "grad_norm": 2.1776721477508545,
+      "learning_rate": 2e-05,
+      "loss": 0.06122681,
+      "step": 2223
+    },
+    {
+      "epoch": 4.448,
+      "grad_norm": 1.7097100019454956,
+      "learning_rate": 2e-05,
+      "loss": 0.02926885,
+      "step": 2224
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 1.3947697877883911,
+      "learning_rate": 2e-05,
+      "loss": 0.03958159,
+      "step": 2225
+    },
+    {
+      "epoch": 4.452,
+      "grad_norm": 1.4915926456451416,
+      "learning_rate": 2e-05,
+      "loss": 0.03779658,
+      "step": 2226
+    },
+    {
+      "epoch": 4.454,
+      "grad_norm": 1.4534882307052612,
+      "learning_rate": 2e-05,
+      "loss": 0.02921861,
+      "step": 2227
+    },
+    {
+      "epoch": 4.456,
+      "grad_norm": 1.5068776607513428,
+      "learning_rate": 2e-05,
+      "loss": 0.03718809,
+      "step": 2228
+    },
+    {
+      "epoch": 4.458,
+      "grad_norm": 1.521750569343567,
+      "learning_rate": 2e-05,
+      "loss": 0.04906987,
+      "step": 2229
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 1.9581565856933594,
+      "learning_rate": 2e-05,
+      "loss": 0.04215892,
+      "step": 2230
+    },
+    {
+      "epoch": 4.462,
+      "grad_norm": 1.3763012886047363,
+      "learning_rate": 2e-05,
+      "loss": 0.03585676,
+      "step": 2231
+    },
+    {
+      "epoch": 4.464,
+      "grad_norm": 1.682054042816162,
+      "learning_rate": 2e-05,
+      "loss": 0.03892795,
+      "step": 2232
+    },
+    {
+      "epoch": 4.466,
+      "grad_norm": 1.1888952255249023,
+      "learning_rate": 2e-05,
+      "loss": 0.02804256,
+      "step": 2233
+    },
+    {
+      "epoch": 4.468,
+      "grad_norm": 1.0312952995300293,
+      "learning_rate": 2e-05,
+      "loss": 0.04241008,
+      "step": 2234
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.9027299284934998,
+      "learning_rate": 2e-05,
+      "loss": 0.02175211,
+      "step": 2235
+    },
+    {
+      "epoch": 4.4719999999999995,
+      "grad_norm": 1.6123967170715332,
+      "learning_rate": 2e-05,
+      "loss": 0.04372674,
+      "step": 2236
+    },
+    {
+      "epoch": 4.474,
+      "grad_norm": 1.2008719444274902,
+      "learning_rate": 2e-05,
+      "loss": 0.03309236,
+      "step": 2237
+    },
+    {
+      "epoch": 4.476,
+      "grad_norm": 1.4866458177566528,
+      "learning_rate": 2e-05,
+      "loss": 0.04360201,
+      "step": 2238
+    },
+    {
+      "epoch": 4.478,
+      "grad_norm": 1.139384150505066,
+      "learning_rate": 2e-05,
+      "loss": 0.02492278,
+      "step": 2239
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 1.314674973487854,
+      "learning_rate": 2e-05,
+      "loss": 0.03167442,
+      "step": 2240
+    },
+    {
+      "epoch": 4.482,
+      "grad_norm": 1.1558375358581543,
+      "learning_rate": 2e-05,
+      "loss": 0.03370375,
+      "step": 2241
+    },
+    {
+      "epoch": 4.484,
+      "grad_norm": 0.98810875415802,
+      "learning_rate": 2e-05,
+      "loss": 0.03298265,
+      "step": 2242
+    },
+    {
+      "epoch": 4.486,
+      "grad_norm": 1.692529559135437,
+      "learning_rate": 2e-05,
+      "loss": 0.043402,
+      "step": 2243
+    },
+    {
+      "epoch": 4.4879999999999995,
+      "grad_norm": 2.226050853729248,
+      "learning_rate": 2e-05,
+      "loss": 0.05941268,
+      "step": 2244
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 1.0783953666687012,
+      "learning_rate": 2e-05,
+      "loss": 0.02816807,
+      "step": 2245
+    },
+    {
+      "epoch": 4.492,
+      "grad_norm": 1.9201335906982422,
+      "learning_rate": 2e-05,
+      "loss": 0.04565334,
+      "step": 2246
+    },
+    {
+      "epoch": 4.494,
+      "grad_norm": 1.2199456691741943,
+      "learning_rate": 2e-05,
+      "loss": 0.03948256,
+      "step": 2247
+    },
+    {
+      "epoch": 4.496,
+      "grad_norm": 2.3383970260620117,
+      "learning_rate": 2e-05,
+      "loss": 0.0555183,
+      "step": 2248
+    },
+    {
+      "epoch": 4.498,
+      "grad_norm": 1.1793116331100464,
+      "learning_rate": 2e-05,
+      "loss": 0.03439267,
+      "step": 2249
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 1.0222777128219604,
+      "learning_rate": 2e-05,
+      "loss": 0.02550181,
+      "step": 2250
+    },
+    {
+      "epoch": 4.502,
+      "grad_norm": 1.751338005065918,
+      "learning_rate": 2e-05,
+      "loss": 0.03421024,
+      "step": 2251
+    },
+    {
+      "epoch": 4.504,
+      "grad_norm": 1.8420368432998657,
+      "learning_rate": 2e-05,
+      "loss": 0.04032079,
+      "step": 2252
+    },
+    {
+      "epoch": 4.506,
+      "grad_norm": 1.5478731393814087,
+      "learning_rate": 2e-05,
+      "loss": 0.04113689,
+      "step": 2253
+    },
+    {
+      "epoch": 4.508,
+      "grad_norm": 1.9176809787750244,
+      "learning_rate": 2e-05,
+      "loss": 0.03211041,
+      "step": 2254
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 1.356545090675354,
+      "learning_rate": 2e-05,
+      "loss": 0.03811408,
+      "step": 2255
+    },
+    {
+      "epoch": 4.5120000000000005,
+      "grad_norm": 1.470837950706482,
+      "learning_rate": 2e-05,
+      "loss": 0.04642755,
+      "step": 2256
+    },
+    {
+      "epoch": 4.514,
+      "grad_norm": 2.1132113933563232,
+      "learning_rate": 2e-05,
+      "loss": 0.06317526,
+      "step": 2257
+    },
+    {
+      "epoch": 4.516,
+      "grad_norm": 1.2848460674285889,
+      "learning_rate": 2e-05,
+      "loss": 0.02209956,
+      "step": 2258
+    },
+    {
+      "epoch": 4.518,
+      "grad_norm": 2.304513454437256,
+      "learning_rate": 2e-05,
+      "loss": 0.05873321,
+      "step": 2259
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 1.1547423601150513,
+      "learning_rate": 2e-05,
+      "loss": 0.03009026,
+      "step": 2260
+    },
+    {
+      "epoch": 4.522,
+      "grad_norm": 2.2654247283935547,
+      "learning_rate": 2e-05,
+      "loss": 0.05305985,
+      "step": 2261
+    },
+    {
+      "epoch": 4.524,
+      "grad_norm": 1.3709189891815186,
+      "learning_rate": 2e-05,
+      "loss": 0.0413008,
+      "step": 2262
+    },
+    {
+      "epoch": 4.526,
+      "grad_norm": 1.2186251878738403,
+      "learning_rate": 2e-05,
+      "loss": 0.04457299,
+      "step": 2263
+    },
+    {
+      "epoch": 4.5280000000000005,
+      "grad_norm": 1.8440145254135132,
+      "learning_rate": 2e-05,
+      "loss": 0.04306265,
+      "step": 2264
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 1.8175358772277832,
+      "learning_rate": 2e-05,
+      "loss": 0.04188886,
+      "step": 2265
+    },
+    {
+      "epoch": 4.532,
+      "grad_norm": 1.7477566003799438,
+      "learning_rate": 2e-05,
+      "loss": 0.03470108,
+      "step": 2266
+    },
+    {
+      "epoch": 4.534,
+      "grad_norm": 1.1122045516967773,
+      "learning_rate": 2e-05,
+      "loss": 0.03090529,
+      "step": 2267
+    },
+    {
+      "epoch": 4.536,
+      "grad_norm": 1.6068907976150513,
+      "learning_rate": 2e-05,
+      "loss": 0.04337908,
+      "step": 2268
+    },
+    {
+      "epoch": 4.538,
+      "grad_norm": 1.5025036334991455,
+      "learning_rate": 2e-05,
+      "loss": 0.04078036,
+      "step": 2269
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 1.6896750926971436,
+      "learning_rate": 2e-05,
+      "loss": 0.05480632,
+      "step": 2270
+    },
+    {
+      "epoch": 4.542,
+      "grad_norm": 1.4953817129135132,
+      "learning_rate": 2e-05,
+      "loss": 0.02734038,
+      "step": 2271
+    },
+    {
+      "epoch": 4.5440000000000005,
+      "grad_norm": 1.010339617729187,
+      "learning_rate": 2e-05,
+      "loss": 0.03597734,
+      "step": 2272
+    },
+    {
+      "epoch": 4.546,
+      "grad_norm": 1.4324558973312378,
+      "learning_rate": 2e-05,
+      "loss": 0.03734388,
+      "step": 2273
+    },
+    {
+      "epoch": 4.548,
+      "grad_norm": 1.7131267786026,
+      "learning_rate": 2e-05,
+      "loss": 0.050556,
+      "step": 2274
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 1.5647190809249878,
+      "learning_rate": 2e-05,
+      "loss": 0.03882205,
+      "step": 2275
+    },
+    {
+      "epoch": 4.552,
+      "grad_norm": 1.6061151027679443,
+      "learning_rate": 2e-05,
+      "loss": 0.05493605,
+      "step": 2276
+    },
+    {
+      "epoch": 4.554,
+      "grad_norm": 1.275571346282959,
+      "learning_rate": 2e-05,
+      "loss": 0.03493945,
+      "step": 2277
+    },
+    {
+      "epoch": 4.556,
+      "grad_norm": 1.767574429512024,
+      "learning_rate": 2e-05,
+      "loss": 0.03032027,
+      "step": 2278
+    },
+    {
+      "epoch": 4.558,
+      "grad_norm": 1.9915244579315186,
+      "learning_rate": 2e-05,
+      "loss": 0.04645093,
+      "step": 2279
+    },
+    {
+      "epoch": 4.5600000000000005,
+      "grad_norm": 1.441829800605774,
+      "learning_rate": 2e-05,
+      "loss": 0.03383998,
+      "step": 2280
+    },
+    {
+      "epoch": 4.562,
+      "grad_norm": 1.1743775606155396,
+      "learning_rate": 2e-05,
+      "loss": 0.03763331,
+      "step": 2281
+    },
+    {
+      "epoch": 4.564,
+      "grad_norm": 1.7172895669937134,
+      "learning_rate": 2e-05,
+      "loss": 0.0662686,
+      "step": 2282
+    },
+    {
+      "epoch": 4.566,
+      "grad_norm": 1.7259050607681274,
+      "learning_rate": 2e-05,
+      "loss": 0.03643933,
+      "step": 2283
+    },
+    {
+      "epoch": 4.568,
+      "grad_norm": 1.2674235105514526,
+      "learning_rate": 2e-05,
+      "loss": 0.03521616,
+      "step": 2284
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 1.095694661140442,
+      "learning_rate": 2e-05,
+      "loss": 0.04045701,
+      "step": 2285
+    },
+    {
+      "epoch": 4.572,
+      "grad_norm": 1.3978180885314941,
+      "learning_rate": 2e-05,
+      "loss": 0.03319689,
+      "step": 2286
+    },
+    {
+      "epoch": 4.574,
+      "grad_norm": 1.4615137577056885,
+      "learning_rate": 2e-05,
+      "loss": 0.04723348,
+      "step": 2287
+    },
+    {
+      "epoch": 4.576,
+      "grad_norm": 2.235830783843994,
+      "learning_rate": 2e-05,
+      "loss": 0.02766045,
+      "step": 2288
+    },
+    {
+      "epoch": 4.578,
+      "grad_norm": 1.0262019634246826,
+      "learning_rate": 2e-05,
+      "loss": 0.02492511,
+      "step": 2289
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 2.410766363143921,
+      "learning_rate": 2e-05,
+      "loss": 0.03837668,
+      "step": 2290
+    },
+    {
+      "epoch": 4.582,
+      "grad_norm": 1.3832848072052002,
+      "learning_rate": 2e-05,
+      "loss": 0.04708455,
+      "step": 2291
+    },
+    {
+      "epoch": 4.584,
+      "grad_norm": 1.5111563205718994,
+      "learning_rate": 2e-05,
+      "loss": 0.03212856,
+      "step": 2292
+    },
+    {
+      "epoch": 4.586,
+      "grad_norm": 1.8234524726867676,
+      "learning_rate": 2e-05,
+      "loss": 0.04155425,
+      "step": 2293
+    },
+    {
+      "epoch": 4.588,
+      "grad_norm": 1.8563976287841797,
+      "learning_rate": 2e-05,
+      "loss": 0.05284409,
+      "step": 2294
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 1.0448572635650635,
+      "learning_rate": 2e-05,
+      "loss": 0.02735084,
+      "step": 2295
+    },
+    {
+      "epoch": 4.592,
+      "grad_norm": 0.9727917909622192,
+      "learning_rate": 2e-05,
+      "loss": 0.03186521,
+      "step": 2296
+    },
+    {
+      "epoch": 4.594,
+      "grad_norm": 1.7344650030136108,
+      "learning_rate": 2e-05,
+      "loss": 0.05732023,
+      "step": 2297
+    },
+    {
+      "epoch": 4.596,
+      "grad_norm": 1.630303144454956,
+      "learning_rate": 2e-05,
+      "loss": 0.03979348,
+      "step": 2298
+    },
+    {
+      "epoch": 4.598,
+      "grad_norm": 1.5773340463638306,
+      "learning_rate": 2e-05,
+      "loss": 0.04727698,
+      "step": 2299
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.4864786863327026,
+      "learning_rate": 2e-05,
+      "loss": 0.0243044,
+      "step": 2300
+    },
+    {
+      "epoch": 4.602,
+      "grad_norm": 1.6385360956192017,
+      "learning_rate": 2e-05,
+      "loss": 0.04001735,
+      "step": 2301
+    },
+    {
+      "epoch": 4.604,
+      "grad_norm": 0.9057992696762085,
+      "learning_rate": 2e-05,
+      "loss": 0.02840018,
+      "step": 2302
+    },
+    {
+      "epoch": 4.606,
+      "grad_norm": 2.1752712726593018,
+      "learning_rate": 2e-05,
+      "loss": 0.02728005,
+      "step": 2303
+    },
+    {
+      "epoch": 4.608,
+      "grad_norm": 1.4201445579528809,
+      "learning_rate": 2e-05,
+      "loss": 0.03847256,
+      "step": 2304
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 1.6967096328735352,
+      "learning_rate": 2e-05,
+      "loss": 0.03130513,
+      "step": 2305
+    },
+    {
+      "epoch": 4.612,
+      "grad_norm": 2.1925299167633057,
+      "learning_rate": 2e-05,
+      "loss": 0.04984703,
+      "step": 2306
+    },
+    {
+      "epoch": 4.614,
+      "grad_norm": 1.6784030199050903,
+      "learning_rate": 2e-05,
+      "loss": 0.03006441,
+      "step": 2307
+    },
+    {
+      "epoch": 4.616,
+      "grad_norm": 1.3622230291366577,
+      "learning_rate": 2e-05,
+      "loss": 0.03182603,
+      "step": 2308
+    },
+    {
+      "epoch": 4.618,
+      "grad_norm": 1.5717957019805908,
+      "learning_rate": 2e-05,
+      "loss": 0.03701077,
+      "step": 2309
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 1.0062084197998047,
+      "learning_rate": 2e-05,
+      "loss": 0.0248526,
+      "step": 2310
+    },
+    {
+      "epoch": 4.622,
+      "grad_norm": 2.1822187900543213,
+      "learning_rate": 2e-05,
+      "loss": 0.03146263,
+      "step": 2311
+    },
+    {
+      "epoch": 4.624,
+      "grad_norm": 3.6029553413391113,
+      "learning_rate": 2e-05,
+      "loss": 0.06397329,
+      "step": 2312
+    },
+    {
+      "epoch": 4.626,
+      "grad_norm": 2.1496500968933105,
+      "learning_rate": 2e-05,
+      "loss": 0.03725152,
+      "step": 2313
+    },
+    {
+      "epoch": 4.628,
+      "grad_norm": 1.8497250080108643,
+      "learning_rate": 2e-05,
+      "loss": 0.03161447,
+      "step": 2314
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 1.3184853792190552,
+      "learning_rate": 2e-05,
+      "loss": 0.03692345,
+      "step": 2315
+    },
+    {
+      "epoch": 4.632,
+      "grad_norm": 1.4427586793899536,
+      "learning_rate": 2e-05,
+      "loss": 0.02692084,
+      "step": 2316
+    },
+    {
+      "epoch": 4.634,
+      "grad_norm": 1.8580230474472046,
+      "learning_rate": 2e-05,
+      "loss": 0.0654839,
+      "step": 2317
+    },
+    {
+      "epoch": 4.636,
+      "grad_norm": 1.1471973657608032,
+      "learning_rate": 2e-05,
+      "loss": 0.03153278,
+      "step": 2318
+    },
+    {
+      "epoch": 4.638,
+      "grad_norm": 1.2644822597503662,
+      "learning_rate": 2e-05,
+      "loss": 0.04941096,
+      "step": 2319
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 2.7720656394958496,
+      "learning_rate": 2e-05,
+      "loss": 0.05895547,
+      "step": 2320
+    },
+    {
+      "epoch": 4.642,
+      "grad_norm": 1.2557971477508545,
+      "learning_rate": 2e-05,
+      "loss": 0.04365244,
+      "step": 2321
+    },
+    {
+      "epoch": 4.644,
+      "grad_norm": 1.6720325946807861,
+      "learning_rate": 2e-05,
+      "loss": 0.05442552,
+      "step": 2322
+    },
+    {
+      "epoch": 4.646,
+      "grad_norm": 1.0071033239364624,
+      "learning_rate": 2e-05,
+      "loss": 0.02480928,
+      "step": 2323
+    },
+    {
+      "epoch": 4.648,
+      "grad_norm": 1.183469295501709,
+      "learning_rate": 2e-05,
+      "loss": 0.03799593,
+      "step": 2324
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 1.3475804328918457,
+      "learning_rate": 2e-05,
+      "loss": 0.05035153,
+      "step": 2325
+    },
+    {
+      "epoch": 4.652,
+      "grad_norm": 2.5962164402008057,
+      "learning_rate": 2e-05,
+      "loss": 0.04414929,
+      "step": 2326
+    },
+    {
+      "epoch": 4.654,
+      "grad_norm": 1.1905382871627808,
+      "learning_rate": 2e-05,
+      "loss": 0.04276285,
+      "step": 2327
+    },
+    {
+      "epoch": 4.656,
+      "grad_norm": 1.5951045751571655,
+      "learning_rate": 2e-05,
+      "loss": 0.03594982,
+      "step": 2328
+    },
+    {
+      "epoch": 4.658,
+      "grad_norm": 1.6772533655166626,
+      "learning_rate": 2e-05,
+      "loss": 0.05392138,
+      "step": 2329
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 1.5733336210250854,
+      "learning_rate": 2e-05,
+      "loss": 0.03919678,
+      "step": 2330
+    },
+    {
+      "epoch": 4.662,
+      "grad_norm": 1.1291429996490479,
+      "learning_rate": 2e-05,
+      "loss": 0.02948717,
+      "step": 2331
+    },
+    {
+      "epoch": 4.664,
+      "grad_norm": 1.38038969039917,
+      "learning_rate": 2e-05,
+      "loss": 0.04529754,
+      "step": 2332
+    },
+    {
+      "epoch": 4.666,
+      "grad_norm": 1.4430439472198486,
+      "learning_rate": 2e-05,
+      "loss": 0.02968345,
+      "step": 2333
+    },
+    {
+      "epoch": 4.668,
+      "grad_norm": 1.0805678367614746,
+      "learning_rate": 2e-05,
+      "loss": 0.03496836,
+      "step": 2334
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 1.4857618808746338,
+      "learning_rate": 2e-05,
+      "loss": 0.04253389,
+      "step": 2335
+    },
+    {
+      "epoch": 4.672,
+      "grad_norm": 1.8443632125854492,
+      "learning_rate": 2e-05,
+      "loss": 0.03913691,
+      "step": 2336
+    },
+    {
+      "epoch": 4.674,
+      "grad_norm": 1.596656084060669,
+      "learning_rate": 2e-05,
+      "loss": 0.03494718,
+      "step": 2337
+    },
+    {
+      "epoch": 4.676,
+      "grad_norm": 2.0523974895477295,
+      "learning_rate": 2e-05,
+      "loss": 0.05468136,
+      "step": 2338
+    },
+    {
+      "epoch": 4.678,
+      "grad_norm": 1.5666359663009644,
+      "learning_rate": 2e-05,
+      "loss": 0.0429353,
+      "step": 2339
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 1.299415946006775,
+      "learning_rate": 2e-05,
+      "loss": 0.02959892,
+      "step": 2340
+    },
+    {
+      "epoch": 4.682,
+      "grad_norm": 1.6929570436477661,
+      "learning_rate": 2e-05,
+      "loss": 0.04222934,
+      "step": 2341
+    },
+    {
+      "epoch": 4.684,
+      "grad_norm": 2.128430128097534,
+      "learning_rate": 2e-05,
+      "loss": 0.04607785,
+      "step": 2342
+    },
+    {
+      "epoch": 4.686,
+      "grad_norm": 2.121476650238037,
+      "learning_rate": 2e-05,
+      "loss": 0.04864143,
+      "step": 2343
+    },
+    {
+      "epoch": 4.688,
+      "grad_norm": 2.3001673221588135,
+      "learning_rate": 2e-05,
+      "loss": 0.04368895,
+      "step": 2344
+    },
+    {
+      "epoch": 4.6899999999999995,
+      "grad_norm": 1.804023265838623,
+      "learning_rate": 2e-05,
+      "loss": 0.03963099,
+      "step": 2345
+    },
+    {
+      "epoch": 4.692,
+      "grad_norm": 1.1703206300735474,
+      "learning_rate": 2e-05,
+      "loss": 0.03038155,
+      "step": 2346
+    },
+    {
+      "epoch": 4.694,
+      "grad_norm": 1.123602032661438,
+      "learning_rate": 2e-05,
+      "loss": 0.04947871,
+      "step": 2347
+    },
+    {
+      "epoch": 4.696,
+      "grad_norm": 1.3002231121063232,
+      "learning_rate": 2e-05,
+      "loss": 0.03265772,
+      "step": 2348
+    },
+    {
+      "epoch": 4.698,
+      "grad_norm": 1.3386176824569702,
+      "learning_rate": 2e-05,
+      "loss": 0.0319346,
+      "step": 2349
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 1.3578490018844604,
+      "learning_rate": 2e-05,
+      "loss": 0.0321762,
+      "step": 2350
+    },
+    {
+      "epoch": 4.702,
+      "grad_norm": 2.0441887378692627,
+      "learning_rate": 2e-05,
+      "loss": 0.0398183,
+      "step": 2351
+    },
+    {
+      "epoch": 4.704,
+      "grad_norm": 1.819500207901001,
+      "learning_rate": 2e-05,
+      "loss": 0.03708476,
+      "step": 2352
+    },
+    {
+      "epoch": 4.7059999999999995,
+      "grad_norm": 1.4998730421066284,
+      "learning_rate": 2e-05,
+      "loss": 0.03444158,
+      "step": 2353
+    },
+    {
+      "epoch": 4.708,
+      "grad_norm": 1.3147387504577637,
+      "learning_rate": 2e-05,
+      "loss": 0.02567411,
+      "step": 2354
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 1.0613209009170532,
+      "learning_rate": 2e-05,
+      "loss": 0.02812085,
+      "step": 2355
+    },
+    {
+      "epoch": 4.712,
+      "grad_norm": 1.6744329929351807,
+      "learning_rate": 2e-05,
+      "loss": 0.04626666,
+      "step": 2356
+    },
+    {
+      "epoch": 4.714,
+      "grad_norm": 1.8703510761260986,
+      "learning_rate": 2e-05,
+      "loss": 0.05195956,
+      "step": 2357
+    },
+    {
+      "epoch": 4.716,
+      "grad_norm": 1.7580175399780273,
+      "learning_rate": 2e-05,
+      "loss": 0.03270845,
+      "step": 2358
+    },
+    {
+      "epoch": 4.718,
+      "grad_norm": 1.9400629997253418,
+      "learning_rate": 2e-05,
+      "loss": 0.05601585,
+      "step": 2359
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 1.4302650690078735,
+      "learning_rate": 2e-05,
+      "loss": 0.03492381,
+      "step": 2360
+    },
+    {
+      "epoch": 4.7219999999999995,
+      "grad_norm": 2.32159161567688,
+      "learning_rate": 2e-05,
+      "loss": 0.04457737,
+      "step": 2361
+    },
+    {
+      "epoch": 4.724,
+      "grad_norm": 3.4112353324890137,
+      "learning_rate": 2e-05,
+      "loss": 0.08344968,
+      "step": 2362
+    },
+    {
+      "epoch": 4.726,
+      "grad_norm": 1.6217650175094604,
+      "learning_rate": 2e-05,
+      "loss": 0.02905689,
+      "step": 2363
+    },
+    {
+      "epoch": 4.728,
+      "grad_norm": 1.6228758096694946,
+      "learning_rate": 2e-05,
+      "loss": 0.04028218,
+      "step": 2364
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 1.1981812715530396,
+      "learning_rate": 2e-05,
+      "loss": 0.04487208,
+      "step": 2365
+    },
+    {
+      "epoch": 4.732,
+      "grad_norm": 1.7020169496536255,
+      "learning_rate": 2e-05,
+      "loss": 0.0399951,
+      "step": 2366
+    },
+    {
+      "epoch": 4.734,
+      "grad_norm": 1.3388174772262573,
+      "learning_rate": 2e-05,
+      "loss": 0.04234082,
+      "step": 2367
+    },
+    {
+      "epoch": 4.736,
+      "grad_norm": 1.5462074279785156,
+      "learning_rate": 2e-05,
+      "loss": 0.0329011,
+      "step": 2368
+    },
+    {
+      "epoch": 4.7379999999999995,
+      "grad_norm": 1.1013768911361694,
+      "learning_rate": 2e-05,
+      "loss": 0.03339402,
+      "step": 2369
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 1.4851659536361694,
+      "learning_rate": 2e-05,
+      "loss": 0.03826153,
+      "step": 2370
+    },
+    {
+      "epoch": 4.742,
+      "grad_norm": 1.2620880603790283,
+      "learning_rate": 2e-05,
+      "loss": 0.03477462,
+      "step": 2371
+    },
+    {
+      "epoch": 4.744,
+      "grad_norm": 1.2920024394989014,
+      "learning_rate": 2e-05,
+      "loss": 0.04607169,
+      "step": 2372
+    },
+    {
+      "epoch": 4.746,
+      "grad_norm": 1.521671175956726,
+      "learning_rate": 2e-05,
+      "loss": 0.04394786,
+      "step": 2373
+    },
+    {
+      "epoch": 4.748,
+      "grad_norm": 1.4648962020874023,
+      "learning_rate": 2e-05,
+      "loss": 0.04500762,
+      "step": 2374
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 1.8178099393844604,
+      "learning_rate": 2e-05,
+      "loss": 0.03826898,
+      "step": 2375
+    },
+    {
+      "epoch": 4.752,
+      "grad_norm": 0.9506421685218811,
+      "learning_rate": 2e-05,
+      "loss": 0.03018279,
+      "step": 2376
+    },
+    {
+      "epoch": 4.754,
+      "grad_norm": 1.1572425365447998,
+      "learning_rate": 2e-05,
+      "loss": 0.03146629,
+      "step": 2377
+    },
+    {
+      "epoch": 4.756,
+      "grad_norm": 0.9127270579338074,
+      "learning_rate": 2e-05,
+      "loss": 0.0217318,
+      "step": 2378
+    },
+    {
+      "epoch": 4.758,
+      "grad_norm": 1.051289677619934,
+      "learning_rate": 2e-05,
+      "loss": 0.03464397,
+      "step": 2379
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 2.1047472953796387,
+      "learning_rate": 2e-05,
+      "loss": 0.05494668,
+      "step": 2380
+    },
+    {
+      "epoch": 4.7620000000000005,
+      "grad_norm": 1.4690008163452148,
+      "learning_rate": 2e-05,
+      "loss": 0.02600802,
+      "step": 2381
+    },
+    {
+      "epoch": 4.764,
+      "grad_norm": 1.9520142078399658,
+      "learning_rate": 2e-05,
+      "loss": 0.03388739,
+      "step": 2382
+    },
+    {
+      "epoch": 4.766,
+      "grad_norm": 2.3827433586120605,
+      "learning_rate": 2e-05,
+      "loss": 0.0476454,
+      "step": 2383
+    },
+    {
+      "epoch": 4.768,
+      "grad_norm": 1.4814623594284058,
+      "learning_rate": 2e-05,
+      "loss": 0.03822245,
+      "step": 2384
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 2.171255350112915,
+      "learning_rate": 2e-05,
+      "loss": 0.06004865,
+      "step": 2385
+    },
+    {
+      "epoch": 4.772,
+      "grad_norm": 1.3165963888168335,
+      "learning_rate": 2e-05,
+      "loss": 0.03820898,
+      "step": 2386
+    },
+    {
+      "epoch": 4.774,
+      "grad_norm": 1.2138549089431763,
+      "learning_rate": 2e-05,
+      "loss": 0.03760416,
+      "step": 2387
+    },
+    {
+      "epoch": 4.776,
+      "grad_norm": 1.4797013998031616,
+      "learning_rate": 2e-05,
+      "loss": 0.03296691,
+      "step": 2388
+    },
+    {
+      "epoch": 4.7780000000000005,
+      "grad_norm": 1.994217038154602,
+      "learning_rate": 2e-05,
+      "loss": 0.05951263,
+      "step": 2389
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 1.6225087642669678,
+      "learning_rate": 2e-05,
+      "loss": 0.03843221,
+      "step": 2390
+    },
+    {
+      "epoch": 4.782,
+      "grad_norm": 1.264367938041687,
+      "learning_rate": 2e-05,
+      "loss": 0.03937426,
+      "step": 2391
+    },
+    {
+      "epoch": 4.784,
+      "grad_norm": 2.4877493381500244,
+      "learning_rate": 2e-05,
+      "loss": 0.04250683,
+      "step": 2392
+    },
+    {
+      "epoch": 4.786,
+      "grad_norm": 1.6016268730163574,
+      "learning_rate": 2e-05,
+      "loss": 0.03200974,
+      "step": 2393
+    },
+    {
+      "epoch": 4.788,
+      "grad_norm": 1.2733501195907593,
+      "learning_rate": 2e-05,
+      "loss": 0.02490179,
+      "step": 2394
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 2.5966386795043945,
+      "learning_rate": 2e-05,
+      "loss": 0.04996984,
+      "step": 2395
+    },
+    {
+      "epoch": 4.792,
+      "grad_norm": 1.5662956237792969,
+      "learning_rate": 2e-05,
+      "loss": 0.05331601,
+      "step": 2396
+    },
+    {
+      "epoch": 4.7940000000000005,
+      "grad_norm": 1.7879033088684082,
+      "learning_rate": 2e-05,
+      "loss": 0.04639386,
+      "step": 2397
+    },
+    {
+      "epoch": 4.796,
+      "grad_norm": 1.5951218605041504,
+      "learning_rate": 2e-05,
+      "loss": 0.04075851,
+      "step": 2398
+    },
+    {
+      "epoch": 4.798,
+      "grad_norm": 1.3353394269943237,
+      "learning_rate": 2e-05,
+      "loss": 0.03567629,
+      "step": 2399
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.5784703493118286,
+      "learning_rate": 2e-05,
+      "loss": 0.0398753,
+      "step": 2400
+    },
+    {
+      "epoch": 4.802,
+      "grad_norm": 1.7636374235153198,
+      "learning_rate": 2e-05,
+      "loss": 0.03577316,
+      "step": 2401
+    },
+    {
+      "epoch": 4.804,
+      "grad_norm": 2.104825496673584,
+      "learning_rate": 2e-05,
+      "loss": 0.05970087,
+      "step": 2402
+    },
+    {
+      "epoch": 4.806,
+      "grad_norm": 1.0066531896591187,
+      "learning_rate": 2e-05,
+      "loss": 0.03378828,
+      "step": 2403
+    },
+    {
+      "epoch": 4.808,
+      "grad_norm": 1.1586098670959473,
+      "learning_rate": 2e-05,
+      "loss": 0.03477071,
+      "step": 2404
+    },
+    {
+      "epoch": 4.8100000000000005,
+      "grad_norm": 1.9183125495910645,
+      "learning_rate": 2e-05,
+      "loss": 0.04798802,
+      "step": 2405
+    },
+    {
+      "epoch": 4.812,
+      "grad_norm": 1.2498745918273926,
+      "learning_rate": 2e-05,
+      "loss": 0.04589451,
+      "step": 2406
+    },
+    {
+      "epoch": 4.814,
+      "grad_norm": 2.4411637783050537,
+      "learning_rate": 2e-05,
+      "loss": 0.06537707,
+      "step": 2407
+    },
+    {
+      "epoch": 4.816,
+      "grad_norm": 1.692878007888794,
+      "learning_rate": 2e-05,
+      "loss": 0.02775835,
+      "step": 2408
+    },
+    {
+      "epoch": 4.818,
+      "grad_norm": 1.5488640069961548,
+      "learning_rate": 2e-05,
+      "loss": 0.04157056,
+      "step": 2409
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 1.456628680229187,
+      "learning_rate": 2e-05,
+      "loss": 0.04453836,
+      "step": 2410
+    },
+    {
+      "epoch": 4.822,
+      "grad_norm": 1.2620590925216675,
+      "learning_rate": 2e-05,
+      "loss": 0.04159302,
+      "step": 2411
+    },
+    {
+      "epoch": 4.824,
+      "grad_norm": 1.6489261388778687,
+      "learning_rate": 2e-05,
+      "loss": 0.03650364,
+      "step": 2412
+    },
+    {
+      "epoch": 4.826,
+      "grad_norm": 1.3085451126098633,
+      "learning_rate": 2e-05,
+      "loss": 0.03413782,
+      "step": 2413
+    },
+    {
+      "epoch": 4.828,
+      "grad_norm": 1.924852967262268,
+      "learning_rate": 2e-05,
+      "loss": 0.06830163,
+      "step": 2414
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 1.020704746246338,
+      "learning_rate": 2e-05,
+      "loss": 0.02394637,
+      "step": 2415
+    },
+    {
+      "epoch": 4.832,
+      "grad_norm": 1.132530689239502,
+      "learning_rate": 2e-05,
+      "loss": 0.0420993,
+      "step": 2416
+    },
+    {
+      "epoch": 4.834,
+      "grad_norm": 0.9570915102958679,
+      "learning_rate": 2e-05,
+      "loss": 0.0311062,
+      "step": 2417
+    },
+    {
+      "epoch": 4.836,
+      "grad_norm": 1.0534124374389648,
+      "learning_rate": 2e-05,
+      "loss": 0.02990963,
+      "step": 2418
+    },
+    {
+      "epoch": 4.838,
+      "grad_norm": 1.1117393970489502,
+      "learning_rate": 2e-05,
+      "loss": 0.03667744,
+      "step": 2419
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 1.32503342628479,
+      "learning_rate": 2e-05,
+      "loss": 0.03586256,
+      "step": 2420
+    },
+    {
+      "epoch": 4.842,
+      "grad_norm": 1.5087800025939941,
+      "learning_rate": 2e-05,
+      "loss": 0.03501411,
+      "step": 2421
+    },
+    {
+      "epoch": 4.844,
+      "grad_norm": 1.7611730098724365,
+      "learning_rate": 2e-05,
+      "loss": 0.03907187,
+      "step": 2422
+    },
+    {
+      "epoch": 4.846,
+      "grad_norm": 1.3522123098373413,
+      "learning_rate": 2e-05,
+      "loss": 0.04923314,
+      "step": 2423
+    },
+    {
+      "epoch": 4.848,
+      "grad_norm": 1.3680315017700195,
+      "learning_rate": 2e-05,
+      "loss": 0.03453797,
+      "step": 2424
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 1.0860910415649414,
+      "learning_rate": 2e-05,
+      "loss": 0.03305909,
+      "step": 2425
+    },
+    {
+      "epoch": 4.852,
+      "grad_norm": 1.2761231660842896,
+      "learning_rate": 2e-05,
+      "loss": 0.03872238,
+      "step": 2426
+    },
+    {
+      "epoch": 4.854,
+      "grad_norm": 1.7196176052093506,
+      "learning_rate": 2e-05,
+      "loss": 0.03280783,
+      "step": 2427
+    },
+    {
+      "epoch": 4.856,
+      "grad_norm": 1.5164096355438232,
+      "learning_rate": 2e-05,
+      "loss": 0.03744507,
+      "step": 2428
+    },
+    {
+      "epoch": 4.858,
+      "grad_norm": 1.294891119003296,
+      "learning_rate": 2e-05,
+      "loss": 0.03750362,
+      "step": 2429
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 1.5655124187469482,
+      "learning_rate": 2e-05,
+      "loss": 0.03739174,
+      "step": 2430
+    },
+    {
+      "epoch": 4.862,
+      "grad_norm": 1.806402325630188,
+      "learning_rate": 2e-05,
+      "loss": 0.04592914,
+      "step": 2431
+    },
+    {
+      "epoch": 4.864,
+      "grad_norm": 2.5879619121551514,
+      "learning_rate": 2e-05,
+      "loss": 0.03984898,
+      "step": 2432
+    },
+    {
+      "epoch": 4.866,
+      "grad_norm": 2.590233564376831,
+      "learning_rate": 2e-05,
+      "loss": 0.03624843,
+      "step": 2433
+    },
+    {
+      "epoch": 4.868,
+      "grad_norm": 1.4250051975250244,
+      "learning_rate": 2e-05,
+      "loss": 0.06627299,
+      "step": 2434
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 2.142298936843872,
+      "learning_rate": 2e-05,
+      "loss": 0.04694505,
+      "step": 2435
+    },
+    {
+      "epoch": 4.872,
+      "grad_norm": 1.6512330770492554,
+      "learning_rate": 2e-05,
+      "loss": 0.04675486,
+      "step": 2436
+    },
+    {
+      "epoch": 4.874,
+      "grad_norm": 1.1466740369796753,
+      "learning_rate": 2e-05,
+      "loss": 0.0404415,
+      "step": 2437
+    },
+    {
+      "epoch": 4.876,
+      "grad_norm": 1.163206934928894,
+      "learning_rate": 2e-05,
+      "loss": 0.04484642,
+      "step": 2438
+    },
+    {
+      "epoch": 4.878,
+      "grad_norm": 1.4623184204101562,
+      "learning_rate": 2e-05,
+      "loss": 0.0501948,
+      "step": 2439
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 1.4037455320358276,
+      "learning_rate": 2e-05,
+      "loss": 0.03190558,
+      "step": 2440
+    },
+    {
+      "epoch": 4.882,
+      "grad_norm": 1.6567379236221313,
+      "learning_rate": 2e-05,
+      "loss": 0.03677463,
+      "step": 2441
+    },
+    {
+      "epoch": 4.884,
+      "grad_norm": 1.5126616954803467,
+      "learning_rate": 2e-05,
+      "loss": 0.03623777,
+      "step": 2442
+    },
+    {
+      "epoch": 4.886,
+      "grad_norm": 1.434916615486145,
+      "learning_rate": 2e-05,
+      "loss": 0.04461944,
+      "step": 2443
+    },
+    {
+      "epoch": 4.888,
+      "grad_norm": 1.0425275564193726,
+      "learning_rate": 2e-05,
+      "loss": 0.03600231,
+      "step": 2444
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 1.3526657819747925,
+      "learning_rate": 2e-05,
+      "loss": 0.05250793,
+      "step": 2445
+    },
+    {
+      "epoch": 4.892,
+      "grad_norm": 2.3579368591308594,
+      "learning_rate": 2e-05,
+      "loss": 0.04716105,
+      "step": 2446
+    },
+    {
+      "epoch": 4.894,
+      "grad_norm": 2.2220516204833984,
+      "learning_rate": 2e-05,
+      "loss": 0.04456067,
+      "step": 2447
+    },
+    {
+      "epoch": 4.896,
+      "grad_norm": 1.631173014640808,
+      "learning_rate": 2e-05,
+      "loss": 0.04533452,
+      "step": 2448
+    },
+    {
+      "epoch": 4.898,
+      "grad_norm": 1.5379027128219604,
+      "learning_rate": 2e-05,
+      "loss": 0.03291882,
+      "step": 2449
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 1.6118535995483398,
+      "learning_rate": 2e-05,
+      "loss": 0.05840709,
+      "step": 2450
+    },
+    {
+      "epoch": 4.902,
+      "grad_norm": 1.0765273571014404,
+      "learning_rate": 2e-05,
+      "loss": 0.03540622,
+      "step": 2451
+    },
+    {
+      "epoch": 4.904,
+      "grad_norm": 1.0575835704803467,
+      "learning_rate": 2e-05,
+      "loss": 0.02790409,
+      "step": 2452
+    },
+    {
+      "epoch": 4.906,
+      "grad_norm": 0.8009814023971558,
+      "learning_rate": 2e-05,
+      "loss": 0.02813254,
+      "step": 2453
+    },
+    {
+      "epoch": 4.908,
+      "grad_norm": 1.1870671510696411,
+      "learning_rate": 2e-05,
+      "loss": 0.0324744,
+      "step": 2454
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 1.583187460899353,
+      "learning_rate": 2e-05,
+      "loss": 0.0449949,
+      "step": 2455
+    },
+    {
+      "epoch": 4.912,
+      "grad_norm": 0.8084712028503418,
+      "learning_rate": 2e-05,
+      "loss": 0.02643416,
+      "step": 2456
+    },
+    {
+      "epoch": 4.914,
+      "grad_norm": 1.0200517177581787,
+      "learning_rate": 2e-05,
+      "loss": 0.02963091,
+      "step": 2457
+    },
+    {
+      "epoch": 4.916,
+      "grad_norm": 1.270112156867981,
+      "learning_rate": 2e-05,
+      "loss": 0.04167578,
+      "step": 2458
+    },
+    {
+      "epoch": 4.918,
+      "grad_norm": 1.7084966897964478,
+      "learning_rate": 2e-05,
+      "loss": 0.0530615,
+      "step": 2459
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.9616237282752991,
+      "learning_rate": 2e-05,
+      "loss": 0.03146474,
+      "step": 2460
+    },
+    {
+      "epoch": 4.922,
+      "grad_norm": 1.698949933052063,
+      "learning_rate": 2e-05,
+      "loss": 0.04134315,
+      "step": 2461
+    },
+    {
+      "epoch": 4.924,
+      "grad_norm": 1.3575299978256226,
+      "learning_rate": 2e-05,
+      "loss": 0.04107556,
+      "step": 2462
+    },
+    {
+      "epoch": 4.926,
+      "grad_norm": 1.2565886974334717,
+      "learning_rate": 2e-05,
+      "loss": 0.02181477,
+      "step": 2463
+    },
+    {
+      "epoch": 4.928,
+      "grad_norm": 0.7854299545288086,
+      "learning_rate": 2e-05,
+      "loss": 0.0208516,
+      "step": 2464
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 1.0150772333145142,
+      "learning_rate": 2e-05,
+      "loss": 0.03809316,
+      "step": 2465
+    },
+    {
+      "epoch": 4.932,
+      "grad_norm": 1.4074971675872803,
+      "learning_rate": 2e-05,
+      "loss": 0.04282606,
+      "step": 2466
+    },
+    {
+      "epoch": 4.934,
+      "grad_norm": 1.378995418548584,
+      "learning_rate": 2e-05,
+      "loss": 0.0498643,
+      "step": 2467
+    },
+    {
+      "epoch": 4.936,
+      "grad_norm": 1.3473279476165771,
+      "learning_rate": 2e-05,
+      "loss": 0.04881741,
+      "step": 2468
+    },
+    {
+      "epoch": 4.938,
+      "grad_norm": 1.5778177976608276,
+      "learning_rate": 2e-05,
+      "loss": 0.05556468,
+      "step": 2469
+    },
+    {
+      "epoch": 4.9399999999999995,
+      "grad_norm": 1.3268343210220337,
+      "learning_rate": 2e-05,
+      "loss": 0.04513773,
+      "step": 2470
+    },
+    {
+      "epoch": 4.942,
+      "grad_norm": 1.551581621170044,
+      "learning_rate": 2e-05,
+      "loss": 0.04969638,
+      "step": 2471
+    },
+    {
+      "epoch": 4.944,
+      "grad_norm": 1.280708909034729,
+      "learning_rate": 2e-05,
+      "loss": 0.04666021,
+      "step": 2472
+    },
+    {
+      "epoch": 4.946,
+      "grad_norm": 1.6760709285736084,
+      "learning_rate": 2e-05,
+      "loss": 0.04832494,
+      "step": 2473
+    },
+    {
+      "epoch": 4.948,
+      "grad_norm": 1.4590950012207031,
+      "learning_rate": 2e-05,
+      "loss": 0.05018194,
+      "step": 2474
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 1.3153926134109497,
+      "learning_rate": 2e-05,
+      "loss": 0.03609195,
+      "step": 2475
+    },
+    {
+      "epoch": 4.952,
+      "grad_norm": 2.7595255374908447,
+      "learning_rate": 2e-05,
+      "loss": 0.02830896,
+      "step": 2476
+    },
+    {
+      "epoch": 4.954,
+      "grad_norm": 1.0296415090560913,
+      "learning_rate": 2e-05,
+      "loss": 0.02910339,
+      "step": 2477
+    },
+    {
+      "epoch": 4.9559999999999995,
+      "grad_norm": 1.205686330795288,
+      "learning_rate": 2e-05,
+      "loss": 0.02493732,
+      "step": 2478
+    },
+    {
+      "epoch": 4.958,
+      "grad_norm": 1.3554532527923584,
+      "learning_rate": 2e-05,
+      "loss": 0.03051371,
+      "step": 2479
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 1.8102985620498657,
+      "learning_rate": 2e-05,
+      "loss": 0.05402344,
+      "step": 2480
+    },
+    {
+      "epoch": 4.962,
+      "grad_norm": 1.090096116065979,
+      "learning_rate": 2e-05,
+      "loss": 0.04083726,
+      "step": 2481
+    },
+    {
+      "epoch": 4.964,
+      "grad_norm": 1.3355975151062012,
+      "learning_rate": 2e-05,
+      "loss": 0.0298448,
+      "step": 2482
+    },
+    {
+      "epoch": 4.966,
+      "grad_norm": 1.1990944147109985,
+      "learning_rate": 2e-05,
+      "loss": 0.04024744,
+      "step": 2483
+    },
+    {
+      "epoch": 4.968,
+      "grad_norm": 1.599819540977478,
+      "learning_rate": 2e-05,
+      "loss": 0.04165243,
+      "step": 2484
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 1.154665470123291,
+      "learning_rate": 2e-05,
+      "loss": 0.03814416,
+      "step": 2485
+    },
+    {
+      "epoch": 4.9719999999999995,
+      "grad_norm": 1.7538753747940063,
+      "learning_rate": 2e-05,
+      "loss": 0.04323652,
+      "step": 2486
+    },
+    {
+      "epoch": 4.974,
+      "grad_norm": 1.2068040370941162,
+      "learning_rate": 2e-05,
+      "loss": 0.02646199,
+      "step": 2487
+    },
+    {
+      "epoch": 4.976,
+      "grad_norm": 1.9645920991897583,
+      "learning_rate": 2e-05,
+      "loss": 0.04496079,
+      "step": 2488
+    },
+    {
+      "epoch": 4.978,
+      "grad_norm": 1.4738874435424805,
+      "learning_rate": 2e-05,
+      "loss": 0.04259995,
+      "step": 2489
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 1.1883742809295654,
+      "learning_rate": 2e-05,
+      "loss": 0.02651443,
+      "step": 2490
+    },
+    {
+      "epoch": 4.982,
+      "grad_norm": 1.4614838361740112,
+      "learning_rate": 2e-05,
+      "loss": 0.0321265,
+      "step": 2491
+    },
+    {
+      "epoch": 4.984,
+      "grad_norm": 2.184630870819092,
+      "learning_rate": 2e-05,
+      "loss": 0.03757413,
+      "step": 2492
+    },
+    {
+      "epoch": 4.986,
+      "grad_norm": 1.9842206239700317,
+      "learning_rate": 2e-05,
+      "loss": 0.05184003,
+      "step": 2493
+    },
+    {
+      "epoch": 4.9879999999999995,
+      "grad_norm": 3.535095453262329,
+      "learning_rate": 2e-05,
+      "loss": 0.04131281,
+      "step": 2494
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 2.6788275241851807,
+      "learning_rate": 2e-05,
+      "loss": 0.04076681,
+      "step": 2495
+    },
+    {
+      "epoch": 4.992,
+      "grad_norm": 1.9538054466247559,
+      "learning_rate": 2e-05,
+      "loss": 0.03534822,
+      "step": 2496
+    },
+    {
+      "epoch": 4.994,
+      "grad_norm": 2.29728364944458,
+      "learning_rate": 2e-05,
+      "loss": 0.0576341,
+      "step": 2497
+    },
+    {
+      "epoch": 4.996,
+      "grad_norm": 1.8153451681137085,
+      "learning_rate": 2e-05,
+      "loss": 0.04025466,
+      "step": 2498
+    },
+    {
+      "epoch": 4.998,
+      "grad_norm": 2.0112109184265137,
+      "learning_rate": 2e-05,
+      "loss": 0.05562697,
+      "step": 2499
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.1833168268203735,
+      "learning_rate": 2e-05,
+      "loss": 0.04619126,
+      "step": 2500
+    },
+    {
+      "epoch": 5.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.988,
+        "AngleClassification_2": 0.99,
+        "AngleClassification_3": 0.7265469061876247,
+        "Equal_1": 0.956,
+        "Equal_2": 0.8602794411177644,
+        "Equal_3": 0.7544910179640718,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 0.9900199600798403,
+        "LineComparison_3": 0.9481037924151696,
+        "Parallel_1": 0.9859719438877755,
+        "Parallel_2": 0.9979959919839679,
+        "Parallel_3": 0.9,
+        "Perpendicular_1": 0.974,
+        "Perpendicular_2": 0.638,
+        "Perpendicular_3": 0.38577154308617234,
+        "PointLiesOnCircle_1": 0.9966599866399466,
+        "PointLiesOnCircle_2": 0.9953333333333334,
+        "PointLiesOnCircle_3": 0.9768,
+        "PointLiesOnLine_1": 0.9939879759519038,
+        "PointLiesOnLine_2": 0.9819639278557114,
+        "PointLiesOnLine_3": 0.6207584830339321
+      },
+      "eval_runtime": 320.4044,
+      "eval_samples_per_second": 32.771,
+      "eval_steps_per_second": 0.655,
+      "step": 2500
+    },
+    {
+      "epoch": 5.002,
+      "grad_norm": 2.2324397563934326,
+      "learning_rate": 2e-05,
+      "loss": 0.05695269,
+      "step": 2501
+    },
+    {
+      "epoch": 5.004,
+      "grad_norm": 1.6174829006195068,
+      "learning_rate": 2e-05,
+      "loss": 0.05285802,
+      "step": 2502
+    },
+    {
+      "epoch": 5.006,
+      "grad_norm": 2.227994680404663,
+      "learning_rate": 2e-05,
+      "loss": 0.08210929,
+      "step": 2503
+    },
+    {
+      "epoch": 5.008,
+      "grad_norm": 2.022165298461914,
+      "learning_rate": 2e-05,
+      "loss": 0.0481383,
+      "step": 2504
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 1.2331695556640625,
+      "learning_rate": 2e-05,
+      "loss": 0.03883846,
+      "step": 2505
+    },
+    {
+      "epoch": 5.012,
+      "grad_norm": 2.4110569953918457,
+      "learning_rate": 2e-05,
+      "loss": 0.0675078,
+      "step": 2506
+    },
+    {
+      "epoch": 5.014,
+      "grad_norm": 1.4140784740447998,
+      "learning_rate": 2e-05,
+      "loss": 0.02794993,
+      "step": 2507
+    },
+    {
+      "epoch": 5.016,
+      "grad_norm": 1.0707343816757202,
+      "learning_rate": 2e-05,
+      "loss": 0.02114813,
+      "step": 2508
+    },
+    {
+      "epoch": 5.018,
+      "grad_norm": 1.7338379621505737,
+      "learning_rate": 2e-05,
+      "loss": 0.04929041,
+      "step": 2509
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 2.7071118354797363,
+      "learning_rate": 2e-05,
+      "loss": 0.06345211,
+      "step": 2510
+    },
+    {
+      "epoch": 5.022,
+      "grad_norm": 1.2723110914230347,
+      "learning_rate": 2e-05,
+      "loss": 0.04290574,
+      "step": 2511
+    },
+    {
+      "epoch": 5.024,
+      "grad_norm": 2.29909348487854,
+      "learning_rate": 2e-05,
+      "loss": 0.05433185,
+      "step": 2512
+    },
+    {
+      "epoch": 5.026,
+      "grad_norm": 1.5173856019973755,
+      "learning_rate": 2e-05,
+      "loss": 0.03184156,
+      "step": 2513
+    },
+    {
+      "epoch": 5.028,
+      "grad_norm": 1.4544750452041626,
+      "learning_rate": 2e-05,
+      "loss": 0.03113187,
+      "step": 2514
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 1.0467116832733154,
+      "learning_rate": 2e-05,
+      "loss": 0.02658311,
+      "step": 2515
+    },
+    {
+      "epoch": 5.032,
+      "grad_norm": 1.4478943347930908,
+      "learning_rate": 2e-05,
+      "loss": 0.04346019,
+      "step": 2516
+    },
+    {
+      "epoch": 5.034,
+      "grad_norm": 1.587401270866394,
+      "learning_rate": 2e-05,
+      "loss": 0.04132868,
+      "step": 2517
+    },
+    {
+      "epoch": 5.036,
+      "grad_norm": 1.247644305229187,
+      "learning_rate": 2e-05,
+      "loss": 0.03109686,
+      "step": 2518
+    },
+    {
+      "epoch": 5.038,
+      "grad_norm": 1.2669824361801147,
+      "learning_rate": 2e-05,
+      "loss": 0.02979152,
+      "step": 2519
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 2.6262893676757812,
+      "learning_rate": 2e-05,
+      "loss": 0.06778932,
+      "step": 2520
+    },
+    {
+      "epoch": 5.042,
+      "grad_norm": 1.6838678121566772,
+      "learning_rate": 2e-05,
+      "loss": 0.06396025,
+      "step": 2521
+    },
+    {
+      "epoch": 5.044,
+      "grad_norm": 1.3168069124221802,
+      "learning_rate": 2e-05,
+      "loss": 0.02535046,
+      "step": 2522
+    },
+    {
+      "epoch": 5.046,
+      "grad_norm": 2.7136900424957275,
+      "learning_rate": 2e-05,
+      "loss": 0.04555402,
+      "step": 2523
+    },
+    {
+      "epoch": 5.048,
+      "grad_norm": 1.461851716041565,
+      "learning_rate": 2e-05,
+      "loss": 0.03833253,
+      "step": 2524
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 1.8288956880569458,
+      "learning_rate": 2e-05,
+      "loss": 0.03891191,
+      "step": 2525
+    },
+    {
+      "epoch": 5.052,
+      "grad_norm": 1.1910473108291626,
+      "learning_rate": 2e-05,
+      "loss": 0.03411533,
+      "step": 2526
+    },
+    {
+      "epoch": 5.054,
+      "grad_norm": 1.5257526636123657,
+      "learning_rate": 2e-05,
+      "loss": 0.04626542,
+      "step": 2527
+    },
+    {
+      "epoch": 5.056,
+      "grad_norm": 1.2953051328659058,
+      "learning_rate": 2e-05,
+      "loss": 0.02808543,
+      "step": 2528
+    },
+    {
+      "epoch": 5.058,
+      "grad_norm": 1.4993410110473633,
+      "learning_rate": 2e-05,
+      "loss": 0.04547715,
+      "step": 2529
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 2.6157491207122803,
+      "learning_rate": 2e-05,
+      "loss": 0.09757001,
+      "step": 2530
+    },
+    {
+      "epoch": 5.062,
+      "grad_norm": 2.16626238822937,
+      "learning_rate": 2e-05,
+      "loss": 0.05185439,
+      "step": 2531
+    },
+    {
+      "epoch": 5.064,
+      "grad_norm": 5.067115783691406,
+      "learning_rate": 2e-05,
+      "loss": 0.04093373,
+      "step": 2532
+    },
+    {
+      "epoch": 5.066,
+      "grad_norm": 3.5406131744384766,
+      "learning_rate": 2e-05,
+      "loss": 0.05016257,
+      "step": 2533
+    },
+    {
+      "epoch": 5.068,
+      "grad_norm": 1.9250370264053345,
+      "learning_rate": 2e-05,
+      "loss": 0.04495588,
+      "step": 2534
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 1.3364894390106201,
+      "learning_rate": 2e-05,
+      "loss": 0.05048854,
+      "step": 2535
+    },
+    {
+      "epoch": 5.072,
+      "grad_norm": 1.1228950023651123,
+      "learning_rate": 2e-05,
+      "loss": 0.03420167,
+      "step": 2536
+    },
+    {
+      "epoch": 5.074,
+      "grad_norm": 2.442943572998047,
+      "learning_rate": 2e-05,
+      "loss": 0.04751071,
+      "step": 2537
+    },
+    {
+      "epoch": 5.076,
+      "grad_norm": 1.1549925804138184,
+      "learning_rate": 2e-05,
+      "loss": 0.03477682,
+      "step": 2538
+    },
+    {
+      "epoch": 5.078,
+      "grad_norm": 1.890607237815857,
+      "learning_rate": 2e-05,
+      "loss": 0.04445625,
+      "step": 2539
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 1.524712324142456,
+      "learning_rate": 2e-05,
+      "loss": 0.04532345,
+      "step": 2540
+    },
+    {
+      "epoch": 5.082,
+      "grad_norm": 1.9938740730285645,
+      "learning_rate": 2e-05,
+      "loss": 0.04844673,
+      "step": 2541
+    },
+    {
+      "epoch": 5.084,
+      "grad_norm": 2.3578040599823,
+      "learning_rate": 2e-05,
+      "loss": 0.06914338,
+      "step": 2542
+    },
+    {
+      "epoch": 5.086,
+      "grad_norm": 1.6852574348449707,
+      "learning_rate": 2e-05,
+      "loss": 0.03590863,
+      "step": 2543
+    },
+    {
+      "epoch": 5.088,
+      "grad_norm": 1.804057240486145,
+      "learning_rate": 2e-05,
+      "loss": 0.0503867,
+      "step": 2544
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 1.832448959350586,
+      "learning_rate": 2e-05,
+      "loss": 0.05448397,
+      "step": 2545
+    },
+    {
+      "epoch": 5.092,
+      "grad_norm": 1.6885632276535034,
+      "learning_rate": 2e-05,
+      "loss": 0.0470449,
+      "step": 2546
+    },
+    {
+      "epoch": 5.094,
+      "grad_norm": 0.9903779625892639,
+      "learning_rate": 2e-05,
+      "loss": 0.03127839,
+      "step": 2547
+    },
+    {
+      "epoch": 5.096,
+      "grad_norm": 1.0127002000808716,
+      "learning_rate": 2e-05,
+      "loss": 0.03280874,
+      "step": 2548
+    },
+    {
+      "epoch": 5.098,
+      "grad_norm": 1.8620246648788452,
+      "learning_rate": 2e-05,
+      "loss": 0.06030571,
+      "step": 2549
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 1.3962090015411377,
+      "learning_rate": 2e-05,
+      "loss": 0.04562517,
+      "step": 2550
+    },
+    {
+      "epoch": 5.102,
+      "grad_norm": 1.8518095016479492,
+      "learning_rate": 2e-05,
+      "loss": 0.06386492,
+      "step": 2551
+    },
+    {
+      "epoch": 5.104,
+      "grad_norm": 1.2915924787521362,
+      "learning_rate": 2e-05,
+      "loss": 0.0317604,
+      "step": 2552
+    },
+    {
+      "epoch": 5.106,
+      "grad_norm": 1.6023142337799072,
+      "learning_rate": 2e-05,
+      "loss": 0.03814955,
+      "step": 2553
+    },
+    {
+      "epoch": 5.108,
+      "grad_norm": 1.1027803421020508,
+      "learning_rate": 2e-05,
+      "loss": 0.02706053,
+      "step": 2554
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 2.314110517501831,
+      "learning_rate": 2e-05,
+      "loss": 0.0609056,
+      "step": 2555
+    },
+    {
+      "epoch": 5.112,
+      "grad_norm": 1.028228759765625,
+      "learning_rate": 2e-05,
+      "loss": 0.02673756,
+      "step": 2556
+    },
+    {
+      "epoch": 5.114,
+      "grad_norm": 1.5030735731124878,
+      "learning_rate": 2e-05,
+      "loss": 0.0449588,
+      "step": 2557
+    },
+    {
+      "epoch": 5.116,
+      "grad_norm": 1.3208764791488647,
+      "learning_rate": 2e-05,
+      "loss": 0.04690504,
+      "step": 2558
+    },
+    {
+      "epoch": 5.118,
+      "grad_norm": 2.08263897895813,
+      "learning_rate": 2e-05,
+      "loss": 0.04718242,
+      "step": 2559
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 1.488623857498169,
+      "learning_rate": 2e-05,
+      "loss": 0.04909895,
+      "step": 2560
+    },
+    {
+      "epoch": 5.122,
+      "grad_norm": 1.8604114055633545,
+      "learning_rate": 2e-05,
+      "loss": 0.04605662,
+      "step": 2561
+    },
+    {
+      "epoch": 5.124,
+      "grad_norm": 1.0743006467819214,
+      "learning_rate": 2e-05,
+      "loss": 0.02543287,
+      "step": 2562
+    },
+    {
+      "epoch": 5.126,
+      "grad_norm": 1.3508437871932983,
+      "learning_rate": 2e-05,
+      "loss": 0.03909838,
+      "step": 2563
+    },
+    {
+      "epoch": 5.128,
+      "grad_norm": 1.4261709451675415,
+      "learning_rate": 2e-05,
+      "loss": 0.03505434,
+      "step": 2564
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 2.1459856033325195,
+      "learning_rate": 2e-05,
+      "loss": 0.05590709,
+      "step": 2565
+    },
+    {
+      "epoch": 5.132,
+      "grad_norm": 2.010373115539551,
+      "learning_rate": 2e-05,
+      "loss": 0.03756604,
+      "step": 2566
+    },
+    {
+      "epoch": 5.134,
+      "grad_norm": 1.6585365533828735,
+      "learning_rate": 2e-05,
+      "loss": 0.03042065,
+      "step": 2567
+    },
+    {
+      "epoch": 5.136,
+      "grad_norm": 2.4542722702026367,
+      "learning_rate": 2e-05,
+      "loss": 0.05523026,
+      "step": 2568
+    },
+    {
+      "epoch": 5.138,
+      "grad_norm": 1.9174667596817017,
+      "learning_rate": 2e-05,
+      "loss": 0.06206827,
+      "step": 2569
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 1.7819212675094604,
+      "learning_rate": 2e-05,
+      "loss": 0.0496488,
+      "step": 2570
+    },
+    {
+      "epoch": 5.142,
+      "grad_norm": 2.3288769721984863,
+      "learning_rate": 2e-05,
+      "loss": 0.05924601,
+      "step": 2571
+    },
+    {
+      "epoch": 5.144,
+      "grad_norm": 4.057807922363281,
+      "learning_rate": 2e-05,
+      "loss": 0.0507498,
+      "step": 2572
+    },
+    {
+      "epoch": 5.146,
+      "grad_norm": 1.5712443590164185,
+      "learning_rate": 2e-05,
+      "loss": 0.05386955,
+      "step": 2573
+    },
+    {
+      "epoch": 5.148,
+      "grad_norm": 1.6841580867767334,
+      "learning_rate": 2e-05,
+      "loss": 0.0409927,
+      "step": 2574
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 2.0877368450164795,
+      "learning_rate": 2e-05,
+      "loss": 0.03837425,
+      "step": 2575
+    },
+    {
+      "epoch": 5.152,
+      "grad_norm": 1.4437499046325684,
+      "learning_rate": 2e-05,
+      "loss": 0.04897504,
+      "step": 2576
+    },
+    {
+      "epoch": 5.154,
+      "grad_norm": 1.4186006784439087,
+      "learning_rate": 2e-05,
+      "loss": 0.05277944,
+      "step": 2577
+    },
+    {
+      "epoch": 5.156,
+      "grad_norm": 6.804627418518066,
+      "learning_rate": 2e-05,
+      "loss": 0.05949213,
+      "step": 2578
+    },
+    {
+      "epoch": 5.158,
+      "grad_norm": 1.6603080034255981,
+      "learning_rate": 2e-05,
+      "loss": 0.05066625,
+      "step": 2579
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 1.7450742721557617,
+      "learning_rate": 2e-05,
+      "loss": 0.05555181,
+      "step": 2580
+    },
+    {
+      "epoch": 5.162,
+      "grad_norm": 2.0960774421691895,
+      "learning_rate": 2e-05,
+      "loss": 0.05067931,
+      "step": 2581
+    },
+    {
+      "epoch": 5.164,
+      "grad_norm": 1.5471669435501099,
+      "learning_rate": 2e-05,
+      "loss": 0.04380035,
+      "step": 2582
+    },
+    {
+      "epoch": 5.166,
+      "grad_norm": 1.5839829444885254,
+      "learning_rate": 2e-05,
+      "loss": 0.0428129,
+      "step": 2583
+    },
+    {
+      "epoch": 5.168,
+      "grad_norm": 1.1979819536209106,
+      "learning_rate": 2e-05,
+      "loss": 0.03231291,
+      "step": 2584
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 0.9489485621452332,
+      "learning_rate": 2e-05,
+      "loss": 0.03189579,
+      "step": 2585
+    },
+    {
+      "epoch": 5.172,
+      "grad_norm": 1.3143651485443115,
+      "learning_rate": 2e-05,
+      "loss": 0.03099883,
+      "step": 2586
+    },
+    {
+      "epoch": 5.174,
+      "grad_norm": 1.7608791589736938,
+      "learning_rate": 2e-05,
+      "loss": 0.06527974,
+      "step": 2587
+    },
+    {
+      "epoch": 5.176,
+      "grad_norm": 1.6358402967453003,
+      "learning_rate": 2e-05,
+      "loss": 0.05212991,
+      "step": 2588
+    },
+    {
+      "epoch": 5.178,
+      "grad_norm": 1.2819510698318481,
+      "learning_rate": 2e-05,
+      "loss": 0.03158101,
+      "step": 2589
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 1.5717135667800903,
+      "learning_rate": 2e-05,
+      "loss": 0.04475843,
+      "step": 2590
+    },
+    {
+      "epoch": 5.182,
+      "grad_norm": 2.0249385833740234,
+      "learning_rate": 2e-05,
+      "loss": 0.05886701,
+      "step": 2591
+    },
+    {
+      "epoch": 5.184,
+      "grad_norm": 1.782658338546753,
+      "learning_rate": 2e-05,
+      "loss": 0.04498136,
+      "step": 2592
+    },
+    {
+      "epoch": 5.186,
+      "grad_norm": 1.6862472295761108,
+      "learning_rate": 2e-05,
+      "loss": 0.03234654,
+      "step": 2593
+    },
+    {
+      "epoch": 5.188,
+      "grad_norm": 1.2082955837249756,
+      "learning_rate": 2e-05,
+      "loss": 0.04066747,
+      "step": 2594
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 1.7842987775802612,
+      "learning_rate": 2e-05,
+      "loss": 0.06287039,
+      "step": 2595
+    },
+    {
+      "epoch": 5.192,
+      "grad_norm": 1.250356674194336,
+      "learning_rate": 2e-05,
+      "loss": 0.04755474,
+      "step": 2596
+    },
+    {
+      "epoch": 5.194,
+      "grad_norm": 2.9072906970977783,
+      "learning_rate": 2e-05,
+      "loss": 0.05345254,
+      "step": 2597
+    },
+    {
+      "epoch": 5.196,
+      "grad_norm": 1.3460503816604614,
+      "learning_rate": 2e-05,
+      "loss": 0.05015785,
+      "step": 2598
+    },
+    {
+      "epoch": 5.198,
+      "grad_norm": 1.5144968032836914,
+      "learning_rate": 2e-05,
+      "loss": 0.05026857,
+      "step": 2599
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 1.0161470174789429,
+      "learning_rate": 2e-05,
+      "loss": 0.03177109,
+      "step": 2600
+    },
+    {
+      "epoch": 5.202,
+      "grad_norm": 1.379644513130188,
+      "learning_rate": 2e-05,
+      "loss": 0.0431831,
+      "step": 2601
+    },
+    {
+      "epoch": 5.204,
+      "grad_norm": 1.5978702306747437,
+      "learning_rate": 2e-05,
+      "loss": 0.04161422,
+      "step": 2602
+    },
+    {
+      "epoch": 5.206,
+      "grad_norm": 1.7875207662582397,
+      "learning_rate": 2e-05,
+      "loss": 0.05222338,
+      "step": 2603
+    },
+    {
+      "epoch": 5.208,
+      "grad_norm": 1.0086873769760132,
+      "learning_rate": 2e-05,
+      "loss": 0.02815355,
+      "step": 2604
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 1.4395867586135864,
+      "learning_rate": 2e-05,
+      "loss": 0.03475678,
+      "step": 2605
+    },
+    {
+      "epoch": 5.212,
+      "grad_norm": 1.7755112648010254,
+      "learning_rate": 2e-05,
+      "loss": 0.05450004,
+      "step": 2606
+    },
+    {
+      "epoch": 5.214,
+      "grad_norm": 1.830944299697876,
+      "learning_rate": 2e-05,
+      "loss": 0.04738111,
+      "step": 2607
+    },
+    {
+      "epoch": 5.216,
+      "grad_norm": 1.3517119884490967,
+      "learning_rate": 2e-05,
+      "loss": 0.0596354,
+      "step": 2608
+    },
+    {
+      "epoch": 5.218,
+      "grad_norm": 1.0986253023147583,
+      "learning_rate": 2e-05,
+      "loss": 0.0300137,
+      "step": 2609
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 2.0288379192352295,
+      "learning_rate": 2e-05,
+      "loss": 0.03250675,
+      "step": 2610
+    },
+    {
+      "epoch": 5.222,
+      "grad_norm": 2.0758872032165527,
+      "learning_rate": 2e-05,
+      "loss": 0.05346858,
+      "step": 2611
+    },
+    {
+      "epoch": 5.224,
+      "grad_norm": 1.4884560108184814,
+      "learning_rate": 2e-05,
+      "loss": 0.04885757,
+      "step": 2612
+    },
+    {
+      "epoch": 5.226,
+      "grad_norm": 2.067258596420288,
+      "learning_rate": 2e-05,
+      "loss": 0.04791706,
+      "step": 2613
+    },
+    {
+      "epoch": 5.228,
+      "grad_norm": 1.7092328071594238,
+      "learning_rate": 2e-05,
+      "loss": 0.06562018,
+      "step": 2614
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 2.985393762588501,
+      "learning_rate": 2e-05,
+      "loss": 0.05072825,
+      "step": 2615
+    },
+    {
+      "epoch": 5.232,
+      "grad_norm": 1.6960763931274414,
+      "learning_rate": 2e-05,
+      "loss": 0.05636974,
+      "step": 2616
+    },
+    {
+      "epoch": 5.234,
+      "grad_norm": 1.1226341724395752,
+      "learning_rate": 2e-05,
+      "loss": 0.03822513,
+      "step": 2617
+    },
+    {
+      "epoch": 5.236,
+      "grad_norm": 1.404388427734375,
+      "learning_rate": 2e-05,
+      "loss": 0.03551162,
+      "step": 2618
+    },
+    {
+      "epoch": 5.2379999999999995,
+      "grad_norm": 1.3654124736785889,
+      "learning_rate": 2e-05,
+      "loss": 0.04120996,
+      "step": 2619
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 2.9756903648376465,
+      "learning_rate": 2e-05,
+      "loss": 0.05003164,
+      "step": 2620
+    },
+    {
+      "epoch": 5.242,
+      "grad_norm": 1.9504529237747192,
+      "learning_rate": 2e-05,
+      "loss": 0.06165954,
+      "step": 2621
+    },
+    {
+      "epoch": 5.244,
+      "grad_norm": 1.5840363502502441,
+      "learning_rate": 2e-05,
+      "loss": 0.0396319,
+      "step": 2622
+    },
+    {
+      "epoch": 5.246,
+      "grad_norm": 1.6869369745254517,
+      "learning_rate": 2e-05,
+      "loss": 0.04542349,
+      "step": 2623
+    },
+    {
+      "epoch": 5.248,
+      "grad_norm": 1.6327499151229858,
+      "learning_rate": 2e-05,
+      "loss": 0.05939788,
+      "step": 2624
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 2.854781150817871,
+      "learning_rate": 2e-05,
+      "loss": 0.03455864,
+      "step": 2625
+    },
+    {
+      "epoch": 5.252,
+      "grad_norm": 1.3008240461349487,
+      "learning_rate": 2e-05,
+      "loss": 0.02595445,
+      "step": 2626
+    },
+    {
+      "epoch": 5.254,
+      "grad_norm": 1.2028790712356567,
+      "learning_rate": 2e-05,
+      "loss": 0.02978789,
+      "step": 2627
+    },
+    {
+      "epoch": 5.256,
+      "grad_norm": 1.6047648191452026,
+      "learning_rate": 2e-05,
+      "loss": 0.05063823,
+      "step": 2628
+    },
+    {
+      "epoch": 5.258,
+      "grad_norm": 2.121851682662964,
+      "learning_rate": 2e-05,
+      "loss": 0.03812759,
+      "step": 2629
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 1.4679371118545532,
+      "learning_rate": 2e-05,
+      "loss": 0.04398443,
+      "step": 2630
+    },
+    {
+      "epoch": 5.2620000000000005,
+      "grad_norm": 1.7496758699417114,
+      "learning_rate": 2e-05,
+      "loss": 0.04756953,
+      "step": 2631
+    },
+    {
+      "epoch": 5.264,
+      "grad_norm": 2.85517954826355,
+      "learning_rate": 2e-05,
+      "loss": 0.04580143,
+      "step": 2632
+    },
+    {
+      "epoch": 5.266,
+      "grad_norm": 1.5564442873001099,
+      "learning_rate": 2e-05,
+      "loss": 0.05306519,
+      "step": 2633
+    },
+    {
+      "epoch": 5.268,
+      "grad_norm": 1.9785518646240234,
+      "learning_rate": 2e-05,
+      "loss": 0.0391811,
+      "step": 2634
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 2.4866366386413574,
+      "learning_rate": 2e-05,
+      "loss": 0.05209032,
+      "step": 2635
+    },
+    {
+      "epoch": 5.272,
+      "grad_norm": 1.0321913957595825,
+      "learning_rate": 2e-05,
+      "loss": 0.02281884,
+      "step": 2636
+    },
+    {
+      "epoch": 5.274,
+      "grad_norm": 0.978119432926178,
+      "learning_rate": 2e-05,
+      "loss": 0.02463586,
+      "step": 2637
+    },
+    {
+      "epoch": 5.276,
+      "grad_norm": 1.338571548461914,
+      "learning_rate": 2e-05,
+      "loss": 0.05107386,
+      "step": 2638
+    },
+    {
+      "epoch": 5.2780000000000005,
+      "grad_norm": 1.172133207321167,
+      "learning_rate": 2e-05,
+      "loss": 0.03545574,
+      "step": 2639
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 1.7161072492599487,
+      "learning_rate": 2e-05,
+      "loss": 0.0608919,
+      "step": 2640
+    },
+    {
+      "epoch": 5.282,
+      "grad_norm": 2.2566652297973633,
+      "learning_rate": 2e-05,
+      "loss": 0.06986575,
+      "step": 2641
+    },
+    {
+      "epoch": 5.284,
+      "grad_norm": 1.5361062288284302,
+      "learning_rate": 2e-05,
+      "loss": 0.0354271,
+      "step": 2642
+    },
+    {
+      "epoch": 5.286,
+      "grad_norm": 1.7138904333114624,
+      "learning_rate": 2e-05,
+      "loss": 0.05330596,
+      "step": 2643
+    },
+    {
+      "epoch": 5.288,
+      "grad_norm": 1.9515053033828735,
+      "learning_rate": 2e-05,
+      "loss": 0.05626545,
+      "step": 2644
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 1.527269721031189,
+      "learning_rate": 2e-05,
+      "loss": 0.04371868,
+      "step": 2645
+    },
+    {
+      "epoch": 5.292,
+      "grad_norm": 1.7075636386871338,
+      "learning_rate": 2e-05,
+      "loss": 0.04032372,
+      "step": 2646
+    },
+    {
+      "epoch": 5.294,
+      "grad_norm": 1.0868964195251465,
+      "learning_rate": 2e-05,
+      "loss": 0.03899564,
+      "step": 2647
+    },
+    {
+      "epoch": 5.296,
+      "grad_norm": 3.068068504333496,
+      "learning_rate": 2e-05,
+      "loss": 0.04235051,
+      "step": 2648
+    },
+    {
+      "epoch": 5.298,
+      "grad_norm": 1.6693544387817383,
+      "learning_rate": 2e-05,
+      "loss": 0.05411904,
+      "step": 2649
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 1.2736790180206299,
+      "learning_rate": 2e-05,
+      "loss": 0.0501396,
+      "step": 2650
+    },
+    {
+      "epoch": 5.302,
+      "grad_norm": 1.770841121673584,
+      "learning_rate": 2e-05,
+      "loss": 0.03934953,
+      "step": 2651
+    },
+    {
+      "epoch": 5.304,
+      "grad_norm": 1.1789344549179077,
+      "learning_rate": 2e-05,
+      "loss": 0.03890145,
+      "step": 2652
+    },
+    {
+      "epoch": 5.306,
+      "grad_norm": 1.8938695192337036,
+      "learning_rate": 2e-05,
+      "loss": 0.05737014,
+      "step": 2653
+    },
+    {
+      "epoch": 5.308,
+      "grad_norm": 1.4466159343719482,
+      "learning_rate": 2e-05,
+      "loss": 0.04105823,
+      "step": 2654
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 1.9113630056381226,
+      "learning_rate": 2e-05,
+      "loss": 0.06063426,
+      "step": 2655
+    },
+    {
+      "epoch": 5.312,
+      "grad_norm": 1.341367483139038,
+      "learning_rate": 2e-05,
+      "loss": 0.03131312,
+      "step": 2656
+    },
+    {
+      "epoch": 5.314,
+      "grad_norm": 1.2381536960601807,
+      "learning_rate": 2e-05,
+      "loss": 0.03998591,
+      "step": 2657
+    },
+    {
+      "epoch": 5.316,
+      "grad_norm": 1.210856318473816,
+      "learning_rate": 2e-05,
+      "loss": 0.03550825,
+      "step": 2658
+    },
+    {
+      "epoch": 5.318,
+      "grad_norm": 1.4786920547485352,
+      "learning_rate": 2e-05,
+      "loss": 0.03781786,
+      "step": 2659
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 1.7087515592575073,
+      "learning_rate": 2e-05,
+      "loss": 0.04967047,
+      "step": 2660
+    },
+    {
+      "epoch": 5.322,
+      "grad_norm": 1.5567851066589355,
+      "learning_rate": 2e-05,
+      "loss": 0.03445258,
+      "step": 2661
+    },
+    {
+      "epoch": 5.324,
+      "grad_norm": 0.9272580146789551,
+      "learning_rate": 2e-05,
+      "loss": 0.02637572,
+      "step": 2662
+    },
+    {
+      "epoch": 5.326,
+      "grad_norm": 2.146501302719116,
+      "learning_rate": 2e-05,
+      "loss": 0.03992863,
+      "step": 2663
+    },
+    {
+      "epoch": 5.328,
+      "grad_norm": 2.161367654800415,
+      "learning_rate": 2e-05,
+      "loss": 0.05867213,
+      "step": 2664
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 1.6468945741653442,
+      "learning_rate": 2e-05,
+      "loss": 0.03345583,
+      "step": 2665
+    },
+    {
+      "epoch": 5.332,
+      "grad_norm": 1.8014037609100342,
+      "learning_rate": 2e-05,
+      "loss": 0.04314873,
+      "step": 2666
+    },
+    {
+      "epoch": 5.334,
+      "grad_norm": 1.926483392715454,
+      "learning_rate": 2e-05,
+      "loss": 0.05389591,
+      "step": 2667
+    },
+    {
+      "epoch": 5.336,
+      "grad_norm": 1.567809820175171,
+      "learning_rate": 2e-05,
+      "loss": 0.05018042,
+      "step": 2668
+    },
+    {
+      "epoch": 5.338,
+      "grad_norm": 0.9763085842132568,
+      "learning_rate": 2e-05,
+      "loss": 0.02220888,
+      "step": 2669
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 1.6590309143066406,
+      "learning_rate": 2e-05,
+      "loss": 0.03101541,
+      "step": 2670
+    },
+    {
+      "epoch": 5.342,
+      "grad_norm": 1.1251074075698853,
+      "learning_rate": 2e-05,
+      "loss": 0.03375977,
+      "step": 2671
+    },
+    {
+      "epoch": 5.344,
+      "grad_norm": 1.7373671531677246,
+      "learning_rate": 2e-05,
+      "loss": 0.02912542,
+      "step": 2672
+    },
+    {
+      "epoch": 5.346,
+      "grad_norm": 1.122557282447815,
+      "learning_rate": 2e-05,
+      "loss": 0.02940472,
+      "step": 2673
+    },
+    {
+      "epoch": 5.348,
+      "grad_norm": 1.277514100074768,
+      "learning_rate": 2e-05,
+      "loss": 0.03232025,
+      "step": 2674
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 1.4601491689682007,
+      "learning_rate": 2e-05,
+      "loss": 0.03994402,
+      "step": 2675
+    },
+    {
+      "epoch": 5.352,
+      "grad_norm": 0.8786239624023438,
+      "learning_rate": 2e-05,
+      "loss": 0.02341381,
+      "step": 2676
+    },
+    {
+      "epoch": 5.354,
+      "grad_norm": 1.1858524084091187,
+      "learning_rate": 2e-05,
+      "loss": 0.04786807,
+      "step": 2677
+    },
+    {
+      "epoch": 5.356,
+      "grad_norm": 1.1991043090820312,
+      "learning_rate": 2e-05,
+      "loss": 0.0410894,
+      "step": 2678
+    },
+    {
+      "epoch": 5.358,
+      "grad_norm": 0.9990038871765137,
+      "learning_rate": 2e-05,
+      "loss": 0.03214614,
+      "step": 2679
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 1.499480128288269,
+      "learning_rate": 2e-05,
+      "loss": 0.04654964,
+      "step": 2680
+    },
+    {
+      "epoch": 5.362,
+      "grad_norm": 1.3104462623596191,
+      "learning_rate": 2e-05,
+      "loss": 0.03467711,
+      "step": 2681
+    },
+    {
+      "epoch": 5.364,
+      "grad_norm": 2.457524538040161,
+      "learning_rate": 2e-05,
+      "loss": 0.05111074,
+      "step": 2682
+    },
+    {
+      "epoch": 5.366,
+      "grad_norm": 1.4979830980300903,
+      "learning_rate": 2e-05,
+      "loss": 0.04499523,
+      "step": 2683
+    },
+    {
+      "epoch": 5.368,
+      "grad_norm": 4.056197166442871,
+      "learning_rate": 2e-05,
+      "loss": 0.02836193,
+      "step": 2684
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 3.501704692840576,
+      "learning_rate": 2e-05,
+      "loss": 0.08720149,
+      "step": 2685
+    },
+    {
+      "epoch": 5.372,
+      "grad_norm": 1.552233099937439,
+      "learning_rate": 2e-05,
+      "loss": 0.04281237,
+      "step": 2686
+    },
+    {
+      "epoch": 5.374,
+      "grad_norm": 1.1963515281677246,
+      "learning_rate": 2e-05,
+      "loss": 0.02805083,
+      "step": 2687
+    },
+    {
+      "epoch": 5.376,
+      "grad_norm": 1.2588318586349487,
+      "learning_rate": 2e-05,
+      "loss": 0.03842116,
+      "step": 2688
+    },
+    {
+      "epoch": 5.378,
+      "grad_norm": 2.016977310180664,
+      "learning_rate": 2e-05,
+      "loss": 0.04068321,
+      "step": 2689
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 1.2053779363632202,
+      "learning_rate": 2e-05,
+      "loss": 0.04211229,
+      "step": 2690
+    },
+    {
+      "epoch": 5.382,
+      "grad_norm": 1.485421895980835,
+      "learning_rate": 2e-05,
+      "loss": 0.05219296,
+      "step": 2691
+    },
+    {
+      "epoch": 5.384,
+      "grad_norm": 1.2090396881103516,
+      "learning_rate": 2e-05,
+      "loss": 0.04025387,
+      "step": 2692
+    },
+    {
+      "epoch": 5.386,
+      "grad_norm": 1.3661246299743652,
+      "learning_rate": 2e-05,
+      "loss": 0.03243088,
+      "step": 2693
+    },
+    {
+      "epoch": 5.388,
+      "grad_norm": 1.7728279829025269,
+      "learning_rate": 2e-05,
+      "loss": 0.04520981,
+      "step": 2694
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 1.839638590812683,
+      "learning_rate": 2e-05,
+      "loss": 0.04922792,
+      "step": 2695
+    },
+    {
+      "epoch": 5.392,
+      "grad_norm": 1.2756119966506958,
+      "learning_rate": 2e-05,
+      "loss": 0.03673655,
+      "step": 2696
+    },
+    {
+      "epoch": 5.394,
+      "grad_norm": 1.812853217124939,
+      "learning_rate": 2e-05,
+      "loss": 0.0451444,
+      "step": 2697
+    },
+    {
+      "epoch": 5.396,
+      "grad_norm": 1.732804298400879,
+      "learning_rate": 2e-05,
+      "loss": 0.0696558,
+      "step": 2698
+    },
+    {
+      "epoch": 5.398,
+      "grad_norm": 2.259817123413086,
+      "learning_rate": 2e-05,
+      "loss": 0.05728294,
+      "step": 2699
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 3.6404404640197754,
+      "learning_rate": 2e-05,
+      "loss": 0.05975555,
+      "step": 2700
+    },
+    {
+      "epoch": 5.402,
+      "grad_norm": 1.1913822889328003,
+      "learning_rate": 2e-05,
+      "loss": 0.04311104,
+      "step": 2701
+    },
+    {
+      "epoch": 5.404,
+      "grad_norm": 1.7628343105316162,
+      "learning_rate": 2e-05,
+      "loss": 0.04324729,
+      "step": 2702
+    },
+    {
+      "epoch": 5.406,
+      "grad_norm": 2.0190365314483643,
+      "learning_rate": 2e-05,
+      "loss": 0.05802897,
+      "step": 2703
+    },
+    {
+      "epoch": 5.408,
+      "grad_norm": 1.1721633672714233,
+      "learning_rate": 2e-05,
+      "loss": 0.03633162,
+      "step": 2704
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 1.600894808769226,
+      "learning_rate": 2e-05,
+      "loss": 0.05752605,
+      "step": 2705
+    },
+    {
+      "epoch": 5.412,
+      "grad_norm": 1.0390570163726807,
+      "learning_rate": 2e-05,
+      "loss": 0.03858446,
+      "step": 2706
+    },
+    {
+      "epoch": 5.414,
+      "grad_norm": 1.4907500743865967,
+      "learning_rate": 2e-05,
+      "loss": 0.02707991,
+      "step": 2707
+    },
+    {
+      "epoch": 5.416,
+      "grad_norm": 2.0031204223632812,
+      "learning_rate": 2e-05,
+      "loss": 0.03021158,
+      "step": 2708
+    },
+    {
+      "epoch": 5.418,
+      "grad_norm": 4.072817325592041,
+      "learning_rate": 2e-05,
+      "loss": 0.05053132,
+      "step": 2709
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 1.7337241172790527,
+      "learning_rate": 2e-05,
+      "loss": 0.03692292,
+      "step": 2710
+    },
+    {
+      "epoch": 5.422,
+      "grad_norm": 1.754112720489502,
+      "learning_rate": 2e-05,
+      "loss": 0.03906107,
+      "step": 2711
+    },
+    {
+      "epoch": 5.424,
+      "grad_norm": 4.156593322753906,
+      "learning_rate": 2e-05,
+      "loss": 0.03578386,
+      "step": 2712
+    },
+    {
+      "epoch": 5.426,
+      "grad_norm": 1.2016631364822388,
+      "learning_rate": 2e-05,
+      "loss": 0.03455806,
+      "step": 2713
+    },
+    {
+      "epoch": 5.428,
+      "grad_norm": 0.9195474982261658,
+      "learning_rate": 2e-05,
+      "loss": 0.02441694,
+      "step": 2714
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 1.091452717781067,
+      "learning_rate": 2e-05,
+      "loss": 0.0270972,
+      "step": 2715
+    },
+    {
+      "epoch": 5.432,
+      "grad_norm": 1.6297458410263062,
+      "learning_rate": 2e-05,
+      "loss": 0.04799194,
+      "step": 2716
+    },
+    {
+      "epoch": 5.434,
+      "grad_norm": 1.3862227201461792,
+      "learning_rate": 2e-05,
+      "loss": 0.04216187,
+      "step": 2717
+    },
+    {
+      "epoch": 5.436,
+      "grad_norm": 1.8939886093139648,
+      "learning_rate": 2e-05,
+      "loss": 0.05403071,
+      "step": 2718
+    },
+    {
+      "epoch": 5.438,
+      "grad_norm": 2.1752545833587646,
+      "learning_rate": 2e-05,
+      "loss": 0.05855702,
+      "step": 2719
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 2.9975874423980713,
+      "learning_rate": 2e-05,
+      "loss": 0.0767413,
+      "step": 2720
+    },
+    {
+      "epoch": 5.442,
+      "grad_norm": 1.3926620483398438,
+      "learning_rate": 2e-05,
+      "loss": 0.03643538,
+      "step": 2721
+    },
+    {
+      "epoch": 5.444,
+      "grad_norm": 1.96034574508667,
+      "learning_rate": 2e-05,
+      "loss": 0.04681049,
+      "step": 2722
+    },
+    {
+      "epoch": 5.446,
+      "grad_norm": 1.1249243021011353,
+      "learning_rate": 2e-05,
+      "loss": 0.02556986,
+      "step": 2723
+    },
+    {
+      "epoch": 5.448,
+      "grad_norm": 1.8713010549545288,
+      "learning_rate": 2e-05,
+      "loss": 0.05697018,
+      "step": 2724
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 1.1329511404037476,
+      "learning_rate": 2e-05,
+      "loss": 0.02677982,
+      "step": 2725
+    },
+    {
+      "epoch": 5.452,
+      "grad_norm": 2.1637654304504395,
+      "learning_rate": 2e-05,
+      "loss": 0.05468876,
+      "step": 2726
+    },
+    {
+      "epoch": 5.454,
+      "grad_norm": 1.6031872034072876,
+      "learning_rate": 2e-05,
+      "loss": 0.04942069,
+      "step": 2727
+    },
+    {
+      "epoch": 5.456,
+      "grad_norm": 1.4238561391830444,
+      "learning_rate": 2e-05,
+      "loss": 0.04586676,
+      "step": 2728
+    },
+    {
+      "epoch": 5.458,
+      "grad_norm": 0.7893091440200806,
+      "learning_rate": 2e-05,
+      "loss": 0.02090956,
+      "step": 2729
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 1.485260248184204,
+      "learning_rate": 2e-05,
+      "loss": 0.04206961,
+      "step": 2730
+    },
+    {
+      "epoch": 5.462,
+      "grad_norm": 0.9618636965751648,
+      "learning_rate": 2e-05,
+      "loss": 0.02777426,
+      "step": 2731
+    },
+    {
+      "epoch": 5.464,
+      "grad_norm": 1.663711428642273,
+      "learning_rate": 2e-05,
+      "loss": 0.04290383,
+      "step": 2732
+    },
+    {
+      "epoch": 5.466,
+      "grad_norm": 2.107558250427246,
+      "learning_rate": 2e-05,
+      "loss": 0.05302426,
+      "step": 2733
+    },
+    {
+      "epoch": 5.468,
+      "grad_norm": 1.3562395572662354,
+      "learning_rate": 2e-05,
+      "loss": 0.04122175,
+      "step": 2734
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 1.3446310758590698,
+      "learning_rate": 2e-05,
+      "loss": 0.04664951,
+      "step": 2735
+    },
+    {
+      "epoch": 5.4719999999999995,
+      "grad_norm": 1.7537676095962524,
+      "learning_rate": 2e-05,
+      "loss": 0.05597606,
+      "step": 2736
+    },
+    {
+      "epoch": 5.474,
+      "grad_norm": 1.2579759359359741,
+      "learning_rate": 2e-05,
+      "loss": 0.04209308,
+      "step": 2737
+    },
+    {
+      "epoch": 5.476,
+      "grad_norm": 1.2744957208633423,
+      "learning_rate": 2e-05,
+      "loss": 0.03771916,
+      "step": 2738
+    },
+    {
+      "epoch": 5.478,
+      "grad_norm": 1.1946172714233398,
+      "learning_rate": 2e-05,
+      "loss": 0.04514795,
+      "step": 2739
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 1.2385889291763306,
+      "learning_rate": 2e-05,
+      "loss": 0.03430807,
+      "step": 2740
+    },
+    {
+      "epoch": 5.482,
+      "grad_norm": 1.1070294380187988,
+      "learning_rate": 2e-05,
+      "loss": 0.02781132,
+      "step": 2741
+    },
+    {
+      "epoch": 5.484,
+      "grad_norm": 2.086735248565674,
+      "learning_rate": 2e-05,
+      "loss": 0.04818981,
+      "step": 2742
+    },
+    {
+      "epoch": 5.486,
+      "grad_norm": 1.2726064920425415,
+      "learning_rate": 2e-05,
+      "loss": 0.04557905,
+      "step": 2743
+    },
+    {
+      "epoch": 5.4879999999999995,
+      "grad_norm": 1.2682517766952515,
+      "learning_rate": 2e-05,
+      "loss": 0.03911246,
+      "step": 2744
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 1.2447410821914673,
+      "learning_rate": 2e-05,
+      "loss": 0.04097291,
+      "step": 2745
+    },
+    {
+      "epoch": 5.492,
+      "grad_norm": 1.2082518339157104,
+      "learning_rate": 2e-05,
+      "loss": 0.04021576,
+      "step": 2746
+    },
+    {
+      "epoch": 5.494,
+      "grad_norm": 1.4685015678405762,
+      "learning_rate": 2e-05,
+      "loss": 0.04541424,
+      "step": 2747
+    },
+    {
+      "epoch": 5.496,
+      "grad_norm": 1.4241282939910889,
+      "learning_rate": 2e-05,
+      "loss": 0.03883777,
+      "step": 2748
+    },
+    {
+      "epoch": 5.498,
+      "grad_norm": 1.5642436742782593,
+      "learning_rate": 2e-05,
+      "loss": 0.05445066,
+      "step": 2749
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 1.600178837776184,
+      "learning_rate": 2e-05,
+      "loss": 0.04767666,
+      "step": 2750
+    },
+    {
+      "epoch": 5.502,
+      "grad_norm": 1.0440183877944946,
+      "learning_rate": 2e-05,
+      "loss": 0.03446297,
+      "step": 2751
+    },
+    {
+      "epoch": 5.504,
+      "grad_norm": 1.2513372898101807,
+      "learning_rate": 2e-05,
+      "loss": 0.03565361,
+      "step": 2752
+    },
+    {
+      "epoch": 5.506,
+      "grad_norm": 1.2894079685211182,
+      "learning_rate": 2e-05,
+      "loss": 0.04235047,
+      "step": 2753
+    },
+    {
+      "epoch": 5.508,
+      "grad_norm": 1.8253742456436157,
+      "learning_rate": 2e-05,
+      "loss": 0.05126983,
+      "step": 2754
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 1.881759762763977,
+      "learning_rate": 2e-05,
+      "loss": 0.05438966,
+      "step": 2755
+    },
+    {
+      "epoch": 5.5120000000000005,
+      "grad_norm": 0.9773890972137451,
+      "learning_rate": 2e-05,
+      "loss": 0.02625747,
+      "step": 2756
+    },
+    {
+      "epoch": 5.514,
+      "grad_norm": 1.1596041917800903,
+      "learning_rate": 2e-05,
+      "loss": 0.03821389,
+      "step": 2757
+    },
+    {
+      "epoch": 5.516,
+      "grad_norm": 1.3653874397277832,
+      "learning_rate": 2e-05,
+      "loss": 0.04838298,
+      "step": 2758
+    },
+    {
+      "epoch": 5.518,
+      "grad_norm": 1.4374557733535767,
+      "learning_rate": 2e-05,
+      "loss": 0.04565241,
+      "step": 2759
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 1.6063563823699951,
+      "learning_rate": 2e-05,
+      "loss": 0.053575,
+      "step": 2760
+    },
+    {
+      "epoch": 5.522,
+      "grad_norm": 1.2797068357467651,
+      "learning_rate": 2e-05,
+      "loss": 0.03941024,
+      "step": 2761
+    },
+    {
+      "epoch": 5.524,
+      "grad_norm": 1.0527799129486084,
+      "learning_rate": 2e-05,
+      "loss": 0.01956801,
+      "step": 2762
+    },
+    {
+      "epoch": 5.526,
+      "grad_norm": 1.218973994255066,
+      "learning_rate": 2e-05,
+      "loss": 0.03965396,
+      "step": 2763
+    },
+    {
+      "epoch": 5.5280000000000005,
+      "grad_norm": 2.3557660579681396,
+      "learning_rate": 2e-05,
+      "loss": 0.03057607,
+      "step": 2764
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 1.2339812517166138,
+      "learning_rate": 2e-05,
+      "loss": 0.0315393,
+      "step": 2765
+    },
+    {
+      "epoch": 5.532,
+      "grad_norm": 1.698705792427063,
+      "learning_rate": 2e-05,
+      "loss": 0.03224694,
+      "step": 2766
+    },
+    {
+      "epoch": 5.534,
+      "grad_norm": 1.9057179689407349,
+      "learning_rate": 2e-05,
+      "loss": 0.05384257,
+      "step": 2767
+    },
+    {
+      "epoch": 5.536,
+      "grad_norm": 1.3799241781234741,
+      "learning_rate": 2e-05,
+      "loss": 0.04452458,
+      "step": 2768
+    },
+    {
+      "epoch": 5.538,
+      "grad_norm": 1.2078452110290527,
+      "learning_rate": 2e-05,
+      "loss": 0.04307923,
+      "step": 2769
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 1.1021215915679932,
+      "learning_rate": 2e-05,
+      "loss": 0.03390051,
+      "step": 2770
+    },
+    {
+      "epoch": 5.542,
+      "grad_norm": 0.9905539751052856,
+      "learning_rate": 2e-05,
+      "loss": 0.03263092,
+      "step": 2771
+    },
+    {
+      "epoch": 5.5440000000000005,
+      "grad_norm": 2.1359174251556396,
+      "learning_rate": 2e-05,
+      "loss": 0.05425924,
+      "step": 2772
+    },
+    {
+      "epoch": 5.546,
+      "grad_norm": 1.9481074810028076,
+      "learning_rate": 2e-05,
+      "loss": 0.05403948,
+      "step": 2773
+    },
+    {
+      "epoch": 5.548,
+      "grad_norm": 2.03695011138916,
+      "learning_rate": 2e-05,
+      "loss": 0.03124078,
+      "step": 2774
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 2.420414447784424,
+      "learning_rate": 2e-05,
+      "loss": 0.05495863,
+      "step": 2775
+    },
+    {
+      "epoch": 5.552,
+      "grad_norm": 2.0198957920074463,
+      "learning_rate": 2e-05,
+      "loss": 0.05129372,
+      "step": 2776
+    },
+    {
+      "epoch": 5.554,
+      "grad_norm": 1.7094142436981201,
+      "learning_rate": 2e-05,
+      "loss": 0.0504851,
+      "step": 2777
+    },
+    {
+      "epoch": 5.556,
+      "grad_norm": 3.827301502227783,
+      "learning_rate": 2e-05,
+      "loss": 0.05337991,
+      "step": 2778
+    },
+    {
+      "epoch": 5.558,
+      "grad_norm": 1.3300044536590576,
+      "learning_rate": 2e-05,
+      "loss": 0.03548397,
+      "step": 2779
+    },
+    {
+      "epoch": 5.5600000000000005,
+      "grad_norm": 1.937286376953125,
+      "learning_rate": 2e-05,
+      "loss": 0.04464417,
+      "step": 2780
+    },
+    {
+      "epoch": 5.562,
+      "grad_norm": 2.370962142944336,
+      "learning_rate": 2e-05,
+      "loss": 0.05774143,
+      "step": 2781
+    },
+    {
+      "epoch": 5.564,
+      "grad_norm": 1.7345980405807495,
+      "learning_rate": 2e-05,
+      "loss": 0.04462703,
+      "step": 2782
+    },
+    {
+      "epoch": 5.566,
+      "grad_norm": 1.5614328384399414,
+      "learning_rate": 2e-05,
+      "loss": 0.04742298,
+      "step": 2783
+    },
+    {
+      "epoch": 5.568,
+      "grad_norm": 1.480557918548584,
+      "learning_rate": 2e-05,
+      "loss": 0.04577837,
+      "step": 2784
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 1.257659912109375,
+      "learning_rate": 2e-05,
+      "loss": 0.03243734,
+      "step": 2785
+    },
+    {
+      "epoch": 5.572,
+      "grad_norm": 2.447787046432495,
+      "learning_rate": 2e-05,
+      "loss": 0.01604201,
+      "step": 2786
+    },
+    {
+      "epoch": 5.574,
+      "grad_norm": 1.1269577741622925,
+      "learning_rate": 2e-05,
+      "loss": 0.02630051,
+      "step": 2787
+    },
+    {
+      "epoch": 5.576,
+      "grad_norm": 1.0902881622314453,
+      "learning_rate": 2e-05,
+      "loss": 0.02891617,
+      "step": 2788
+    },
+    {
+      "epoch": 5.578,
+      "grad_norm": 2.013070821762085,
+      "learning_rate": 2e-05,
+      "loss": 0.04982309,
+      "step": 2789
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 1.520807147026062,
+      "learning_rate": 2e-05,
+      "loss": 0.06273451,
+      "step": 2790
+    },
+    {
+      "epoch": 5.582,
+      "grad_norm": 1.0118508338928223,
+      "learning_rate": 2e-05,
+      "loss": 0.0323143,
+      "step": 2791
+    },
+    {
+      "epoch": 5.584,
+      "grad_norm": 1.589319109916687,
+      "learning_rate": 2e-05,
+      "loss": 0.04532817,
+      "step": 2792
+    },
+    {
+      "epoch": 5.586,
+      "grad_norm": 1.4283770322799683,
+      "learning_rate": 2e-05,
+      "loss": 0.03695945,
+      "step": 2793
+    },
+    {
+      "epoch": 5.588,
+      "grad_norm": 1.5165592432022095,
+      "learning_rate": 2e-05,
+      "loss": 0.04063965,
+      "step": 2794
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 2.1748976707458496,
+      "learning_rate": 2e-05,
+      "loss": 0.0483332,
+      "step": 2795
+    },
+    {
+      "epoch": 5.592,
+      "grad_norm": 1.4957975149154663,
+      "learning_rate": 2e-05,
+      "loss": 0.03734215,
+      "step": 2796
+    },
+    {
+      "epoch": 5.594,
+      "grad_norm": 1.5057545900344849,
+      "learning_rate": 2e-05,
+      "loss": 0.0448417,
+      "step": 2797
+    },
+    {
+      "epoch": 5.596,
+      "grad_norm": 1.6289266347885132,
+      "learning_rate": 2e-05,
+      "loss": 0.04035756,
+      "step": 2798
+    },
+    {
+      "epoch": 5.598,
+      "grad_norm": 1.945152759552002,
+      "learning_rate": 2e-05,
+      "loss": 0.04505145,
+      "step": 2799
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 3.9745194911956787,
+      "learning_rate": 2e-05,
+      "loss": 0.05695663,
+      "step": 2800
+    },
+    {
+      "epoch": 5.602,
+      "grad_norm": 1.1209356784820557,
+      "learning_rate": 2e-05,
+      "loss": 0.04382028,
+      "step": 2801
+    },
+    {
+      "epoch": 5.604,
+      "grad_norm": 1.7319954633712769,
+      "learning_rate": 2e-05,
+      "loss": 0.07861836,
+      "step": 2802
+    },
+    {
+      "epoch": 5.606,
+      "grad_norm": 1.2670855522155762,
+      "learning_rate": 2e-05,
+      "loss": 0.03723117,
+      "step": 2803
+    },
+    {
+      "epoch": 5.608,
+      "grad_norm": 1.5532779693603516,
+      "learning_rate": 2e-05,
+      "loss": 0.03009872,
+      "step": 2804
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 1.9622262716293335,
+      "learning_rate": 2e-05,
+      "loss": 0.06740649,
+      "step": 2805
+    },
+    {
+      "epoch": 5.612,
+      "grad_norm": 1.1806871891021729,
+      "learning_rate": 2e-05,
+      "loss": 0.03466694,
+      "step": 2806
+    },
+    {
+      "epoch": 5.614,
+      "grad_norm": 2.1496856212615967,
+      "learning_rate": 2e-05,
+      "loss": 0.060385,
+      "step": 2807
+    },
+    {
+      "epoch": 5.616,
+      "grad_norm": 1.7937800884246826,
+      "learning_rate": 2e-05,
+      "loss": 0.04280568,
+      "step": 2808
+    },
+    {
+      "epoch": 5.618,
+      "grad_norm": 1.664449691772461,
+      "learning_rate": 2e-05,
+      "loss": 0.0573493,
+      "step": 2809
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 2.113910436630249,
+      "learning_rate": 2e-05,
+      "loss": 0.05896275,
+      "step": 2810
+    },
+    {
+      "epoch": 5.622,
+      "grad_norm": 1.4774582386016846,
+      "learning_rate": 2e-05,
+      "loss": 0.05064362,
+      "step": 2811
+    },
+    {
+      "epoch": 5.624,
+      "grad_norm": 1.077660083770752,
+      "learning_rate": 2e-05,
+      "loss": 0.03990199,
+      "step": 2812
+    },
+    {
+      "epoch": 5.626,
+      "grad_norm": 2.0504021644592285,
+      "learning_rate": 2e-05,
+      "loss": 0.04368483,
+      "step": 2813
+    },
+    {
+      "epoch": 5.628,
+      "grad_norm": 1.3828272819519043,
+      "learning_rate": 2e-05,
+      "loss": 0.0452172,
+      "step": 2814
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 1.3679569959640503,
+      "learning_rate": 2e-05,
+      "loss": 0.04204818,
+      "step": 2815
+    },
+    {
+      "epoch": 5.632,
+      "grad_norm": 1.904268503189087,
+      "learning_rate": 2e-05,
+      "loss": 0.03786581,
+      "step": 2816
+    },
+    {
+      "epoch": 5.634,
+      "grad_norm": 1.5074198246002197,
+      "learning_rate": 2e-05,
+      "loss": 0.05239385,
+      "step": 2817
+    },
+    {
+      "epoch": 5.636,
+      "grad_norm": 1.603095531463623,
+      "learning_rate": 2e-05,
+      "loss": 0.04265279,
+      "step": 2818
+    },
+    {
+      "epoch": 5.638,
+      "grad_norm": 1.2010900974273682,
+      "learning_rate": 2e-05,
+      "loss": 0.0357244,
+      "step": 2819
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 1.8262673616409302,
+      "learning_rate": 2e-05,
+      "loss": 0.05815496,
+      "step": 2820
+    },
+    {
+      "epoch": 5.642,
+      "grad_norm": 1.6634687185287476,
+      "learning_rate": 2e-05,
+      "loss": 0.04178239,
+      "step": 2821
+    },
+    {
+      "epoch": 5.644,
+      "grad_norm": 1.9540555477142334,
+      "learning_rate": 2e-05,
+      "loss": 0.0514649,
+      "step": 2822
+    },
+    {
+      "epoch": 5.646,
+      "grad_norm": 1.463693618774414,
+      "learning_rate": 2e-05,
+      "loss": 0.0404924,
+      "step": 2823
+    },
+    {
+      "epoch": 5.648,
+      "grad_norm": 2.447908639907837,
+      "learning_rate": 2e-05,
+      "loss": 0.03566689,
+      "step": 2824
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 1.290624737739563,
+      "learning_rate": 2e-05,
+      "loss": 0.03801401,
+      "step": 2825
+    },
+    {
+      "epoch": 5.652,
+      "grad_norm": 1.4527415037155151,
+      "learning_rate": 2e-05,
+      "loss": 0.03635446,
+      "step": 2826
+    },
+    {
+      "epoch": 5.654,
+      "grad_norm": 1.6195863485336304,
+      "learning_rate": 2e-05,
+      "loss": 0.05585508,
+      "step": 2827
+    },
+    {
+      "epoch": 5.656,
+      "grad_norm": 1.6211501359939575,
+      "learning_rate": 2e-05,
+      "loss": 0.06746172,
+      "step": 2828
+    },
+    {
+      "epoch": 5.658,
+      "grad_norm": 1.346248745918274,
+      "learning_rate": 2e-05,
+      "loss": 0.04963518,
+      "step": 2829
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 1.062984585762024,
+      "learning_rate": 2e-05,
+      "loss": 0.03912539,
+      "step": 2830
+    },
+    {
+      "epoch": 5.662,
+      "grad_norm": 1.6748669147491455,
+      "learning_rate": 2e-05,
+      "loss": 0.06159768,
+      "step": 2831
+    },
+    {
+      "epoch": 5.664,
+      "grad_norm": 1.3549000024795532,
+      "learning_rate": 2e-05,
+      "loss": 0.04779611,
+      "step": 2832
+    },
+    {
+      "epoch": 5.666,
+      "grad_norm": 1.2861626148223877,
+      "learning_rate": 2e-05,
+      "loss": 0.03312145,
+      "step": 2833
+    },
+    {
+      "epoch": 5.668,
+      "grad_norm": 2.1982734203338623,
+      "learning_rate": 2e-05,
+      "loss": 0.05529946,
+      "step": 2834
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 1.2915890216827393,
+      "learning_rate": 2e-05,
+      "loss": 0.03705323,
+      "step": 2835
+    },
+    {
+      "epoch": 5.672,
+      "grad_norm": 1.5575082302093506,
+      "learning_rate": 2e-05,
+      "loss": 0.03686603,
+      "step": 2836
+    },
+    {
+      "epoch": 5.674,
+      "grad_norm": 1.038826823234558,
+      "learning_rate": 2e-05,
+      "loss": 0.03793515,
+      "step": 2837
+    },
+    {
+      "epoch": 5.676,
+      "grad_norm": 1.6118172407150269,
+      "learning_rate": 2e-05,
+      "loss": 0.03095808,
+      "step": 2838
+    },
+    {
+      "epoch": 5.678,
+      "grad_norm": 1.6042510271072388,
+      "learning_rate": 2e-05,
+      "loss": 0.05740377,
+      "step": 2839
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 1.7100886106491089,
+      "learning_rate": 2e-05,
+      "loss": 0.05264761,
+      "step": 2840
+    },
+    {
+      "epoch": 5.682,
+      "grad_norm": 1.575619101524353,
+      "learning_rate": 2e-05,
+      "loss": 0.03263854,
+      "step": 2841
+    },
+    {
+      "epoch": 5.684,
+      "grad_norm": 1.4796384572982788,
+      "learning_rate": 2e-05,
+      "loss": 0.04719574,
+      "step": 2842
+    },
+    {
+      "epoch": 5.686,
+      "grad_norm": 1.3210619688034058,
+      "learning_rate": 2e-05,
+      "loss": 0.0271394,
+      "step": 2843
+    },
+    {
+      "epoch": 5.688,
+      "grad_norm": 1.0833356380462646,
+      "learning_rate": 2e-05,
+      "loss": 0.03883443,
+      "step": 2844
+    },
+    {
+      "epoch": 5.6899999999999995,
+      "grad_norm": 2.5778627395629883,
+      "learning_rate": 2e-05,
+      "loss": 0.05159726,
+      "step": 2845
+    },
+    {
+      "epoch": 5.692,
+      "grad_norm": 2.201190948486328,
+      "learning_rate": 2e-05,
+      "loss": 0.04031905,
+      "step": 2846
+    },
+    {
+      "epoch": 5.694,
+      "grad_norm": 1.1688532829284668,
+      "learning_rate": 2e-05,
+      "loss": 0.03291119,
+      "step": 2847
+    },
+    {
+      "epoch": 5.696,
+      "grad_norm": 1.796081304550171,
+      "learning_rate": 2e-05,
+      "loss": 0.05050949,
+      "step": 2848
+    },
+    {
+      "epoch": 5.698,
+      "grad_norm": 1.016455888748169,
+      "learning_rate": 2e-05,
+      "loss": 0.03628581,
+      "step": 2849
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 1.9447333812713623,
+      "learning_rate": 2e-05,
+      "loss": 0.04152183,
+      "step": 2850
+    },
+    {
+      "epoch": 5.702,
+      "grad_norm": 1.3478142023086548,
+      "learning_rate": 2e-05,
+      "loss": 0.03443387,
+      "step": 2851
+    },
+    {
+      "epoch": 5.704,
+      "grad_norm": 1.6059876680374146,
+      "learning_rate": 2e-05,
+      "loss": 0.05241629,
+      "step": 2852
+    },
+    {
+      "epoch": 5.7059999999999995,
+      "grad_norm": 1.2685903310775757,
+      "learning_rate": 2e-05,
+      "loss": 0.03086682,
+      "step": 2853
+    },
+    {
+      "epoch": 5.708,
+      "grad_norm": 2.984708547592163,
+      "learning_rate": 2e-05,
+      "loss": 0.05078296,
+      "step": 2854
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 3.1346423625946045,
+      "learning_rate": 2e-05,
+      "loss": 0.04925366,
+      "step": 2855
+    },
+    {
+      "epoch": 5.712,
+      "grad_norm": 1.1678012609481812,
+      "learning_rate": 2e-05,
+      "loss": 0.03549577,
+      "step": 2856
+    },
+    {
+      "epoch": 5.714,
+      "grad_norm": 1.408694863319397,
+      "learning_rate": 2e-05,
+      "loss": 0.03885608,
+      "step": 2857
+    },
+    {
+      "epoch": 5.716,
+      "grad_norm": 0.9981144070625305,
+      "learning_rate": 2e-05,
+      "loss": 0.02968807,
+      "step": 2858
+    },
+    {
+      "epoch": 5.718,
+      "grad_norm": 1.307047724723816,
+      "learning_rate": 2e-05,
+      "loss": 0.03957972,
+      "step": 2859
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 2.804670572280884,
+      "learning_rate": 2e-05,
+      "loss": 0.04513305,
+      "step": 2860
+    },
+    {
+      "epoch": 5.7219999999999995,
+      "grad_norm": 1.1364901065826416,
+      "learning_rate": 2e-05,
+      "loss": 0.03480239,
+      "step": 2861
+    },
+    {
+      "epoch": 5.724,
+      "grad_norm": 1.2232093811035156,
+      "learning_rate": 2e-05,
+      "loss": 0.03162998,
+      "step": 2862
+    },
+    {
+      "epoch": 5.726,
+      "grad_norm": 1.298130989074707,
+      "learning_rate": 2e-05,
+      "loss": 0.04124122,
+      "step": 2863
+    },
+    {
+      "epoch": 5.728,
+      "grad_norm": 1.2249360084533691,
+      "learning_rate": 2e-05,
+      "loss": 0.04093556,
+      "step": 2864
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 1.261306881904602,
+      "learning_rate": 2e-05,
+      "loss": 0.04419369,
+      "step": 2865
+    },
+    {
+      "epoch": 5.732,
+      "grad_norm": 1.0616569519042969,
+      "learning_rate": 2e-05,
+      "loss": 0.02921419,
+      "step": 2866
+    },
+    {
+      "epoch": 5.734,
+      "grad_norm": 1.7513487339019775,
+      "learning_rate": 2e-05,
+      "loss": 0.03050586,
+      "step": 2867
+    },
+    {
+      "epoch": 5.736,
+      "grad_norm": 1.4477100372314453,
+      "learning_rate": 2e-05,
+      "loss": 0.0365976,
+      "step": 2868
+    },
+    {
+      "epoch": 5.7379999999999995,
+      "grad_norm": 2.1464011669158936,
+      "learning_rate": 2e-05,
+      "loss": 0.05555936,
+      "step": 2869
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 1.7182115316390991,
+      "learning_rate": 2e-05,
+      "loss": 0.04245908,
+      "step": 2870
+    },
+    {
+      "epoch": 5.742,
+      "grad_norm": 1.0961130857467651,
+      "learning_rate": 2e-05,
+      "loss": 0.03913336,
+      "step": 2871
+    },
+    {
+      "epoch": 5.744,
+      "grad_norm": 1.87662935256958,
+      "learning_rate": 2e-05,
+      "loss": 0.04560525,
+      "step": 2872
+    },
+    {
+      "epoch": 5.746,
+      "grad_norm": 1.343624234199524,
+      "learning_rate": 2e-05,
+      "loss": 0.03066914,
+      "step": 2873
+    },
+    {
+      "epoch": 5.748,
+      "grad_norm": 2.7794783115386963,
+      "learning_rate": 2e-05,
+      "loss": 0.05809136,
+      "step": 2874
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 1.3323760032653809,
+      "learning_rate": 2e-05,
+      "loss": 0.05136766,
+      "step": 2875
+    },
+    {
+      "epoch": 5.752,
+      "grad_norm": 1.9432049989700317,
+      "learning_rate": 2e-05,
+      "loss": 0.05197339,
+      "step": 2876
+    },
+    {
+      "epoch": 5.754,
+      "grad_norm": 1.1205425262451172,
+      "learning_rate": 2e-05,
+      "loss": 0.03465851,
+      "step": 2877
+    },
+    {
+      "epoch": 5.756,
+      "grad_norm": 1.6224393844604492,
+      "learning_rate": 2e-05,
+      "loss": 0.03670148,
+      "step": 2878
+    },
+    {
+      "epoch": 5.758,
+      "grad_norm": 1.3828800916671753,
+      "learning_rate": 2e-05,
+      "loss": 0.04724912,
+      "step": 2879
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 1.6124157905578613,
+      "learning_rate": 2e-05,
+      "loss": 0.03899434,
+      "step": 2880
+    },
+    {
+      "epoch": 5.7620000000000005,
+      "grad_norm": 1.2006202936172485,
+      "learning_rate": 2e-05,
+      "loss": 0.04298317,
+      "step": 2881
+    },
+    {
+      "epoch": 5.764,
+      "grad_norm": 1.8390953540802002,
+      "learning_rate": 2e-05,
+      "loss": 0.04603451,
+      "step": 2882
+    },
+    {
+      "epoch": 5.766,
+      "grad_norm": 3.418121099472046,
+      "learning_rate": 2e-05,
+      "loss": 0.03081621,
+      "step": 2883
+    },
+    {
+      "epoch": 5.768,
+      "grad_norm": 0.9251627922058105,
+      "learning_rate": 2e-05,
+      "loss": 0.02699123,
+      "step": 2884
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 1.5069046020507812,
+      "learning_rate": 2e-05,
+      "loss": 0.03988301,
+      "step": 2885
+    },
+    {
+      "epoch": 5.772,
+      "grad_norm": 2.365957498550415,
+      "learning_rate": 2e-05,
+      "loss": 0.03983999,
+      "step": 2886
+    },
+    {
+      "epoch": 5.774,
+      "grad_norm": 1.5599316358566284,
+      "learning_rate": 2e-05,
+      "loss": 0.04989076,
+      "step": 2887
+    },
+    {
+      "epoch": 5.776,
+      "grad_norm": 1.1228059530258179,
+      "learning_rate": 2e-05,
+      "loss": 0.03197101,
+      "step": 2888
+    },
+    {
+      "epoch": 5.7780000000000005,
+      "grad_norm": 1.517126202583313,
+      "learning_rate": 2e-05,
+      "loss": 0.03176156,
+      "step": 2889
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 1.023742914199829,
+      "learning_rate": 2e-05,
+      "loss": 0.035194,
+      "step": 2890
+    },
+    {
+      "epoch": 5.782,
+      "grad_norm": 1.5021580457687378,
+      "learning_rate": 2e-05,
+      "loss": 0.0346195,
+      "step": 2891
+    },
+    {
+      "epoch": 5.784,
+      "grad_norm": 1.5000685453414917,
+      "learning_rate": 2e-05,
+      "loss": 0.03589319,
+      "step": 2892
+    },
+    {
+      "epoch": 5.786,
+      "grad_norm": 1.2418376207351685,
+      "learning_rate": 2e-05,
+      "loss": 0.03027872,
+      "step": 2893
+    },
+    {
+      "epoch": 5.788,
+      "grad_norm": 1.425329566001892,
+      "learning_rate": 2e-05,
+      "loss": 0.03400642,
+      "step": 2894
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 1.4583309888839722,
+      "learning_rate": 2e-05,
+      "loss": 0.03597549,
+      "step": 2895
+    },
+    {
+      "epoch": 5.792,
+      "grad_norm": 1.5168836116790771,
+      "learning_rate": 2e-05,
+      "loss": 0.02261955,
+      "step": 2896
+    },
+    {
+      "epoch": 5.7940000000000005,
+      "grad_norm": 1.7382346391677856,
+      "learning_rate": 2e-05,
+      "loss": 0.03887838,
+      "step": 2897
+    },
+    {
+      "epoch": 5.796,
+      "grad_norm": 3.2048113346099854,
+      "learning_rate": 2e-05,
+      "loss": 0.07382256,
+      "step": 2898
+    },
+    {
+      "epoch": 5.798,
+      "grad_norm": 1.1313360929489136,
+      "learning_rate": 2e-05,
+      "loss": 0.02457133,
+      "step": 2899
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 2.6181559562683105,
+      "learning_rate": 2e-05,
+      "loss": 0.04528946,
+      "step": 2900
+    },
+    {
+      "epoch": 5.802,
+      "grad_norm": 2.0275521278381348,
+      "learning_rate": 2e-05,
+      "loss": 0.04249582,
+      "step": 2901
+    },
+    {
+      "epoch": 5.804,
+      "grad_norm": 3.3684051036834717,
+      "learning_rate": 2e-05,
+      "loss": 0.04949026,
+      "step": 2902
+    },
+    {
+      "epoch": 5.806,
+      "grad_norm": 1.1052809953689575,
+      "learning_rate": 2e-05,
+      "loss": 0.03748845,
+      "step": 2903
+    },
+    {
+      "epoch": 5.808,
+      "grad_norm": 1.147633671760559,
+      "learning_rate": 2e-05,
+      "loss": 0.02890555,
+      "step": 2904
+    },
+    {
+      "epoch": 5.8100000000000005,
+      "grad_norm": 2.285583257675171,
+      "learning_rate": 2e-05,
+      "loss": 0.04941707,
+      "step": 2905
+    },
+    {
+      "epoch": 5.812,
+      "grad_norm": 2.404172658920288,
+      "learning_rate": 2e-05,
+      "loss": 0.02909764,
+      "step": 2906
+    },
+    {
+      "epoch": 5.814,
+      "grad_norm": 2.2711870670318604,
+      "learning_rate": 2e-05,
+      "loss": 0.06077723,
+      "step": 2907
+    },
+    {
+      "epoch": 5.816,
+      "grad_norm": 1.342374563217163,
+      "learning_rate": 2e-05,
+      "loss": 0.03685233,
+      "step": 2908
+    },
+    {
+      "epoch": 5.818,
+      "grad_norm": 1.1346458196640015,
+      "learning_rate": 2e-05,
+      "loss": 0.02703713,
+      "step": 2909
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 0.9739170670509338,
+      "learning_rate": 2e-05,
+      "loss": 0.0288388,
+      "step": 2910
+    },
+    {
+      "epoch": 5.822,
+      "grad_norm": 1.538444995880127,
+      "learning_rate": 2e-05,
+      "loss": 0.0610106,
+      "step": 2911
+    },
+    {
+      "epoch": 5.824,
+      "grad_norm": 1.5088497400283813,
+      "learning_rate": 2e-05,
+      "loss": 0.03801049,
+      "step": 2912
+    },
+    {
+      "epoch": 5.826,
+      "grad_norm": 2.691092014312744,
+      "learning_rate": 2e-05,
+      "loss": 0.05319805,
+      "step": 2913
+    },
+    {
+      "epoch": 5.828,
+      "grad_norm": 1.7784383296966553,
+      "learning_rate": 2e-05,
+      "loss": 0.03665651,
+      "step": 2914
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 2.699039936065674,
+      "learning_rate": 2e-05,
+      "loss": 0.06110234,
+      "step": 2915
+    },
+    {
+      "epoch": 5.832,
+      "grad_norm": 1.0186821222305298,
+      "learning_rate": 2e-05,
+      "loss": 0.0389097,
+      "step": 2916
+    },
+    {
+      "epoch": 5.834,
+      "grad_norm": 1.5855401754379272,
+      "learning_rate": 2e-05,
+      "loss": 0.05706736,
+      "step": 2917
+    },
+    {
+      "epoch": 5.836,
+      "grad_norm": 1.5186421871185303,
+      "learning_rate": 2e-05,
+      "loss": 0.04286093,
+      "step": 2918
+    },
+    {
+      "epoch": 5.838,
+      "grad_norm": 1.8473025560379028,
+      "learning_rate": 2e-05,
+      "loss": 0.04118644,
+      "step": 2919
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 1.2848973274230957,
+      "learning_rate": 2e-05,
+      "loss": 0.03644439,
+      "step": 2920
+    },
+    {
+      "epoch": 5.842,
+      "grad_norm": 1.198118805885315,
+      "learning_rate": 2e-05,
+      "loss": 0.04760751,
+      "step": 2921
+    },
+    {
+      "epoch": 5.844,
+      "grad_norm": 1.7948402166366577,
+      "learning_rate": 2e-05,
+      "loss": 0.05257875,
+      "step": 2922
+    },
+    {
+      "epoch": 5.846,
+      "grad_norm": 1.2989088296890259,
+      "learning_rate": 2e-05,
+      "loss": 0.0287516,
+      "step": 2923
+    },
+    {
+      "epoch": 5.848,
+      "grad_norm": 0.9115301966667175,
+      "learning_rate": 2e-05,
+      "loss": 0.0334456,
+      "step": 2924
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 11.285009384155273,
+      "learning_rate": 2e-05,
+      "loss": 0.03400384,
+      "step": 2925
+    },
+    {
+      "epoch": 5.852,
+      "grad_norm": 5.87296199798584,
+      "learning_rate": 2e-05,
+      "loss": 0.10033739,
+      "step": 2926
+    },
+    {
+      "epoch": 5.854,
+      "grad_norm": 1.882878065109253,
+      "learning_rate": 2e-05,
+      "loss": 0.04880483,
+      "step": 2927
+    },
+    {
+      "epoch": 5.856,
+      "grad_norm": 11.981039047241211,
+      "learning_rate": 2e-05,
+      "loss": 0.166832,
+      "step": 2928
+    },
+    {
+      "epoch": 5.858,
+      "grad_norm": 1.1964349746704102,
+      "learning_rate": 2e-05,
+      "loss": 0.02282231,
+      "step": 2929
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 3.9903624057769775,
+      "learning_rate": 2e-05,
+      "loss": 0.06060695,
+      "step": 2930
+    },
+    {
+      "epoch": 5.862,
+      "grad_norm": 1.5105197429656982,
+      "learning_rate": 2e-05,
+      "loss": 0.04793704,
+      "step": 2931
+    },
+    {
+      "epoch": 5.864,
+      "grad_norm": 1.41489839553833,
+      "learning_rate": 2e-05,
+      "loss": 0.03938361,
+      "step": 2932
+    },
+    {
+      "epoch": 5.866,
+      "grad_norm": 2.26653790473938,
+      "learning_rate": 2e-05,
+      "loss": 0.0598973,
+      "step": 2933
+    },
+    {
+      "epoch": 5.868,
+      "grad_norm": 2.4137723445892334,
+      "learning_rate": 2e-05,
+      "loss": 0.05444659,
+      "step": 2934
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 1.2001371383666992,
+      "learning_rate": 2e-05,
+      "loss": 0.03985805,
+      "step": 2935
+    },
+    {
+      "epoch": 5.872,
+      "grad_norm": 1.5971794128417969,
+      "learning_rate": 2e-05,
+      "loss": 0.06079929,
+      "step": 2936
+    },
+    {
+      "epoch": 5.874,
+      "grad_norm": 1.318945050239563,
+      "learning_rate": 2e-05,
+      "loss": 0.04854849,
+      "step": 2937
+    },
+    {
+      "epoch": 5.876,
+      "grad_norm": 1.246225118637085,
+      "learning_rate": 2e-05,
+      "loss": 0.03397983,
+      "step": 2938
+    },
+    {
+      "epoch": 5.878,
+      "grad_norm": 2.4939327239990234,
+      "learning_rate": 2e-05,
+      "loss": 0.04911096,
+      "step": 2939
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 1.0084093809127808,
+      "learning_rate": 2e-05,
+      "loss": 0.03702591,
+      "step": 2940
+    },
+    {
+      "epoch": 5.882,
+      "grad_norm": 1.044545292854309,
+      "learning_rate": 2e-05,
+      "loss": 0.03148817,
+      "step": 2941
+    },
+    {
+      "epoch": 5.884,
+      "grad_norm": 1.6608963012695312,
+      "learning_rate": 2e-05,
+      "loss": 0.05824044,
+      "step": 2942
+    },
+    {
+      "epoch": 5.886,
+      "grad_norm": 2.168658494949341,
+      "learning_rate": 2e-05,
+      "loss": 0.05980438,
+      "step": 2943
+    },
+    {
+      "epoch": 5.888,
+      "grad_norm": 1.3249305486679077,
+      "learning_rate": 2e-05,
+      "loss": 0.04346288,
+      "step": 2944
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 1.1912715435028076,
+      "learning_rate": 2e-05,
+      "loss": 0.03074624,
+      "step": 2945
+    },
+    {
+      "epoch": 5.892,
+      "grad_norm": 2.0439677238464355,
+      "learning_rate": 2e-05,
+      "loss": 0.06351136,
+      "step": 2946
+    },
+    {
+      "epoch": 5.894,
+      "grad_norm": 1.2318497896194458,
+      "learning_rate": 2e-05,
+      "loss": 0.03561502,
+      "step": 2947
+    },
+    {
+      "epoch": 5.896,
+      "grad_norm": 1.001007318496704,
+      "learning_rate": 2e-05,
+      "loss": 0.0264318,
+      "step": 2948
+    },
+    {
+      "epoch": 5.898,
+      "grad_norm": 1.6429765224456787,
+      "learning_rate": 2e-05,
+      "loss": 0.05054122,
+      "step": 2949
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 1.8154584169387817,
+      "learning_rate": 2e-05,
+      "loss": 0.05611707,
+      "step": 2950
+    },
+    {
+      "epoch": 5.902,
+      "grad_norm": 1.8020012378692627,
+      "learning_rate": 2e-05,
+      "loss": 0.0445966,
+      "step": 2951
+    },
+    {
+      "epoch": 5.904,
+      "grad_norm": 1.1561640501022339,
+      "learning_rate": 2e-05,
+      "loss": 0.042308,
+      "step": 2952
+    },
+    {
+      "epoch": 5.906,
+      "grad_norm": 1.3049297332763672,
+      "learning_rate": 2e-05,
+      "loss": 0.02226608,
+      "step": 2953
+    },
+    {
+      "epoch": 5.908,
+      "grad_norm": 1.9397492408752441,
+      "learning_rate": 2e-05,
+      "loss": 0.04040186,
+      "step": 2954
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 2.0794615745544434,
+      "learning_rate": 2e-05,
+      "loss": 0.04332725,
+      "step": 2955
+    },
+    {
+      "epoch": 5.912,
+      "grad_norm": 1.1256343126296997,
+      "learning_rate": 2e-05,
+      "loss": 0.03192374,
+      "step": 2956
+    },
+    {
+      "epoch": 5.914,
+      "grad_norm": 1.629341959953308,
+      "learning_rate": 2e-05,
+      "loss": 0.03662695,
+      "step": 2957
+    },
+    {
+      "epoch": 5.916,
+      "grad_norm": 1.8990249633789062,
+      "learning_rate": 2e-05,
+      "loss": 0.04180484,
+      "step": 2958
+    },
+    {
+      "epoch": 5.918,
+      "grad_norm": 1.3357110023498535,
+      "learning_rate": 2e-05,
+      "loss": 0.05098827,
+      "step": 2959
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 2.420962333679199,
+      "learning_rate": 2e-05,
+      "loss": 0.04422121,
+      "step": 2960
+    },
+    {
+      "epoch": 5.922,
+      "grad_norm": 1.0313847064971924,
+      "learning_rate": 2e-05,
+      "loss": 0.02736597,
+      "step": 2961
+    },
+    {
+      "epoch": 5.924,
+      "grad_norm": 2.0497448444366455,
+      "learning_rate": 2e-05,
+      "loss": 0.04987185,
+      "step": 2962
+    },
+    {
+      "epoch": 5.926,
+      "grad_norm": 1.508475661277771,
+      "learning_rate": 2e-05,
+      "loss": 0.02963543,
+      "step": 2963
+    },
+    {
+      "epoch": 5.928,
+      "grad_norm": 2.0732882022857666,
+      "learning_rate": 2e-05,
+      "loss": 0.04629382,
+      "step": 2964
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 1.1313819885253906,
+      "learning_rate": 2e-05,
+      "loss": 0.02837754,
+      "step": 2965
+    },
+    {
+      "epoch": 5.932,
+      "grad_norm": 1.3886733055114746,
+      "learning_rate": 2e-05,
+      "loss": 0.03276101,
+      "step": 2966
+    },
+    {
+      "epoch": 5.934,
+      "grad_norm": 1.8589578866958618,
+      "learning_rate": 2e-05,
+      "loss": 0.05601101,
+      "step": 2967
+    },
+    {
+      "epoch": 5.936,
+      "grad_norm": 1.4239290952682495,
+      "learning_rate": 2e-05,
+      "loss": 0.03772479,
+      "step": 2968
+    },
+    {
+      "epoch": 5.938,
+      "grad_norm": 1.3066622018814087,
+      "learning_rate": 2e-05,
+      "loss": 0.03739661,
+      "step": 2969
+    },
+    {
+      "epoch": 5.9399999999999995,
+      "grad_norm": 1.04325270652771,
+      "learning_rate": 2e-05,
+      "loss": 0.03995933,
+      "step": 2970
+    },
+    {
+      "epoch": 5.942,
+      "grad_norm": 1.5223791599273682,
+      "learning_rate": 2e-05,
+      "loss": 0.04225756,
+      "step": 2971
+    },
+    {
+      "epoch": 5.944,
+      "grad_norm": 2.6733899116516113,
+      "learning_rate": 2e-05,
+      "loss": 0.04441323,
+      "step": 2972
+    },
+    {
+      "epoch": 5.946,
+      "grad_norm": 1.6275644302368164,
+      "learning_rate": 2e-05,
+      "loss": 0.04762278,
+      "step": 2973
+    },
+    {
+      "epoch": 5.948,
+      "grad_norm": 1.2216103076934814,
+      "learning_rate": 2e-05,
+      "loss": 0.03107932,
+      "step": 2974
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 1.3730415105819702,
+      "learning_rate": 2e-05,
+      "loss": 0.02952298,
+      "step": 2975
+    },
+    {
+      "epoch": 5.952,
+      "grad_norm": 1.223415493965149,
+      "learning_rate": 2e-05,
+      "loss": 0.03430998,
+      "step": 2976
+    },
+    {
+      "epoch": 5.954,
+      "grad_norm": 1.5398921966552734,
+      "learning_rate": 2e-05,
+      "loss": 0.05110143,
+      "step": 2977
+    },
+    {
+      "epoch": 5.9559999999999995,
+      "grad_norm": 1.7357475757598877,
+      "learning_rate": 2e-05,
+      "loss": 0.04328895,
+      "step": 2978
+    },
+    {
+      "epoch": 5.958,
+      "grad_norm": 1.7618504762649536,
+      "learning_rate": 2e-05,
+      "loss": 0.05958952,
+      "step": 2979
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 2.2981998920440674,
+      "learning_rate": 2e-05,
+      "loss": 0.06046209,
+      "step": 2980
+    },
+    {
+      "epoch": 5.962,
+      "grad_norm": 1.7709860801696777,
+      "learning_rate": 2e-05,
+      "loss": 0.05310146,
+      "step": 2981
+    },
+    {
+      "epoch": 5.964,
+      "grad_norm": 1.4711432456970215,
+      "learning_rate": 2e-05,
+      "loss": 0.03057422,
+      "step": 2982
+    },
+    {
+      "epoch": 5.966,
+      "grad_norm": 2.1963133811950684,
+      "learning_rate": 2e-05,
+      "loss": 0.07748948,
+      "step": 2983
+    },
+    {
+      "epoch": 5.968,
+      "grad_norm": 1.2445276975631714,
+      "learning_rate": 2e-05,
+      "loss": 0.03456353,
+      "step": 2984
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 1.2059154510498047,
+      "learning_rate": 2e-05,
+      "loss": 0.03177342,
+      "step": 2985
+    },
+    {
+      "epoch": 5.9719999999999995,
+      "grad_norm": 2.007486581802368,
+      "learning_rate": 2e-05,
+      "loss": 0.03973828,
+      "step": 2986
+    },
+    {
+      "epoch": 5.974,
+      "grad_norm": 1.5894452333450317,
+      "learning_rate": 2e-05,
+      "loss": 0.04387932,
+      "step": 2987
+    },
+    {
+      "epoch": 5.976,
+      "grad_norm": 1.9093830585479736,
+      "learning_rate": 2e-05,
+      "loss": 0.04532828,
+      "step": 2988
+    },
+    {
+      "epoch": 5.978,
+      "grad_norm": 1.9382355213165283,
+      "learning_rate": 2e-05,
+      "loss": 0.05527761,
+      "step": 2989
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 1.3034459352493286,
+      "learning_rate": 2e-05,
+      "loss": 0.05401599,
+      "step": 2990
+    },
+    {
+      "epoch": 5.982,
+      "grad_norm": 1.4245580434799194,
+      "learning_rate": 2e-05,
+      "loss": 0.04059825,
+      "step": 2991
+    },
+    {
+      "epoch": 5.984,
+      "grad_norm": 2.453721523284912,
+      "learning_rate": 2e-05,
+      "loss": 0.04975076,
+      "step": 2992
+    },
+    {
+      "epoch": 5.986,
+      "grad_norm": 1.6775089502334595,
+      "learning_rate": 2e-05,
+      "loss": 0.03426765,
+      "step": 2993
+    },
+    {
+      "epoch": 5.9879999999999995,
+      "grad_norm": 1.31791090965271,
+      "learning_rate": 2e-05,
+      "loss": 0.0269012,
+      "step": 2994
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 1.2786681652069092,
+      "learning_rate": 2e-05,
+      "loss": 0.04798415,
+      "step": 2995
+    },
+    {
+      "epoch": 5.992,
+      "grad_norm": 1.375746488571167,
+      "learning_rate": 2e-05,
+      "loss": 0.03704346,
+      "step": 2996
+    },
+    {
+      "epoch": 5.994,
+      "grad_norm": 2.2422120571136475,
+      "learning_rate": 2e-05,
+      "loss": 0.03538365,
+      "step": 2997
+    },
+    {
+      "epoch": 5.996,
+      "grad_norm": 1.6498358249664307,
+      "learning_rate": 2e-05,
+      "loss": 0.05676982,
+      "step": 2998
+    },
+    {
+      "epoch": 5.998,
+      "grad_norm": 0.9439936280250549,
+      "learning_rate": 2e-05,
+      "loss": 0.03124115,
+      "step": 2999
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.866321086883545,
+      "learning_rate": 2e-05,
+      "loss": 0.04968661,
+      "step": 3000
+    },
+    {
+      "epoch": 6.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.996,
+        "AngleClassification_2": 0.996,
+        "AngleClassification_3": 0.7944111776447106,
+        "Equal_1": 0.99,
+        "Equal_2": 0.9121756487025948,
+        "Equal_3": 0.7964071856287425,
+        "LineComparison_1": 0.986,
+        "LineComparison_2": 0.9960079840319361,
+        "LineComparison_3": 0.9720558882235529,
+        "Parallel_1": 0.9819639278557114,
+        "Parallel_2": 0.9959919839679359,
+        "Parallel_3": 0.986,
+        "Perpendicular_1": 0.984,
+        "Perpendicular_2": 0.726,
+        "Perpendicular_3": 0.3717434869739479,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9976666666666667,
+        "PointLiesOnCircle_3": 0.9796,
+        "PointLiesOnLine_1": 0.9879759519038076,
+        "PointLiesOnLine_2": 0.9839679358717435,
+        "PointLiesOnLine_3": 0.7724550898203593
+      },
+      "eval_runtime": 319.7917,
+      "eval_samples_per_second": 32.834,
+      "eval_steps_per_second": 0.657,
+      "step": 3000
+    },
+    {
+      "epoch": 6.002,
+      "grad_norm": 1.5802181959152222,
+      "learning_rate": 2e-05,
+      "loss": 0.02955408,
+      "step": 3001
+    },
+    {
+      "epoch": 6.004,
+      "grad_norm": 1.1106842756271362,
+      "learning_rate": 2e-05,
+      "loss": 0.03170551,
+      "step": 3002
+    },
+    {
+      "epoch": 6.006,
+      "grad_norm": 1.0765694379806519,
+      "learning_rate": 2e-05,
+      "loss": 0.032282,
+      "step": 3003
+    },
+    {
+      "epoch": 6.008,
+      "grad_norm": 1.2182323932647705,
+      "learning_rate": 2e-05,
+      "loss": 0.04121379,
+      "step": 3004
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 1.2825361490249634,
+      "learning_rate": 2e-05,
+      "loss": 0.04134614,
+      "step": 3005
+    },
+    {
+      "epoch": 6.012,
+      "grad_norm": 2.202105760574341,
+      "learning_rate": 2e-05,
+      "loss": 0.04897365,
+      "step": 3006
+    },
+    {
+      "epoch": 6.014,
+      "grad_norm": 2.587782382965088,
+      "learning_rate": 2e-05,
+      "loss": 0.06023916,
+      "step": 3007
+    },
+    {
+      "epoch": 6.016,
+      "grad_norm": 2.555541753768921,
+      "learning_rate": 2e-05,
+      "loss": 0.03511619,
+      "step": 3008
+    },
+    {
+      "epoch": 6.018,
+      "grad_norm": 1.417189598083496,
+      "learning_rate": 2e-05,
+      "loss": 0.05029716,
+      "step": 3009
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 1.7279114723205566,
+      "learning_rate": 2e-05,
+      "loss": 0.0623187,
+      "step": 3010
+    },
+    {
+      "epoch": 6.022,
+      "grad_norm": 1.6738758087158203,
+      "learning_rate": 2e-05,
+      "loss": 0.03575193,
+      "step": 3011
+    },
+    {
+      "epoch": 6.024,
+      "grad_norm": 1.7566310167312622,
+      "learning_rate": 2e-05,
+      "loss": 0.05152703,
+      "step": 3012
+    },
+    {
+      "epoch": 6.026,
+      "grad_norm": 1.1162376403808594,
+      "learning_rate": 2e-05,
+      "loss": 0.03532304,
+      "step": 3013
+    },
+    {
+      "epoch": 6.028,
+      "grad_norm": 1.4076039791107178,
+      "learning_rate": 2e-05,
+      "loss": 0.04677567,
+      "step": 3014
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 1.133365273475647,
+      "learning_rate": 2e-05,
+      "loss": 0.02753913,
+      "step": 3015
+    },
+    {
+      "epoch": 6.032,
+      "grad_norm": 1.4458855390548706,
+      "learning_rate": 2e-05,
+      "loss": 0.03868459,
+      "step": 3016
+    },
+    {
+      "epoch": 6.034,
+      "grad_norm": 2.3904032707214355,
+      "learning_rate": 2e-05,
+      "loss": 0.03503593,
+      "step": 3017
+    },
+    {
+      "epoch": 6.036,
+      "grad_norm": 1.949507713317871,
+      "learning_rate": 2e-05,
+      "loss": 0.04720395,
+      "step": 3018
+    },
+    {
+      "epoch": 6.038,
+      "grad_norm": 1.8610894680023193,
+      "learning_rate": 2e-05,
+      "loss": 0.05176724,
+      "step": 3019
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 1.7843246459960938,
+      "learning_rate": 2e-05,
+      "loss": 0.0451156,
+      "step": 3020
+    },
+    {
+      "epoch": 6.042,
+      "grad_norm": 1.3640440702438354,
+      "learning_rate": 2e-05,
+      "loss": 0.04446036,
+      "step": 3021
+    },
+    {
+      "epoch": 6.044,
+      "grad_norm": 1.3423798084259033,
+      "learning_rate": 2e-05,
+      "loss": 0.04165779,
+      "step": 3022
+    },
+    {
+      "epoch": 6.046,
+      "grad_norm": 1.4849580526351929,
+      "learning_rate": 2e-05,
+      "loss": 0.03914479,
+      "step": 3023
+    },
+    {
+      "epoch": 6.048,
+      "grad_norm": 1.4148266315460205,
+      "learning_rate": 2e-05,
+      "loss": 0.03579354,
+      "step": 3024
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 1.197055459022522,
+      "learning_rate": 2e-05,
+      "loss": 0.04568943,
+      "step": 3025
+    },
+    {
+      "epoch": 6.052,
+      "grad_norm": 0.9507676959037781,
+      "learning_rate": 2e-05,
+      "loss": 0.03244974,
+      "step": 3026
+    },
+    {
+      "epoch": 6.054,
+      "grad_norm": 1.1743727922439575,
+      "learning_rate": 2e-05,
+      "loss": 0.04001627,
+      "step": 3027
+    },
+    {
+      "epoch": 6.056,
+      "grad_norm": 1.06550931930542,
+      "learning_rate": 2e-05,
+      "loss": 0.0358919,
+      "step": 3028
+    },
+    {
+      "epoch": 6.058,
+      "grad_norm": 1.3072738647460938,
+      "learning_rate": 2e-05,
+      "loss": 0.04308274,
+      "step": 3029
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 1.368091106414795,
+      "learning_rate": 2e-05,
+      "loss": 0.0464918,
+      "step": 3030
+    },
+    {
+      "epoch": 6.062,
+      "grad_norm": 1.4518086910247803,
+      "learning_rate": 2e-05,
+      "loss": 0.04207553,
+      "step": 3031
+    },
+    {
+      "epoch": 6.064,
+      "grad_norm": 0.9633432030677795,
+      "learning_rate": 2e-05,
+      "loss": 0.02657059,
+      "step": 3032
+    },
+    {
+      "epoch": 6.066,
+      "grad_norm": 2.7227838039398193,
+      "learning_rate": 2e-05,
+      "loss": 0.04998293,
+      "step": 3033
+    },
+    {
+      "epoch": 6.068,
+      "grad_norm": 3.5516669750213623,
+      "learning_rate": 2e-05,
+      "loss": 0.03489901,
+      "step": 3034
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 1.5370341539382935,
+      "learning_rate": 2e-05,
+      "loss": 0.04453697,
+      "step": 3035
+    },
+    {
+      "epoch": 6.072,
+      "grad_norm": 1.7417138814926147,
+      "learning_rate": 2e-05,
+      "loss": 0.03877822,
+      "step": 3036
+    },
+    {
+      "epoch": 6.074,
+      "grad_norm": 1.3443036079406738,
+      "learning_rate": 2e-05,
+      "loss": 0.03986344,
+      "step": 3037
+    },
+    {
+      "epoch": 6.076,
+      "grad_norm": 1.4446654319763184,
+      "learning_rate": 2e-05,
+      "loss": 0.03199197,
+      "step": 3038
+    },
+    {
+      "epoch": 6.078,
+      "grad_norm": 1.7269781827926636,
+      "learning_rate": 2e-05,
+      "loss": 0.04078665,
+      "step": 3039
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 1.3139877319335938,
+      "learning_rate": 2e-05,
+      "loss": 0.04576064,
+      "step": 3040
+    },
+    {
+      "epoch": 6.082,
+      "grad_norm": 1.6032078266143799,
+      "learning_rate": 2e-05,
+      "loss": 0.03815749,
+      "step": 3041
+    },
+    {
+      "epoch": 6.084,
+      "grad_norm": 2.1997110843658447,
+      "learning_rate": 2e-05,
+      "loss": 0.04005154,
+      "step": 3042
+    },
+    {
+      "epoch": 6.086,
+      "grad_norm": 1.4301249980926514,
+      "learning_rate": 2e-05,
+      "loss": 0.03681426,
+      "step": 3043
+    },
+    {
+      "epoch": 6.088,
+      "grad_norm": 0.9475594758987427,
+      "learning_rate": 2e-05,
+      "loss": 0.03235501,
+      "step": 3044
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 1.1631615161895752,
+      "learning_rate": 2e-05,
+      "loss": 0.04111622,
+      "step": 3045
+    },
+    {
+      "epoch": 6.092,
+      "grad_norm": 1.8243446350097656,
+      "learning_rate": 2e-05,
+      "loss": 0.03791495,
+      "step": 3046
+    },
+    {
+      "epoch": 6.094,
+      "grad_norm": 1.5762276649475098,
+      "learning_rate": 2e-05,
+      "loss": 0.05622587,
+      "step": 3047
+    },
+    {
+      "epoch": 6.096,
+      "grad_norm": 1.136772632598877,
+      "learning_rate": 2e-05,
+      "loss": 0.03879281,
+      "step": 3048
+    },
+    {
+      "epoch": 6.098,
+      "grad_norm": 1.1505334377288818,
+      "learning_rate": 2e-05,
+      "loss": 0.03384344,
+      "step": 3049
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 1.9480074644088745,
+      "learning_rate": 2e-05,
+      "loss": 0.03820986,
+      "step": 3050
+    },
+    {
+      "epoch": 6.102,
+      "grad_norm": 1.3107194900512695,
+      "learning_rate": 2e-05,
+      "loss": 0.03161674,
+      "step": 3051
+    },
+    {
+      "epoch": 6.104,
+      "grad_norm": 1.3566876649856567,
+      "learning_rate": 2e-05,
+      "loss": 0.0397422,
+      "step": 3052
+    },
+    {
+      "epoch": 6.106,
+      "grad_norm": 2.022648811340332,
+      "learning_rate": 2e-05,
+      "loss": 0.05766981,
+      "step": 3053
+    },
+    {
+      "epoch": 6.108,
+      "grad_norm": 1.7005928754806519,
+      "learning_rate": 2e-05,
+      "loss": 0.03167719,
+      "step": 3054
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 3.042736768722534,
+      "learning_rate": 2e-05,
+      "loss": 0.04883799,
+      "step": 3055
+    },
+    {
+      "epoch": 6.112,
+      "grad_norm": 1.171810269355774,
+      "learning_rate": 2e-05,
+      "loss": 0.03449272,
+      "step": 3056
+    },
+    {
+      "epoch": 6.114,
+      "grad_norm": 1.3196911811828613,
+      "learning_rate": 2e-05,
+      "loss": 0.03630933,
+      "step": 3057
+    },
+    {
+      "epoch": 6.116,
+      "grad_norm": 1.3632961511611938,
+      "learning_rate": 2e-05,
+      "loss": 0.02563158,
+      "step": 3058
+    },
+    {
+      "epoch": 6.118,
+      "grad_norm": 1.1939222812652588,
+      "learning_rate": 2e-05,
+      "loss": 0.03741849,
+      "step": 3059
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 1.714534044265747,
+      "learning_rate": 2e-05,
+      "loss": 0.05540673,
+      "step": 3060
+    },
+    {
+      "epoch": 6.122,
+      "grad_norm": 1.329563021659851,
+      "learning_rate": 2e-05,
+      "loss": 0.04123599,
+      "step": 3061
+    },
+    {
+      "epoch": 6.124,
+      "grad_norm": 1.8710254430770874,
+      "learning_rate": 2e-05,
+      "loss": 0.04224695,
+      "step": 3062
+    },
+    {
+      "epoch": 6.126,
+      "grad_norm": 1.2661737203598022,
+      "learning_rate": 2e-05,
+      "loss": 0.03803711,
+      "step": 3063
+    },
+    {
+      "epoch": 6.128,
+      "grad_norm": 1.2511358261108398,
+      "learning_rate": 2e-05,
+      "loss": 0.04244345,
+      "step": 3064
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 0.9187764525413513,
+      "learning_rate": 2e-05,
+      "loss": 0.03632679,
+      "step": 3065
+    },
+    {
+      "epoch": 6.132,
+      "grad_norm": 2.078164577484131,
+      "learning_rate": 2e-05,
+      "loss": 0.04006804,
+      "step": 3066
+    },
+    {
+      "epoch": 6.134,
+      "grad_norm": 1.1497598886489868,
+      "learning_rate": 2e-05,
+      "loss": 0.03985286,
+      "step": 3067
+    },
+    {
+      "epoch": 6.136,
+      "grad_norm": 1.7147870063781738,
+      "learning_rate": 2e-05,
+      "loss": 0.03666584,
+      "step": 3068
+    },
+    {
+      "epoch": 6.138,
+      "grad_norm": 1.467666506767273,
+      "learning_rate": 2e-05,
+      "loss": 0.03018375,
+      "step": 3069
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 2.5385053157806396,
+      "learning_rate": 2e-05,
+      "loss": 0.05757154,
+      "step": 3070
+    },
+    {
+      "epoch": 6.142,
+      "grad_norm": 2.1830251216888428,
+      "learning_rate": 2e-05,
+      "loss": 0.0400609,
+      "step": 3071
+    },
+    {
+      "epoch": 6.144,
+      "grad_norm": 2.2541887760162354,
+      "learning_rate": 2e-05,
+      "loss": 0.04700291,
+      "step": 3072
+    },
+    {
+      "epoch": 6.146,
+      "grad_norm": 1.2078704833984375,
+      "learning_rate": 2e-05,
+      "loss": 0.03664304,
+      "step": 3073
+    },
+    {
+      "epoch": 6.148,
+      "grad_norm": 1.6265748739242554,
+      "learning_rate": 2e-05,
+      "loss": 0.03525054,
+      "step": 3074
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 1.9453312158584595,
+      "learning_rate": 2e-05,
+      "loss": 0.05236892,
+      "step": 3075
+    },
+    {
+      "epoch": 6.152,
+      "grad_norm": 1.7163499593734741,
+      "learning_rate": 2e-05,
+      "loss": 0.03309094,
+      "step": 3076
+    },
+    {
+      "epoch": 6.154,
+      "grad_norm": 1.738598108291626,
+      "learning_rate": 2e-05,
+      "loss": 0.03554305,
+      "step": 3077
+    },
+    {
+      "epoch": 6.156,
+      "grad_norm": 1.6464965343475342,
+      "learning_rate": 2e-05,
+      "loss": 0.05227431,
+      "step": 3078
+    },
+    {
+      "epoch": 6.158,
+      "grad_norm": 2.7797718048095703,
+      "learning_rate": 2e-05,
+      "loss": 0.04286774,
+      "step": 3079
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 1.9953299760818481,
+      "learning_rate": 2e-05,
+      "loss": 0.05062944,
+      "step": 3080
+    },
+    {
+      "epoch": 6.162,
+      "grad_norm": 1.4432344436645508,
+      "learning_rate": 2e-05,
+      "loss": 0.03490534,
+      "step": 3081
+    },
+    {
+      "epoch": 6.164,
+      "grad_norm": 1.4732416868209839,
+      "learning_rate": 2e-05,
+      "loss": 0.04236253,
+      "step": 3082
+    },
+    {
+      "epoch": 6.166,
+      "grad_norm": 2.069814682006836,
+      "learning_rate": 2e-05,
+      "loss": 0.04514054,
+      "step": 3083
+    },
+    {
+      "epoch": 6.168,
+      "grad_norm": 1.8395639657974243,
+      "learning_rate": 2e-05,
+      "loss": 0.04167663,
+      "step": 3084
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 1.5798648595809937,
+      "learning_rate": 2e-05,
+      "loss": 0.05062833,
+      "step": 3085
+    },
+    {
+      "epoch": 6.172,
+      "grad_norm": 1.718724012374878,
+      "learning_rate": 2e-05,
+      "loss": 0.03674094,
+      "step": 3086
+    },
+    {
+      "epoch": 6.174,
+      "grad_norm": 1.8063639402389526,
+      "learning_rate": 2e-05,
+      "loss": 0.05062528,
+      "step": 3087
+    },
+    {
+      "epoch": 6.176,
+      "grad_norm": 1.3898078203201294,
+      "learning_rate": 2e-05,
+      "loss": 0.04505751,
+      "step": 3088
+    },
+    {
+      "epoch": 6.178,
+      "grad_norm": 1.5374902486801147,
+      "learning_rate": 2e-05,
+      "loss": 0.05387596,
+      "step": 3089
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 1.6485440731048584,
+      "learning_rate": 2e-05,
+      "loss": 0.05111238,
+      "step": 3090
+    },
+    {
+      "epoch": 6.182,
+      "grad_norm": 1.6274720430374146,
+      "learning_rate": 2e-05,
+      "loss": 0.03595363,
+      "step": 3091
+    },
+    {
+      "epoch": 6.184,
+      "grad_norm": 1.5677777528762817,
+      "learning_rate": 2e-05,
+      "loss": 0.03086212,
+      "step": 3092
+    },
+    {
+      "epoch": 6.186,
+      "grad_norm": 2.0702250003814697,
+      "learning_rate": 2e-05,
+      "loss": 0.03619816,
+      "step": 3093
+    },
+    {
+      "epoch": 6.188,
+      "grad_norm": 1.2100623846054077,
+      "learning_rate": 2e-05,
+      "loss": 0.02968932,
+      "step": 3094
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 1.6770201921463013,
+      "learning_rate": 2e-05,
+      "loss": 0.05100476,
+      "step": 3095
+    },
+    {
+      "epoch": 6.192,
+      "grad_norm": 1.4166276454925537,
+      "learning_rate": 2e-05,
+      "loss": 0.04519867,
+      "step": 3096
+    },
+    {
+      "epoch": 6.194,
+      "grad_norm": 1.882489562034607,
+      "learning_rate": 2e-05,
+      "loss": 0.04186785,
+      "step": 3097
+    },
+    {
+      "epoch": 6.196,
+      "grad_norm": 1.4164977073669434,
+      "learning_rate": 2e-05,
+      "loss": 0.04563984,
+      "step": 3098
+    },
+    {
+      "epoch": 6.198,
+      "grad_norm": 1.6581916809082031,
+      "learning_rate": 2e-05,
+      "loss": 0.04879126,
+      "step": 3099
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 1.4858304262161255,
+      "learning_rate": 2e-05,
+      "loss": 0.03413234,
+      "step": 3100
+    },
+    {
+      "epoch": 6.202,
+      "grad_norm": 1.121500849723816,
+      "learning_rate": 2e-05,
+      "loss": 0.02813966,
+      "step": 3101
+    },
+    {
+      "epoch": 6.204,
+      "grad_norm": 1.5754960775375366,
+      "learning_rate": 2e-05,
+      "loss": 0.04899345,
+      "step": 3102
+    },
+    {
+      "epoch": 6.206,
+      "grad_norm": 1.356394648551941,
+      "learning_rate": 2e-05,
+      "loss": 0.03853102,
+      "step": 3103
+    },
+    {
+      "epoch": 6.208,
+      "grad_norm": 1.0517022609710693,
+      "learning_rate": 2e-05,
+      "loss": 0.03538657,
+      "step": 3104
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 1.3921403884887695,
+      "learning_rate": 2e-05,
+      "loss": 0.03444414,
+      "step": 3105
+    },
+    {
+      "epoch": 6.212,
+      "grad_norm": 1.133500099182129,
+      "learning_rate": 2e-05,
+      "loss": 0.03595487,
+      "step": 3106
+    },
+    {
+      "epoch": 6.214,
+      "grad_norm": 1.1165953874588013,
+      "learning_rate": 2e-05,
+      "loss": 0.02767484,
+      "step": 3107
+    },
+    {
+      "epoch": 6.216,
+      "grad_norm": 2.7160420417785645,
+      "learning_rate": 2e-05,
+      "loss": 0.03668471,
+      "step": 3108
+    },
+    {
+      "epoch": 6.218,
+      "grad_norm": 1.8742766380310059,
+      "learning_rate": 2e-05,
+      "loss": 0.04762148,
+      "step": 3109
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 1.108489751815796,
+      "learning_rate": 2e-05,
+      "loss": 0.03803735,
+      "step": 3110
+    },
+    {
+      "epoch": 6.222,
+      "grad_norm": 1.4708930253982544,
+      "learning_rate": 2e-05,
+      "loss": 0.04263173,
+      "step": 3111
+    },
+    {
+      "epoch": 6.224,
+      "grad_norm": 2.706648349761963,
+      "learning_rate": 2e-05,
+      "loss": 0.05671699,
+      "step": 3112
+    },
+    {
+      "epoch": 6.226,
+      "grad_norm": 1.8484437465667725,
+      "learning_rate": 2e-05,
+      "loss": 0.02719938,
+      "step": 3113
+    },
+    {
+      "epoch": 6.228,
+      "grad_norm": 1.539387822151184,
+      "learning_rate": 2e-05,
+      "loss": 0.03141466,
+      "step": 3114
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 1.7777189016342163,
+      "learning_rate": 2e-05,
+      "loss": 0.0490169,
+      "step": 3115
+    },
+    {
+      "epoch": 6.232,
+      "grad_norm": 1.6402760744094849,
+      "learning_rate": 2e-05,
+      "loss": 0.0498919,
+      "step": 3116
+    },
+    {
+      "epoch": 6.234,
+      "grad_norm": 0.8348276019096375,
+      "learning_rate": 2e-05,
+      "loss": 0.01968635,
+      "step": 3117
+    },
+    {
+      "epoch": 6.236,
+      "grad_norm": 1.0284675359725952,
+      "learning_rate": 2e-05,
+      "loss": 0.02943014,
+      "step": 3118
+    },
+    {
+      "epoch": 6.2379999999999995,
+      "grad_norm": 1.6384459733963013,
+      "learning_rate": 2e-05,
+      "loss": 0.05340105,
+      "step": 3119
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 1.3227370977401733,
+      "learning_rate": 2e-05,
+      "loss": 0.03396504,
+      "step": 3120
+    },
+    {
+      "epoch": 6.242,
+      "grad_norm": 1.28560471534729,
+      "learning_rate": 2e-05,
+      "loss": 0.04309334,
+      "step": 3121
+    },
+    {
+      "epoch": 6.244,
+      "grad_norm": 1.1973211765289307,
+      "learning_rate": 2e-05,
+      "loss": 0.03165175,
+      "step": 3122
+    },
+    {
+      "epoch": 6.246,
+      "grad_norm": 1.492358922958374,
+      "learning_rate": 2e-05,
+      "loss": 0.04523762,
+      "step": 3123
+    },
+    {
+      "epoch": 6.248,
+      "grad_norm": 1.4751956462860107,
+      "learning_rate": 2e-05,
+      "loss": 0.04451723,
+      "step": 3124
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 1.1065033674240112,
+      "learning_rate": 2e-05,
+      "loss": 0.04365182,
+      "step": 3125
+    },
+    {
+      "epoch": 6.252,
+      "grad_norm": 1.341362714767456,
+      "learning_rate": 2e-05,
+      "loss": 0.02606768,
+      "step": 3126
+    },
+    {
+      "epoch": 6.254,
+      "grad_norm": 1.7384096384048462,
+      "learning_rate": 2e-05,
+      "loss": 0.03969992,
+      "step": 3127
+    },
+    {
+      "epoch": 6.256,
+      "grad_norm": 1.5155085325241089,
+      "learning_rate": 2e-05,
+      "loss": 0.03229823,
+      "step": 3128
+    },
+    {
+      "epoch": 6.258,
+      "grad_norm": 1.0733712911605835,
+      "learning_rate": 2e-05,
+      "loss": 0.0306533,
+      "step": 3129
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 2.0977604389190674,
+      "learning_rate": 2e-05,
+      "loss": 0.05317961,
+      "step": 3130
+    },
+    {
+      "epoch": 6.2620000000000005,
+      "grad_norm": 1.4438533782958984,
+      "learning_rate": 2e-05,
+      "loss": 0.05355418,
+      "step": 3131
+    },
+    {
+      "epoch": 6.264,
+      "grad_norm": 2.2710464000701904,
+      "learning_rate": 2e-05,
+      "loss": 0.04868567,
+      "step": 3132
+    },
+    {
+      "epoch": 6.266,
+      "grad_norm": 1.7031066417694092,
+      "learning_rate": 2e-05,
+      "loss": 0.04030811,
+      "step": 3133
+    },
+    {
+      "epoch": 6.268,
+      "grad_norm": 1.0892564058303833,
+      "learning_rate": 2e-05,
+      "loss": 0.0400371,
+      "step": 3134
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 1.327901840209961,
+      "learning_rate": 2e-05,
+      "loss": 0.05018859,
+      "step": 3135
+    },
+    {
+      "epoch": 6.272,
+      "grad_norm": 1.1799272298812866,
+      "learning_rate": 2e-05,
+      "loss": 0.03606693,
+      "step": 3136
+    },
+    {
+      "epoch": 6.274,
+      "grad_norm": 1.553155541419983,
+      "learning_rate": 2e-05,
+      "loss": 0.06281506,
+      "step": 3137
+    },
+    {
+      "epoch": 6.276,
+      "grad_norm": 1.075716257095337,
+      "learning_rate": 2e-05,
+      "loss": 0.02753101,
+      "step": 3138
+    },
+    {
+      "epoch": 6.2780000000000005,
+      "grad_norm": 2.481788396835327,
+      "learning_rate": 2e-05,
+      "loss": 0.03427035,
+      "step": 3139
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 1.606927752494812,
+      "learning_rate": 2e-05,
+      "loss": 0.03269802,
+      "step": 3140
+    },
+    {
+      "epoch": 6.282,
+      "grad_norm": 1.5751926898956299,
+      "learning_rate": 2e-05,
+      "loss": 0.03891,
+      "step": 3141
+    },
+    {
+      "epoch": 6.284,
+      "grad_norm": 2.1597113609313965,
+      "learning_rate": 2e-05,
+      "loss": 0.04709848,
+      "step": 3142
+    },
+    {
+      "epoch": 6.286,
+      "grad_norm": 1.4305531978607178,
+      "learning_rate": 2e-05,
+      "loss": 0.04866112,
+      "step": 3143
+    },
+    {
+      "epoch": 6.288,
+      "grad_norm": 1.1876941919326782,
+      "learning_rate": 2e-05,
+      "loss": 0.03890382,
+      "step": 3144
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 2.109853744506836,
+      "learning_rate": 2e-05,
+      "loss": 0.04726175,
+      "step": 3145
+    },
+    {
+      "epoch": 6.292,
+      "grad_norm": 1.5774122476577759,
+      "learning_rate": 2e-05,
+      "loss": 0.04510861,
+      "step": 3146
+    },
+    {
+      "epoch": 6.294,
+      "grad_norm": 2.0001204013824463,
+      "learning_rate": 2e-05,
+      "loss": 0.03729511,
+      "step": 3147
+    },
+    {
+      "epoch": 6.296,
+      "grad_norm": 1.9148621559143066,
+      "learning_rate": 2e-05,
+      "loss": 0.05043909,
+      "step": 3148
+    },
+    {
+      "epoch": 6.298,
+      "grad_norm": 1.77232825756073,
+      "learning_rate": 2e-05,
+      "loss": 0.05203857,
+      "step": 3149
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 1.7277973890304565,
+      "learning_rate": 2e-05,
+      "loss": 0.0506317,
+      "step": 3150
+    },
+    {
+      "epoch": 6.302,
+      "grad_norm": 1.567769169807434,
+      "learning_rate": 2e-05,
+      "loss": 0.04894867,
+      "step": 3151
+    },
+    {
+      "epoch": 6.304,
+      "grad_norm": 1.5426815748214722,
+      "learning_rate": 2e-05,
+      "loss": 0.04491549,
+      "step": 3152
+    },
+    {
+      "epoch": 6.306,
+      "grad_norm": 2.3619165420532227,
+      "learning_rate": 2e-05,
+      "loss": 0.04702245,
+      "step": 3153
+    },
+    {
+      "epoch": 6.308,
+      "grad_norm": 1.696313500404358,
+      "learning_rate": 2e-05,
+      "loss": 0.05087637,
+      "step": 3154
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 1.100569248199463,
+      "learning_rate": 2e-05,
+      "loss": 0.03571167,
+      "step": 3155
+    },
+    {
+      "epoch": 6.312,
+      "grad_norm": 1.4361978769302368,
+      "learning_rate": 2e-05,
+      "loss": 0.03643346,
+      "step": 3156
+    },
+    {
+      "epoch": 6.314,
+      "grad_norm": 1.2051647901535034,
+      "learning_rate": 2e-05,
+      "loss": 0.05133547,
+      "step": 3157
+    },
+    {
+      "epoch": 6.316,
+      "grad_norm": 1.9132866859436035,
+      "learning_rate": 2e-05,
+      "loss": 0.03743671,
+      "step": 3158
+    },
+    {
+      "epoch": 6.318,
+      "grad_norm": 2.0840513706207275,
+      "learning_rate": 2e-05,
+      "loss": 0.05485996,
+      "step": 3159
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 1.9084033966064453,
+      "learning_rate": 2e-05,
+      "loss": 0.05698278,
+      "step": 3160
+    },
+    {
+      "epoch": 6.322,
+      "grad_norm": 1.0253159999847412,
+      "learning_rate": 2e-05,
+      "loss": 0.02836904,
+      "step": 3161
+    },
+    {
+      "epoch": 6.324,
+      "grad_norm": 1.5270940065383911,
+      "learning_rate": 2e-05,
+      "loss": 0.0500167,
+      "step": 3162
+    },
+    {
+      "epoch": 6.326,
+      "grad_norm": 1.3842896223068237,
+      "learning_rate": 2e-05,
+      "loss": 0.03779293,
+      "step": 3163
+    },
+    {
+      "epoch": 6.328,
+      "grad_norm": 1.4352202415466309,
+      "learning_rate": 2e-05,
+      "loss": 0.03724783,
+      "step": 3164
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 1.2283622026443481,
+      "learning_rate": 2e-05,
+      "loss": 0.04922156,
+      "step": 3165
+    },
+    {
+      "epoch": 6.332,
+      "grad_norm": 1.2028950452804565,
+      "learning_rate": 2e-05,
+      "loss": 0.03399231,
+      "step": 3166
+    },
+    {
+      "epoch": 6.334,
+      "grad_norm": 1.8364872932434082,
+      "learning_rate": 2e-05,
+      "loss": 0.05715834,
+      "step": 3167
+    },
+    {
+      "epoch": 6.336,
+      "grad_norm": 1.926184058189392,
+      "learning_rate": 2e-05,
+      "loss": 0.052489,
+      "step": 3168
+    },
+    {
+      "epoch": 6.338,
+      "grad_norm": 2.94455623626709,
+      "learning_rate": 2e-05,
+      "loss": 0.05718009,
+      "step": 3169
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 1.064375877380371,
+      "learning_rate": 2e-05,
+      "loss": 0.03787267,
+      "step": 3170
+    },
+    {
+      "epoch": 6.342,
+      "grad_norm": 1.227028250694275,
+      "learning_rate": 2e-05,
+      "loss": 0.0388754,
+      "step": 3171
+    },
+    {
+      "epoch": 6.344,
+      "grad_norm": 1.999516487121582,
+      "learning_rate": 2e-05,
+      "loss": 0.02867444,
+      "step": 3172
+    },
+    {
+      "epoch": 6.346,
+      "grad_norm": 1.4541538953781128,
+      "learning_rate": 2e-05,
+      "loss": 0.04546948,
+      "step": 3173
+    },
+    {
+      "epoch": 6.348,
+      "grad_norm": 1.3081698417663574,
+      "learning_rate": 2e-05,
+      "loss": 0.05439358,
+      "step": 3174
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 1.7099684476852417,
+      "learning_rate": 2e-05,
+      "loss": 0.04584994,
+      "step": 3175
+    },
+    {
+      "epoch": 6.352,
+      "grad_norm": 1.3694244623184204,
+      "learning_rate": 2e-05,
+      "loss": 0.05263059,
+      "step": 3176
+    },
+    {
+      "epoch": 6.354,
+      "grad_norm": 2.0050711631774902,
+      "learning_rate": 2e-05,
+      "loss": 0.06513149,
+      "step": 3177
+    },
+    {
+      "epoch": 6.356,
+      "grad_norm": 2.2543318271636963,
+      "learning_rate": 2e-05,
+      "loss": 0.07143942,
+      "step": 3178
+    },
+    {
+      "epoch": 6.358,
+      "grad_norm": 0.9358059167861938,
+      "learning_rate": 2e-05,
+      "loss": 0.02930567,
+      "step": 3179
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 1.4434269666671753,
+      "learning_rate": 2e-05,
+      "loss": 0.02489997,
+      "step": 3180
+    },
+    {
+      "epoch": 6.362,
+      "grad_norm": 1.233762264251709,
+      "learning_rate": 2e-05,
+      "loss": 0.0335849,
+      "step": 3181
+    },
+    {
+      "epoch": 6.364,
+      "grad_norm": 1.5462418794631958,
+      "learning_rate": 2e-05,
+      "loss": 0.03041675,
+      "step": 3182
+    },
+    {
+      "epoch": 6.366,
+      "grad_norm": 2.0180463790893555,
+      "learning_rate": 2e-05,
+      "loss": 0.05690164,
+      "step": 3183
+    },
+    {
+      "epoch": 6.368,
+      "grad_norm": 1.6697211265563965,
+      "learning_rate": 2e-05,
+      "loss": 0.04623037,
+      "step": 3184
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 1.815030574798584,
+      "learning_rate": 2e-05,
+      "loss": 0.03556263,
+      "step": 3185
+    },
+    {
+      "epoch": 6.372,
+      "grad_norm": 1.826423168182373,
+      "learning_rate": 2e-05,
+      "loss": 0.04994491,
+      "step": 3186
+    },
+    {
+      "epoch": 6.374,
+      "grad_norm": 1.6917457580566406,
+      "learning_rate": 2e-05,
+      "loss": 0.03865033,
+      "step": 3187
+    },
+    {
+      "epoch": 6.376,
+      "grad_norm": 1.7738280296325684,
+      "learning_rate": 2e-05,
+      "loss": 0.04079893,
+      "step": 3188
+    },
+    {
+      "epoch": 6.378,
+      "grad_norm": 1.0954397916793823,
+      "learning_rate": 2e-05,
+      "loss": 0.03440213,
+      "step": 3189
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 1.043359637260437,
+      "learning_rate": 2e-05,
+      "loss": 0.03818577,
+      "step": 3190
+    },
+    {
+      "epoch": 6.382,
+      "grad_norm": 0.9938315153121948,
+      "learning_rate": 2e-05,
+      "loss": 0.03229547,
+      "step": 3191
+    },
+    {
+      "epoch": 6.384,
+      "grad_norm": 2.5036749839782715,
+      "learning_rate": 2e-05,
+      "loss": 0.04423948,
+      "step": 3192
+    },
+    {
+      "epoch": 6.386,
+      "grad_norm": 1.164203405380249,
+      "learning_rate": 2e-05,
+      "loss": 0.03556975,
+      "step": 3193
+    },
+    {
+      "epoch": 6.388,
+      "grad_norm": 1.4395684003829956,
+      "learning_rate": 2e-05,
+      "loss": 0.04390024,
+      "step": 3194
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 1.4073199033737183,
+      "learning_rate": 2e-05,
+      "loss": 0.02870702,
+      "step": 3195
+    },
+    {
+      "epoch": 6.392,
+      "grad_norm": 1.905430793762207,
+      "learning_rate": 2e-05,
+      "loss": 0.05451259,
+      "step": 3196
+    },
+    {
+      "epoch": 6.394,
+      "grad_norm": 2.2426390647888184,
+      "learning_rate": 2e-05,
+      "loss": 0.04760137,
+      "step": 3197
+    },
+    {
+      "epoch": 6.396,
+      "grad_norm": 1.6508835554122925,
+      "learning_rate": 2e-05,
+      "loss": 0.03574256,
+      "step": 3198
+    },
+    {
+      "epoch": 6.398,
+      "grad_norm": 1.2062972784042358,
+      "learning_rate": 2e-05,
+      "loss": 0.03633127,
+      "step": 3199
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 1.1305536031723022,
+      "learning_rate": 2e-05,
+      "loss": 0.03830415,
+      "step": 3200
+    },
+    {
+      "epoch": 6.402,
+      "grad_norm": 1.6477299928665161,
+      "learning_rate": 2e-05,
+      "loss": 0.04004838,
+      "step": 3201
+    },
+    {
+      "epoch": 6.404,
+      "grad_norm": 1.4459377527236938,
+      "learning_rate": 2e-05,
+      "loss": 0.0500464,
+      "step": 3202
+    },
+    {
+      "epoch": 6.406,
+      "grad_norm": 1.643071174621582,
+      "learning_rate": 2e-05,
+      "loss": 0.03596194,
+      "step": 3203
+    },
+    {
+      "epoch": 6.408,
+      "grad_norm": 1.4056624174118042,
+      "learning_rate": 2e-05,
+      "loss": 0.0451113,
+      "step": 3204
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 1.7545050382614136,
+      "learning_rate": 2e-05,
+      "loss": 0.04658743,
+      "step": 3205
+    },
+    {
+      "epoch": 6.412,
+      "grad_norm": 1.7296162843704224,
+      "learning_rate": 2e-05,
+      "loss": 0.04593526,
+      "step": 3206
+    },
+    {
+      "epoch": 6.414,
+      "grad_norm": 1.6219792366027832,
+      "learning_rate": 2e-05,
+      "loss": 0.0362427,
+      "step": 3207
+    },
+    {
+      "epoch": 6.416,
+      "grad_norm": 1.0432740449905396,
+      "learning_rate": 2e-05,
+      "loss": 0.03033496,
+      "step": 3208
+    },
+    {
+      "epoch": 6.418,
+      "grad_norm": 1.3447455167770386,
+      "learning_rate": 2e-05,
+      "loss": 0.04312557,
+      "step": 3209
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 1.140795350074768,
+      "learning_rate": 2e-05,
+      "loss": 0.02616213,
+      "step": 3210
+    },
+    {
+      "epoch": 6.422,
+      "grad_norm": 1.3150559663772583,
+      "learning_rate": 2e-05,
+      "loss": 0.03334409,
+      "step": 3211
+    },
+    {
+      "epoch": 6.424,
+      "grad_norm": 1.9361340999603271,
+      "learning_rate": 2e-05,
+      "loss": 0.06632831,
+      "step": 3212
+    },
+    {
+      "epoch": 6.426,
+      "grad_norm": 2.0635571479797363,
+      "learning_rate": 2e-05,
+      "loss": 0.04785598,
+      "step": 3213
+    },
+    {
+      "epoch": 6.428,
+      "grad_norm": 1.896443247795105,
+      "learning_rate": 2e-05,
+      "loss": 0.04902334,
+      "step": 3214
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 1.3984146118164062,
+      "learning_rate": 2e-05,
+      "loss": 0.03529963,
+      "step": 3215
+    },
+    {
+      "epoch": 6.432,
+      "grad_norm": 1.1594336032867432,
+      "learning_rate": 2e-05,
+      "loss": 0.03067735,
+      "step": 3216
+    },
+    {
+      "epoch": 6.434,
+      "grad_norm": 1.5906693935394287,
+      "learning_rate": 2e-05,
+      "loss": 0.04803975,
+      "step": 3217
+    },
+    {
+      "epoch": 6.436,
+      "grad_norm": 1.211396336555481,
+      "learning_rate": 2e-05,
+      "loss": 0.02445906,
+      "step": 3218
+    },
+    {
+      "epoch": 6.438,
+      "grad_norm": 1.4170849323272705,
+      "learning_rate": 2e-05,
+      "loss": 0.03652653,
+      "step": 3219
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 2.7853238582611084,
+      "learning_rate": 2e-05,
+      "loss": 0.05708388,
+      "step": 3220
+    },
+    {
+      "epoch": 6.442,
+      "grad_norm": 1.129001498222351,
+      "learning_rate": 2e-05,
+      "loss": 0.03645102,
+      "step": 3221
+    },
+    {
+      "epoch": 6.444,
+      "grad_norm": 1.4778531789779663,
+      "learning_rate": 2e-05,
+      "loss": 0.02655194,
+      "step": 3222
+    },
+    {
+      "epoch": 6.446,
+      "grad_norm": 1.6517877578735352,
+      "learning_rate": 2e-05,
+      "loss": 0.03924521,
+      "step": 3223
+    },
+    {
+      "epoch": 6.448,
+      "grad_norm": 2.0939931869506836,
+      "learning_rate": 2e-05,
+      "loss": 0.03322685,
+      "step": 3224
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 1.4540835618972778,
+      "learning_rate": 2e-05,
+      "loss": 0.04464151,
+      "step": 3225
+    },
+    {
+      "epoch": 6.452,
+      "grad_norm": 1.3042082786560059,
+      "learning_rate": 2e-05,
+      "loss": 0.04663192,
+      "step": 3226
+    },
+    {
+      "epoch": 6.454,
+      "grad_norm": 1.1038719415664673,
+      "learning_rate": 2e-05,
+      "loss": 0.03338126,
+      "step": 3227
+    },
+    {
+      "epoch": 6.456,
+      "grad_norm": 1.4457974433898926,
+      "learning_rate": 2e-05,
+      "loss": 0.03782085,
+      "step": 3228
+    },
+    {
+      "epoch": 6.458,
+      "grad_norm": 1.439251184463501,
+      "learning_rate": 2e-05,
+      "loss": 0.04930679,
+      "step": 3229
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 1.6377997398376465,
+      "learning_rate": 2e-05,
+      "loss": 0.04300382,
+      "step": 3230
+    },
+    {
+      "epoch": 6.462,
+      "grad_norm": 3.27187442779541,
+      "learning_rate": 2e-05,
+      "loss": 0.03758849,
+      "step": 3231
+    },
+    {
+      "epoch": 6.464,
+      "grad_norm": 1.257215976715088,
+      "learning_rate": 2e-05,
+      "loss": 0.03490548,
+      "step": 3232
+    },
+    {
+      "epoch": 6.466,
+      "grad_norm": 1.3594287633895874,
+      "learning_rate": 2e-05,
+      "loss": 0.03412262,
+      "step": 3233
+    },
+    {
+      "epoch": 6.468,
+      "grad_norm": 1.1700828075408936,
+      "learning_rate": 2e-05,
+      "loss": 0.03058089,
+      "step": 3234
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 0.9974095225334167,
+      "learning_rate": 2e-05,
+      "loss": 0.03268985,
+      "step": 3235
+    },
+    {
+      "epoch": 6.4719999999999995,
+      "grad_norm": 1.6861094236373901,
+      "learning_rate": 2e-05,
+      "loss": 0.04692501,
+      "step": 3236
+    },
+    {
+      "epoch": 6.474,
+      "grad_norm": 1.220467209815979,
+      "learning_rate": 2e-05,
+      "loss": 0.04217067,
+      "step": 3237
+    },
+    {
+      "epoch": 6.476,
+      "grad_norm": 1.1765286922454834,
+      "learning_rate": 2e-05,
+      "loss": 0.03455502,
+      "step": 3238
+    },
+    {
+      "epoch": 6.478,
+      "grad_norm": 1.4939866065979004,
+      "learning_rate": 2e-05,
+      "loss": 0.05220221,
+      "step": 3239
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 1.1564301252365112,
+      "learning_rate": 2e-05,
+      "loss": 0.03657191,
+      "step": 3240
+    },
+    {
+      "epoch": 6.482,
+      "grad_norm": 1.5989612340927124,
+      "learning_rate": 2e-05,
+      "loss": 0.04166878,
+      "step": 3241
+    },
+    {
+      "epoch": 6.484,
+      "grad_norm": 1.5818308591842651,
+      "learning_rate": 2e-05,
+      "loss": 0.04151236,
+      "step": 3242
+    },
+    {
+      "epoch": 6.486,
+      "grad_norm": 1.6136821508407593,
+      "learning_rate": 2e-05,
+      "loss": 0.03971895,
+      "step": 3243
+    },
+    {
+      "epoch": 6.4879999999999995,
+      "grad_norm": 1.5467586517333984,
+      "learning_rate": 2e-05,
+      "loss": 0.04718712,
+      "step": 3244
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 0.9672493934631348,
+      "learning_rate": 2e-05,
+      "loss": 0.02750623,
+      "step": 3245
+    },
+    {
+      "epoch": 6.492,
+      "grad_norm": 1.5886257886886597,
+      "learning_rate": 2e-05,
+      "loss": 0.043421,
+      "step": 3246
+    },
+    {
+      "epoch": 6.494,
+      "grad_norm": 1.9581189155578613,
+      "learning_rate": 2e-05,
+      "loss": 0.04324682,
+      "step": 3247
+    },
+    {
+      "epoch": 6.496,
+      "grad_norm": 1.1795440912246704,
+      "learning_rate": 2e-05,
+      "loss": 0.04421395,
+      "step": 3248
+    },
+    {
+      "epoch": 6.498,
+      "grad_norm": 1.8300138711929321,
+      "learning_rate": 2e-05,
+      "loss": 0.05202936,
+      "step": 3249
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 1.1608667373657227,
+      "learning_rate": 2e-05,
+      "loss": 0.03892054,
+      "step": 3250
+    },
+    {
+      "epoch": 6.502,
+      "grad_norm": 1.8854470252990723,
+      "learning_rate": 2e-05,
+      "loss": 0.05796061,
+      "step": 3251
+    },
+    {
+      "epoch": 6.504,
+      "grad_norm": 5.84080696105957,
+      "learning_rate": 2e-05,
+      "loss": 0.04915443,
+      "step": 3252
+    },
+    {
+      "epoch": 6.506,
+      "grad_norm": 1.2369056940078735,
+      "learning_rate": 2e-05,
+      "loss": 0.04068791,
+      "step": 3253
+    },
+    {
+      "epoch": 6.508,
+      "grad_norm": 2.1469545364379883,
+      "learning_rate": 2e-05,
+      "loss": 0.0260004,
+      "step": 3254
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 1.370782494544983,
+      "learning_rate": 2e-05,
+      "loss": 0.03866827,
+      "step": 3255
+    },
+    {
+      "epoch": 6.5120000000000005,
+      "grad_norm": 1.2568261623382568,
+      "learning_rate": 2e-05,
+      "loss": 0.038529,
+      "step": 3256
+    },
+    {
+      "epoch": 6.514,
+      "grad_norm": 1.186782717704773,
+      "learning_rate": 2e-05,
+      "loss": 0.04556369,
+      "step": 3257
+    },
+    {
+      "epoch": 6.516,
+      "grad_norm": 0.8793965578079224,
+      "learning_rate": 2e-05,
+      "loss": 0.02259767,
+      "step": 3258
+    },
+    {
+      "epoch": 6.518,
+      "grad_norm": 1.2058748006820679,
+      "learning_rate": 2e-05,
+      "loss": 0.03813112,
+      "step": 3259
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 1.2766646146774292,
+      "learning_rate": 2e-05,
+      "loss": 0.04175682,
+      "step": 3260
+    },
+    {
+      "epoch": 6.522,
+      "grad_norm": 1.3827567100524902,
+      "learning_rate": 2e-05,
+      "loss": 0.03931576,
+      "step": 3261
+    },
+    {
+      "epoch": 6.524,
+      "grad_norm": 1.741408109664917,
+      "learning_rate": 2e-05,
+      "loss": 0.03503319,
+      "step": 3262
+    },
+    {
+      "epoch": 6.526,
+      "grad_norm": 1.5188173055648804,
+      "learning_rate": 2e-05,
+      "loss": 0.03888669,
+      "step": 3263
+    },
+    {
+      "epoch": 6.5280000000000005,
+      "grad_norm": 1.9496771097183228,
+      "learning_rate": 2e-05,
+      "loss": 0.04322708,
+      "step": 3264
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 2.4623851776123047,
+      "learning_rate": 2e-05,
+      "loss": 0.04068457,
+      "step": 3265
+    },
+    {
+      "epoch": 6.532,
+      "grad_norm": 0.9864256381988525,
+      "learning_rate": 2e-05,
+      "loss": 0.03565934,
+      "step": 3266
+    },
+    {
+      "epoch": 6.534,
+      "grad_norm": 0.9321268796920776,
+      "learning_rate": 2e-05,
+      "loss": 0.02948448,
+      "step": 3267
+    },
+    {
+      "epoch": 6.536,
+      "grad_norm": 1.1365487575531006,
+      "learning_rate": 2e-05,
+      "loss": 0.03684638,
+      "step": 3268
+    },
+    {
+      "epoch": 6.538,
+      "grad_norm": 1.8974064588546753,
+      "learning_rate": 2e-05,
+      "loss": 0.03136589,
+      "step": 3269
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 1.6937148571014404,
+      "learning_rate": 2e-05,
+      "loss": 0.04182105,
+      "step": 3270
+    },
+    {
+      "epoch": 6.542,
+      "grad_norm": 1.4892710447311401,
+      "learning_rate": 2e-05,
+      "loss": 0.05436049,
+      "step": 3271
+    },
+    {
+      "epoch": 6.5440000000000005,
+      "grad_norm": 2.192354917526245,
+      "learning_rate": 2e-05,
+      "loss": 0.03655919,
+      "step": 3272
+    },
+    {
+      "epoch": 6.546,
+      "grad_norm": 1.3368217945098877,
+      "learning_rate": 2e-05,
+      "loss": 0.02830843,
+      "step": 3273
+    },
+    {
+      "epoch": 6.548,
+      "grad_norm": 1.422207236289978,
+      "learning_rate": 2e-05,
+      "loss": 0.02914708,
+      "step": 3274
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 3.405113935470581,
+      "learning_rate": 2e-05,
+      "loss": 0.04792117,
+      "step": 3275
+    },
+    {
+      "epoch": 6.552,
+      "grad_norm": 2.973292112350464,
+      "learning_rate": 2e-05,
+      "loss": 0.050552,
+      "step": 3276
+    },
+    {
+      "epoch": 6.554,
+      "grad_norm": 1.7082622051239014,
+      "learning_rate": 2e-05,
+      "loss": 0.03328468,
+      "step": 3277
+    },
+    {
+      "epoch": 6.556,
+      "grad_norm": 1.2330087423324585,
+      "learning_rate": 2e-05,
+      "loss": 0.0288753,
+      "step": 3278
+    },
+    {
+      "epoch": 6.558,
+      "grad_norm": 1.4605454206466675,
+      "learning_rate": 2e-05,
+      "loss": 0.03361526,
+      "step": 3279
+    },
+    {
+      "epoch": 6.5600000000000005,
+      "grad_norm": 1.288313627243042,
+      "learning_rate": 2e-05,
+      "loss": 0.04549043,
+      "step": 3280
+    },
+    {
+      "epoch": 6.562,
+      "grad_norm": 2.2067997455596924,
+      "learning_rate": 2e-05,
+      "loss": 0.0399694,
+      "step": 3281
+    },
+    {
+      "epoch": 6.564,
+      "grad_norm": 1.6156538724899292,
+      "learning_rate": 2e-05,
+      "loss": 0.07278688,
+      "step": 3282
+    },
+    {
+      "epoch": 6.566,
+      "grad_norm": 2.2792131900787354,
+      "learning_rate": 2e-05,
+      "loss": 0.05353578,
+      "step": 3283
+    },
+    {
+      "epoch": 6.568,
+      "grad_norm": 1.9242584705352783,
+      "learning_rate": 2e-05,
+      "loss": 0.0594984,
+      "step": 3284
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 1.4676686525344849,
+      "learning_rate": 2e-05,
+      "loss": 0.04927154,
+      "step": 3285
+    },
+    {
+      "epoch": 6.572,
+      "grad_norm": 1.2734441757202148,
+      "learning_rate": 2e-05,
+      "loss": 0.04294922,
+      "step": 3286
+    },
+    {
+      "epoch": 6.574,
+      "grad_norm": 1.739187240600586,
+      "learning_rate": 2e-05,
+      "loss": 0.04884008,
+      "step": 3287
+    },
+    {
+      "epoch": 6.576,
+      "grad_norm": 1.88695228099823,
+      "learning_rate": 2e-05,
+      "loss": 0.05116354,
+      "step": 3288
+    },
+    {
+      "epoch": 6.578,
+      "grad_norm": 1.28707754611969,
+      "learning_rate": 2e-05,
+      "loss": 0.04390291,
+      "step": 3289
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 1.3685258626937866,
+      "learning_rate": 2e-05,
+      "loss": 0.03953566,
+      "step": 3290
+    },
+    {
+      "epoch": 6.582,
+      "grad_norm": 1.109430193901062,
+      "learning_rate": 2e-05,
+      "loss": 0.04132858,
+      "step": 3291
+    },
+    {
+      "epoch": 6.584,
+      "grad_norm": 1.8510801792144775,
+      "learning_rate": 2e-05,
+      "loss": 0.05430602,
+      "step": 3292
+    },
+    {
+      "epoch": 6.586,
+      "grad_norm": 1.2035598754882812,
+      "learning_rate": 2e-05,
+      "loss": 0.03927475,
+      "step": 3293
+    },
+    {
+      "epoch": 6.588,
+      "grad_norm": 2.5072784423828125,
+      "learning_rate": 2e-05,
+      "loss": 0.04595596,
+      "step": 3294
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 1.213221549987793,
+      "learning_rate": 2e-05,
+      "loss": 0.04527979,
+      "step": 3295
+    },
+    {
+      "epoch": 6.592,
+      "grad_norm": 1.2507351636886597,
+      "learning_rate": 2e-05,
+      "loss": 0.03918802,
+      "step": 3296
+    },
+    {
+      "epoch": 6.594,
+      "grad_norm": 1.0590146780014038,
+      "learning_rate": 2e-05,
+      "loss": 0.03508047,
+      "step": 3297
+    },
+    {
+      "epoch": 6.596,
+      "grad_norm": 1.703151822090149,
+      "learning_rate": 2e-05,
+      "loss": 0.04405899,
+      "step": 3298
+    },
+    {
+      "epoch": 6.598,
+      "grad_norm": 1.1094061136245728,
+      "learning_rate": 2e-05,
+      "loss": 0.03539534,
+      "step": 3299
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 1.1324583292007446,
+      "learning_rate": 2e-05,
+      "loss": 0.04019186,
+      "step": 3300
+    },
+    {
+      "epoch": 6.602,
+      "grad_norm": 3.1161935329437256,
+      "learning_rate": 2e-05,
+      "loss": 0.04159467,
+      "step": 3301
+    },
+    {
+      "epoch": 6.604,
+      "grad_norm": 1.558793306350708,
+      "learning_rate": 2e-05,
+      "loss": 0.04492222,
+      "step": 3302
+    },
+    {
+      "epoch": 6.606,
+      "grad_norm": 1.1711833477020264,
+      "learning_rate": 2e-05,
+      "loss": 0.03407395,
+      "step": 3303
+    },
+    {
+      "epoch": 6.608,
+      "grad_norm": 1.5858479738235474,
+      "learning_rate": 2e-05,
+      "loss": 0.06137519,
+      "step": 3304
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 1.5749893188476562,
+      "learning_rate": 2e-05,
+      "loss": 0.06341276,
+      "step": 3305
+    },
+    {
+      "epoch": 6.612,
+      "grad_norm": 1.4319220781326294,
+      "learning_rate": 2e-05,
+      "loss": 0.03183598,
+      "step": 3306
+    },
+    {
+      "epoch": 6.614,
+      "grad_norm": 1.9067035913467407,
+      "learning_rate": 2e-05,
+      "loss": 0.03418519,
+      "step": 3307
+    },
+    {
+      "epoch": 6.616,
+      "grad_norm": 1.5378934144973755,
+      "learning_rate": 2e-05,
+      "loss": 0.05683891,
+      "step": 3308
+    },
+    {
+      "epoch": 6.618,
+      "grad_norm": 1.4987341165542603,
+      "learning_rate": 2e-05,
+      "loss": 0.0428179,
+      "step": 3309
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 0.9848178625106812,
+      "learning_rate": 2e-05,
+      "loss": 0.03192403,
+      "step": 3310
+    },
+    {
+      "epoch": 6.622,
+      "grad_norm": 1.1798021793365479,
+      "learning_rate": 2e-05,
+      "loss": 0.04244391,
+      "step": 3311
+    },
+    {
+      "epoch": 6.624,
+      "grad_norm": 1.2790571451187134,
+      "learning_rate": 2e-05,
+      "loss": 0.05533645,
+      "step": 3312
+    },
+    {
+      "epoch": 6.626,
+      "grad_norm": 1.6922457218170166,
+      "learning_rate": 2e-05,
+      "loss": 0.04536498,
+      "step": 3313
+    },
+    {
+      "epoch": 6.628,
+      "grad_norm": 1.0179157257080078,
+      "learning_rate": 2e-05,
+      "loss": 0.03572208,
+      "step": 3314
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 1.8201943635940552,
+      "learning_rate": 2e-05,
+      "loss": 0.03994879,
+      "step": 3315
+    },
+    {
+      "epoch": 6.632,
+      "grad_norm": 1.4098727703094482,
+      "learning_rate": 2e-05,
+      "loss": 0.04173999,
+      "step": 3316
+    },
+    {
+      "epoch": 6.634,
+      "grad_norm": 1.0485551357269287,
+      "learning_rate": 2e-05,
+      "loss": 0.03973451,
+      "step": 3317
+    },
+    {
+      "epoch": 6.636,
+      "grad_norm": 1.4638235569000244,
+      "learning_rate": 2e-05,
+      "loss": 0.04760775,
+      "step": 3318
+    },
+    {
+      "epoch": 6.638,
+      "grad_norm": 1.5323985815048218,
+      "learning_rate": 2e-05,
+      "loss": 0.04406105,
+      "step": 3319
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 1.126651644706726,
+      "learning_rate": 2e-05,
+      "loss": 0.03304296,
+      "step": 3320
+    },
+    {
+      "epoch": 6.642,
+      "grad_norm": 1.278250813484192,
+      "learning_rate": 2e-05,
+      "loss": 0.02966734,
+      "step": 3321
+    },
+    {
+      "epoch": 6.644,
+      "grad_norm": 1.018739938735962,
+      "learning_rate": 2e-05,
+      "loss": 0.03991082,
+      "step": 3322
+    },
+    {
+      "epoch": 6.646,
+      "grad_norm": 1.2465643882751465,
+      "learning_rate": 2e-05,
+      "loss": 0.0458714,
+      "step": 3323
+    },
+    {
+      "epoch": 6.648,
+      "grad_norm": 29.821657180786133,
+      "learning_rate": 2e-05,
+      "loss": 0.0503132,
+      "step": 3324
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 3.7819323539733887,
+      "learning_rate": 2e-05,
+      "loss": 0.04644343,
+      "step": 3325
+    },
+    {
+      "epoch": 6.652,
+      "grad_norm": 0.8932387232780457,
+      "learning_rate": 2e-05,
+      "loss": 0.02574427,
+      "step": 3326
+    },
+    {
+      "epoch": 6.654,
+      "grad_norm": 1.2899101972579956,
+      "learning_rate": 2e-05,
+      "loss": 0.0327214,
+      "step": 3327
+    },
+    {
+      "epoch": 6.656,
+      "grad_norm": 2.6038079261779785,
+      "learning_rate": 2e-05,
+      "loss": 0.05023351,
+      "step": 3328
+    },
+    {
+      "epoch": 6.658,
+      "grad_norm": 1.7134565114974976,
+      "learning_rate": 2e-05,
+      "loss": 0.04856233,
+      "step": 3329
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 1.8361741304397583,
+      "learning_rate": 2e-05,
+      "loss": 0.06476304,
+      "step": 3330
+    },
+    {
+      "epoch": 6.662,
+      "grad_norm": 1.2037334442138672,
+      "learning_rate": 2e-05,
+      "loss": 0.04044564,
+      "step": 3331
+    },
+    {
+      "epoch": 6.664,
+      "grad_norm": 1.0349493026733398,
+      "learning_rate": 2e-05,
+      "loss": 0.03656833,
+      "step": 3332
+    },
+    {
+      "epoch": 6.666,
+      "grad_norm": 2.0848727226257324,
+      "learning_rate": 2e-05,
+      "loss": 0.03080843,
+      "step": 3333
+    },
+    {
+      "epoch": 6.668,
+      "grad_norm": 1.4211640357971191,
+      "learning_rate": 2e-05,
+      "loss": 0.04820876,
+      "step": 3334
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 0.9767024517059326,
+      "learning_rate": 2e-05,
+      "loss": 0.03227487,
+      "step": 3335
+    },
+    {
+      "epoch": 6.672,
+      "grad_norm": 1.7927852869033813,
+      "learning_rate": 2e-05,
+      "loss": 0.03349268,
+      "step": 3336
+    },
+    {
+      "epoch": 6.674,
+      "grad_norm": 1.3334259986877441,
+      "learning_rate": 2e-05,
+      "loss": 0.03758476,
+      "step": 3337
+    },
+    {
+      "epoch": 6.676,
+      "grad_norm": 1.1497528553009033,
+      "learning_rate": 2e-05,
+      "loss": 0.0366753,
+      "step": 3338
+    },
+    {
+      "epoch": 6.678,
+      "grad_norm": 1.3280630111694336,
+      "learning_rate": 2e-05,
+      "loss": 0.04031434,
+      "step": 3339
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 1.5619535446166992,
+      "learning_rate": 2e-05,
+      "loss": 0.02004341,
+      "step": 3340
+    },
+    {
+      "epoch": 6.682,
+      "grad_norm": 1.5152435302734375,
+      "learning_rate": 2e-05,
+      "loss": 0.04714475,
+      "step": 3341
+    },
+    {
+      "epoch": 6.684,
+      "grad_norm": 1.263704776763916,
+      "learning_rate": 2e-05,
+      "loss": 0.04945374,
+      "step": 3342
+    },
+    {
+      "epoch": 6.686,
+      "grad_norm": 1.4276520013809204,
+      "learning_rate": 2e-05,
+      "loss": 0.03332176,
+      "step": 3343
+    },
+    {
+      "epoch": 6.688,
+      "grad_norm": 0.9606978297233582,
+      "learning_rate": 2e-05,
+      "loss": 0.03038597,
+      "step": 3344
+    },
+    {
+      "epoch": 6.6899999999999995,
+      "grad_norm": 1.5653752088546753,
+      "learning_rate": 2e-05,
+      "loss": 0.04423832,
+      "step": 3345
+    },
+    {
+      "epoch": 6.692,
+      "grad_norm": 1.6132287979125977,
+      "learning_rate": 2e-05,
+      "loss": 0.04920972,
+      "step": 3346
+    },
+    {
+      "epoch": 6.694,
+      "grad_norm": 2.4168407917022705,
+      "learning_rate": 2e-05,
+      "loss": 0.05681578,
+      "step": 3347
+    },
+    {
+      "epoch": 6.696,
+      "grad_norm": 1.8760751485824585,
+      "learning_rate": 2e-05,
+      "loss": 0.0534036,
+      "step": 3348
+    },
+    {
+      "epoch": 6.698,
+      "grad_norm": 1.2305333614349365,
+      "learning_rate": 2e-05,
+      "loss": 0.02969261,
+      "step": 3349
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 1.4671058654785156,
+      "learning_rate": 2e-05,
+      "loss": 0.02868845,
+      "step": 3350
+    },
+    {
+      "epoch": 6.702,
+      "grad_norm": 1.1077961921691895,
+      "learning_rate": 2e-05,
+      "loss": 0.0313329,
+      "step": 3351
+    },
+    {
+      "epoch": 6.704,
+      "grad_norm": 1.2289212942123413,
+      "learning_rate": 2e-05,
+      "loss": 0.04141407,
+      "step": 3352
+    },
+    {
+      "epoch": 6.7059999999999995,
+      "grad_norm": 1.0401378870010376,
+      "learning_rate": 2e-05,
+      "loss": 0.02844977,
+      "step": 3353
+    },
+    {
+      "epoch": 6.708,
+      "grad_norm": 1.6821035146713257,
+      "learning_rate": 2e-05,
+      "loss": 0.033038,
+      "step": 3354
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 1.4683276414871216,
+      "learning_rate": 2e-05,
+      "loss": 0.03478734,
+      "step": 3355
+    },
+    {
+      "epoch": 6.712,
+      "grad_norm": 0.9681516885757446,
+      "learning_rate": 2e-05,
+      "loss": 0.02334734,
+      "step": 3356
+    },
+    {
+      "epoch": 6.714,
+      "grad_norm": 1.9194122552871704,
+      "learning_rate": 2e-05,
+      "loss": 0.05143356,
+      "step": 3357
+    },
+    {
+      "epoch": 6.716,
+      "grad_norm": 1.9872151613235474,
+      "learning_rate": 2e-05,
+      "loss": 0.0525632,
+      "step": 3358
+    },
+    {
+      "epoch": 6.718,
+      "grad_norm": 1.8832703828811646,
+      "learning_rate": 2e-05,
+      "loss": 0.04416494,
+      "step": 3359
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 1.498202919960022,
+      "learning_rate": 2e-05,
+      "loss": 0.04000438,
+      "step": 3360
+    },
+    {
+      "epoch": 6.7219999999999995,
+      "grad_norm": 1.6321178674697876,
+      "learning_rate": 2e-05,
+      "loss": 0.05430613,
+      "step": 3361
+    },
+    {
+      "epoch": 6.724,
+      "grad_norm": 2.441689968109131,
+      "learning_rate": 2e-05,
+      "loss": 0.0355437,
+      "step": 3362
+    },
+    {
+      "epoch": 6.726,
+      "grad_norm": 1.513107419013977,
+      "learning_rate": 2e-05,
+      "loss": 0.03850288,
+      "step": 3363
+    },
+    {
+      "epoch": 6.728,
+      "grad_norm": 0.9349620938301086,
+      "learning_rate": 2e-05,
+      "loss": 0.02798152,
+      "step": 3364
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 1.3889509439468384,
+      "learning_rate": 2e-05,
+      "loss": 0.04935129,
+      "step": 3365
+    },
+    {
+      "epoch": 6.732,
+      "grad_norm": 1.4558771848678589,
+      "learning_rate": 2e-05,
+      "loss": 0.04684127,
+      "step": 3366
+    },
+    {
+      "epoch": 6.734,
+      "grad_norm": 1.614352822303772,
+      "learning_rate": 2e-05,
+      "loss": 0.05351394,
+      "step": 3367
+    },
+    {
+      "epoch": 6.736,
+      "grad_norm": 1.7101666927337646,
+      "learning_rate": 2e-05,
+      "loss": 0.05229837,
+      "step": 3368
+    },
+    {
+      "epoch": 6.7379999999999995,
+      "grad_norm": 1.225295901298523,
+      "learning_rate": 2e-05,
+      "loss": 0.04323405,
+      "step": 3369
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 1.0662435293197632,
+      "learning_rate": 2e-05,
+      "loss": 0.03803292,
+      "step": 3370
+    },
+    {
+      "epoch": 6.742,
+      "grad_norm": 1.4984309673309326,
+      "learning_rate": 2e-05,
+      "loss": 0.043239,
+      "step": 3371
+    },
+    {
+      "epoch": 6.744,
+      "grad_norm": 1.9732307195663452,
+      "learning_rate": 2e-05,
+      "loss": 0.06063057,
+      "step": 3372
+    },
+    {
+      "epoch": 6.746,
+      "grad_norm": 0.8619161248207092,
+      "learning_rate": 2e-05,
+      "loss": 0.02019028,
+      "step": 3373
+    },
+    {
+      "epoch": 6.748,
+      "grad_norm": 1.5696371793746948,
+      "learning_rate": 2e-05,
+      "loss": 0.04602835,
+      "step": 3374
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 1.1177750825881958,
+      "learning_rate": 2e-05,
+      "loss": 0.03560708,
+      "step": 3375
+    },
+    {
+      "epoch": 6.752,
+      "grad_norm": 0.9110558032989502,
+      "learning_rate": 2e-05,
+      "loss": 0.01933064,
+      "step": 3376
+    },
+    {
+      "epoch": 6.754,
+      "grad_norm": 1.2915785312652588,
+      "learning_rate": 2e-05,
+      "loss": 0.04963629,
+      "step": 3377
+    },
+    {
+      "epoch": 6.756,
+      "grad_norm": 1.2118951082229614,
+      "learning_rate": 2e-05,
+      "loss": 0.03098101,
+      "step": 3378
+    },
+    {
+      "epoch": 6.758,
+      "grad_norm": 1.0286513566970825,
+      "learning_rate": 2e-05,
+      "loss": 0.02711356,
+      "step": 3379
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 1.4874508380889893,
+      "learning_rate": 2e-05,
+      "loss": 0.03703025,
+      "step": 3380
+    },
+    {
+      "epoch": 6.7620000000000005,
+      "grad_norm": 1.0302430391311646,
+      "learning_rate": 2e-05,
+      "loss": 0.02999306,
+      "step": 3381
+    },
+    {
+      "epoch": 6.764,
+      "grad_norm": 1.1743347644805908,
+      "learning_rate": 2e-05,
+      "loss": 0.03301746,
+      "step": 3382
+    },
+    {
+      "epoch": 6.766,
+      "grad_norm": 1.2912507057189941,
+      "learning_rate": 2e-05,
+      "loss": 0.02964731,
+      "step": 3383
+    },
+    {
+      "epoch": 6.768,
+      "grad_norm": 1.8387466669082642,
+      "learning_rate": 2e-05,
+      "loss": 0.03587981,
+      "step": 3384
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 1.2493430376052856,
+      "learning_rate": 2e-05,
+      "loss": 0.03021685,
+      "step": 3385
+    },
+    {
+      "epoch": 6.772,
+      "grad_norm": 1.8405768871307373,
+      "learning_rate": 2e-05,
+      "loss": 0.03573044,
+      "step": 3386
+    },
+    {
+      "epoch": 6.774,
+      "grad_norm": 1.5832135677337646,
+      "learning_rate": 2e-05,
+      "loss": 0.05767541,
+      "step": 3387
+    },
+    {
+      "epoch": 6.776,
+      "grad_norm": 1.0562320947647095,
+      "learning_rate": 2e-05,
+      "loss": 0.02797987,
+      "step": 3388
+    },
+    {
+      "epoch": 6.7780000000000005,
+      "grad_norm": 1.513688325881958,
+      "learning_rate": 2e-05,
+      "loss": 0.04148609,
+      "step": 3389
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 1.0690497159957886,
+      "learning_rate": 2e-05,
+      "loss": 0.03502911,
+      "step": 3390
+    },
+    {
+      "epoch": 6.782,
+      "grad_norm": 1.8190356492996216,
+      "learning_rate": 2e-05,
+      "loss": 0.03984679,
+      "step": 3391
+    },
+    {
+      "epoch": 6.784,
+      "grad_norm": 1.2146891355514526,
+      "learning_rate": 2e-05,
+      "loss": 0.02811867,
+      "step": 3392
+    },
+    {
+      "epoch": 6.786,
+      "grad_norm": 2.19791579246521,
+      "learning_rate": 2e-05,
+      "loss": 0.02688298,
+      "step": 3393
+    },
+    {
+      "epoch": 6.788,
+      "grad_norm": 2.66430926322937,
+      "learning_rate": 2e-05,
+      "loss": 0.05289562,
+      "step": 3394
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 1.111362338066101,
+      "learning_rate": 2e-05,
+      "loss": 0.03009887,
+      "step": 3395
+    },
+    {
+      "epoch": 6.792,
+      "grad_norm": 2.333982229232788,
+      "learning_rate": 2e-05,
+      "loss": 0.0382598,
+      "step": 3396
+    },
+    {
+      "epoch": 6.7940000000000005,
+      "grad_norm": 1.1318389177322388,
+      "learning_rate": 2e-05,
+      "loss": 0.03336748,
+      "step": 3397
+    },
+    {
+      "epoch": 6.796,
+      "grad_norm": 1.1174737215042114,
+      "learning_rate": 2e-05,
+      "loss": 0.03350708,
+      "step": 3398
+    },
+    {
+      "epoch": 6.798,
+      "grad_norm": 1.5082918405532837,
+      "learning_rate": 2e-05,
+      "loss": 0.03707235,
+      "step": 3399
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 2.6029469966888428,
+      "learning_rate": 2e-05,
+      "loss": 0.0542256,
+      "step": 3400
+    },
+    {
+      "epoch": 6.802,
+      "grad_norm": 1.5858310461044312,
+      "learning_rate": 2e-05,
+      "loss": 0.03691325,
+      "step": 3401
+    },
+    {
+      "epoch": 6.804,
+      "grad_norm": 1.3726555109024048,
+      "learning_rate": 2e-05,
+      "loss": 0.04375734,
+      "step": 3402
+    },
+    {
+      "epoch": 6.806,
+      "grad_norm": 1.2137669324874878,
+      "learning_rate": 2e-05,
+      "loss": 0.03287685,
+      "step": 3403
+    },
+    {
+      "epoch": 6.808,
+      "grad_norm": 1.421502709388733,
+      "learning_rate": 2e-05,
+      "loss": 0.04229927,
+      "step": 3404
+    },
+    {
+      "epoch": 6.8100000000000005,
+      "grad_norm": 0.9239016771316528,
+      "learning_rate": 2e-05,
+      "loss": 0.02439195,
+      "step": 3405
+    },
+    {
+      "epoch": 6.812,
+      "grad_norm": 1.5839835405349731,
+      "learning_rate": 2e-05,
+      "loss": 0.04999159,
+      "step": 3406
+    },
+    {
+      "epoch": 6.814,
+      "grad_norm": 1.8794617652893066,
+      "learning_rate": 2e-05,
+      "loss": 0.04069415,
+      "step": 3407
+    },
+    {
+      "epoch": 6.816,
+      "grad_norm": 1.6794978380203247,
+      "learning_rate": 2e-05,
+      "loss": 0.042319,
+      "step": 3408
+    },
+    {
+      "epoch": 6.818,
+      "grad_norm": 4.976267337799072,
+      "learning_rate": 2e-05,
+      "loss": 0.03227082,
+      "step": 3409
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 1.4886870384216309,
+      "learning_rate": 2e-05,
+      "loss": 0.04275644,
+      "step": 3410
+    },
+    {
+      "epoch": 6.822,
+      "grad_norm": 7.621867656707764,
+      "learning_rate": 2e-05,
+      "loss": 0.04230917,
+      "step": 3411
+    },
+    {
+      "epoch": 6.824,
+      "grad_norm": 1.1310099363327026,
+      "learning_rate": 2e-05,
+      "loss": 0.038154,
+      "step": 3412
+    },
+    {
+      "epoch": 6.826,
+      "grad_norm": 0.9763595461845398,
+      "learning_rate": 2e-05,
+      "loss": 0.0353835,
+      "step": 3413
+    },
+    {
+      "epoch": 6.828,
+      "grad_norm": 1.6774576902389526,
+      "learning_rate": 2e-05,
+      "loss": 0.04004671,
+      "step": 3414
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 1.4653470516204834,
+      "learning_rate": 2e-05,
+      "loss": 0.04310292,
+      "step": 3415
+    },
+    {
+      "epoch": 6.832,
+      "grad_norm": 1.3005256652832031,
+      "learning_rate": 2e-05,
+      "loss": 0.04270222,
+      "step": 3416
+    },
+    {
+      "epoch": 6.834,
+      "grad_norm": 1.2623956203460693,
+      "learning_rate": 2e-05,
+      "loss": 0.04610048,
+      "step": 3417
+    },
+    {
+      "epoch": 6.836,
+      "grad_norm": 3.5397067070007324,
+      "learning_rate": 2e-05,
+      "loss": 0.03764736,
+      "step": 3418
+    },
+    {
+      "epoch": 6.838,
+      "grad_norm": 1.3481019735336304,
+      "learning_rate": 2e-05,
+      "loss": 0.0253664,
+      "step": 3419
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 1.1361665725708008,
+      "learning_rate": 2e-05,
+      "loss": 0.03342405,
+      "step": 3420
+    },
+    {
+      "epoch": 6.842,
+      "grad_norm": 1.141860842704773,
+      "learning_rate": 2e-05,
+      "loss": 0.03326521,
+      "step": 3421
+    },
+    {
+      "epoch": 6.844,
+      "grad_norm": 1.170899748802185,
+      "learning_rate": 2e-05,
+      "loss": 0.03952732,
+      "step": 3422
+    },
+    {
+      "epoch": 6.846,
+      "grad_norm": 1.3554216623306274,
+      "learning_rate": 2e-05,
+      "loss": 0.04645754,
+      "step": 3423
+    },
+    {
+      "epoch": 6.848,
+      "grad_norm": 1.1913604736328125,
+      "learning_rate": 2e-05,
+      "loss": 0.02899063,
+      "step": 3424
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 0.9958671927452087,
+      "learning_rate": 2e-05,
+      "loss": 0.03145525,
+      "step": 3425
+    },
+    {
+      "epoch": 6.852,
+      "grad_norm": 1.9133530855178833,
+      "learning_rate": 2e-05,
+      "loss": 0.05315136,
+      "step": 3426
+    },
+    {
+      "epoch": 6.854,
+      "grad_norm": 1.4105165004730225,
+      "learning_rate": 2e-05,
+      "loss": 0.03489037,
+      "step": 3427
+    },
+    {
+      "epoch": 6.856,
+      "grad_norm": 2.0334339141845703,
+      "learning_rate": 2e-05,
+      "loss": 0.03431166,
+      "step": 3428
+    },
+    {
+      "epoch": 6.858,
+      "grad_norm": 1.2631068229675293,
+      "learning_rate": 2e-05,
+      "loss": 0.04041663,
+      "step": 3429
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 0.889681875705719,
+      "learning_rate": 2e-05,
+      "loss": 0.02558031,
+      "step": 3430
+    },
+    {
+      "epoch": 6.862,
+      "grad_norm": 1.6193792819976807,
+      "learning_rate": 2e-05,
+      "loss": 0.04380007,
+      "step": 3431
+    },
+    {
+      "epoch": 6.864,
+      "grad_norm": 1.2849256992340088,
+      "learning_rate": 2e-05,
+      "loss": 0.03113081,
+      "step": 3432
+    },
+    {
+      "epoch": 6.866,
+      "grad_norm": 1.6333054304122925,
+      "learning_rate": 2e-05,
+      "loss": 0.02930366,
+      "step": 3433
+    },
+    {
+      "epoch": 6.868,
+      "grad_norm": 1.2340630292892456,
+      "learning_rate": 2e-05,
+      "loss": 0.03456418,
+      "step": 3434
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 1.9538198709487915,
+      "learning_rate": 2e-05,
+      "loss": 0.037586,
+      "step": 3435
+    },
+    {
+      "epoch": 6.872,
+      "grad_norm": 1.0432629585266113,
+      "learning_rate": 2e-05,
+      "loss": 0.02851261,
+      "step": 3436
+    },
+    {
+      "epoch": 6.874,
+      "grad_norm": 1.1603375673294067,
+      "learning_rate": 2e-05,
+      "loss": 0.034657,
+      "step": 3437
+    },
+    {
+      "epoch": 6.876,
+      "grad_norm": 1.1851812601089478,
+      "learning_rate": 2e-05,
+      "loss": 0.03753354,
+      "step": 3438
+    },
+    {
+      "epoch": 6.878,
+      "grad_norm": 1.6592788696289062,
+      "learning_rate": 2e-05,
+      "loss": 0.04491906,
+      "step": 3439
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 1.2274245023727417,
+      "learning_rate": 2e-05,
+      "loss": 0.0340904,
+      "step": 3440
+    },
+    {
+      "epoch": 6.882,
+      "grad_norm": 1.3614306449890137,
+      "learning_rate": 2e-05,
+      "loss": 0.0444744,
+      "step": 3441
+    },
+    {
+      "epoch": 6.884,
+      "grad_norm": 1.531777262687683,
+      "learning_rate": 2e-05,
+      "loss": 0.04471203,
+      "step": 3442
+    },
+    {
+      "epoch": 6.886,
+      "grad_norm": 1.5285937786102295,
+      "learning_rate": 2e-05,
+      "loss": 0.05250137,
+      "step": 3443
+    },
+    {
+      "epoch": 6.888,
+      "grad_norm": 1.1851426362991333,
+      "learning_rate": 2e-05,
+      "loss": 0.02804793,
+      "step": 3444
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 1.1896464824676514,
+      "learning_rate": 2e-05,
+      "loss": 0.03605233,
+      "step": 3445
+    },
+    {
+      "epoch": 6.892,
+      "grad_norm": 1.2750282287597656,
+      "learning_rate": 2e-05,
+      "loss": 0.03346939,
+      "step": 3446
+    },
+    {
+      "epoch": 6.894,
+      "grad_norm": 1.1519888639450073,
+      "learning_rate": 2e-05,
+      "loss": 0.0394849,
+      "step": 3447
+    },
+    {
+      "epoch": 6.896,
+      "grad_norm": 1.385860800743103,
+      "learning_rate": 2e-05,
+      "loss": 0.03840117,
+      "step": 3448
+    },
+    {
+      "epoch": 6.898,
+      "grad_norm": 1.8041616678237915,
+      "learning_rate": 2e-05,
+      "loss": 0.04675615,
+      "step": 3449
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 2.1062817573547363,
+      "learning_rate": 2e-05,
+      "loss": 0.06331439,
+      "step": 3450
+    },
+    {
+      "epoch": 6.902,
+      "grad_norm": 1.432571530342102,
+      "learning_rate": 2e-05,
+      "loss": 0.04114764,
+      "step": 3451
+    },
+    {
+      "epoch": 6.904,
+      "grad_norm": 1.6515777111053467,
+      "learning_rate": 2e-05,
+      "loss": 0.03684805,
+      "step": 3452
+    },
+    {
+      "epoch": 6.906,
+      "grad_norm": 1.2222999334335327,
+      "learning_rate": 2e-05,
+      "loss": 0.03101808,
+      "step": 3453
+    },
+    {
+      "epoch": 6.908,
+      "grad_norm": 1.390168309211731,
+      "learning_rate": 2e-05,
+      "loss": 0.0516277,
+      "step": 3454
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 0.9776650667190552,
+      "learning_rate": 2e-05,
+      "loss": 0.02990855,
+      "step": 3455
+    },
+    {
+      "epoch": 6.912,
+      "grad_norm": 2.2317774295806885,
+      "learning_rate": 2e-05,
+      "loss": 0.03915323,
+      "step": 3456
+    },
+    {
+      "epoch": 6.914,
+      "grad_norm": 1.5829832553863525,
+      "learning_rate": 2e-05,
+      "loss": 0.04086168,
+      "step": 3457
+    },
+    {
+      "epoch": 6.916,
+      "grad_norm": 1.0172293186187744,
+      "learning_rate": 2e-05,
+      "loss": 0.02923257,
+      "step": 3458
+    },
+    {
+      "epoch": 6.918,
+      "grad_norm": 1.2105185985565186,
+      "learning_rate": 2e-05,
+      "loss": 0.02814864,
+      "step": 3459
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 2.631154775619507,
+      "learning_rate": 2e-05,
+      "loss": 0.05200697,
+      "step": 3460
+    },
+    {
+      "epoch": 6.922,
+      "grad_norm": 1.0045074224472046,
+      "learning_rate": 2e-05,
+      "loss": 0.03085414,
+      "step": 3461
+    },
+    {
+      "epoch": 6.924,
+      "grad_norm": 1.0970081090927124,
+      "learning_rate": 2e-05,
+      "loss": 0.03664451,
+      "step": 3462
+    },
+    {
+      "epoch": 6.926,
+      "grad_norm": 1.9260581731796265,
+      "learning_rate": 2e-05,
+      "loss": 0.05451979,
+      "step": 3463
+    },
+    {
+      "epoch": 6.928,
+      "grad_norm": 1.4260892868041992,
+      "learning_rate": 2e-05,
+      "loss": 0.03202819,
+      "step": 3464
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 1.867712378501892,
+      "learning_rate": 2e-05,
+      "loss": 0.0472446,
+      "step": 3465
+    },
+    {
+      "epoch": 6.932,
+      "grad_norm": 1.2931119203567505,
+      "learning_rate": 2e-05,
+      "loss": 0.03627722,
+      "step": 3466
+    },
+    {
+      "epoch": 6.934,
+      "grad_norm": 1.267653226852417,
+      "learning_rate": 2e-05,
+      "loss": 0.04178291,
+      "step": 3467
+    },
+    {
+      "epoch": 6.936,
+      "grad_norm": 1.8989514112472534,
+      "learning_rate": 2e-05,
+      "loss": 0.06069873,
+      "step": 3468
+    },
+    {
+      "epoch": 6.938,
+      "grad_norm": 1.369830846786499,
+      "learning_rate": 2e-05,
+      "loss": 0.03865648,
+      "step": 3469
+    },
+    {
+      "epoch": 6.9399999999999995,
+      "grad_norm": 2.1145782470703125,
+      "learning_rate": 2e-05,
+      "loss": 0.03145633,
+      "step": 3470
+    },
+    {
+      "epoch": 6.942,
+      "grad_norm": 1.4967544078826904,
+      "learning_rate": 2e-05,
+      "loss": 0.04484332,
+      "step": 3471
+    },
+    {
+      "epoch": 6.944,
+      "grad_norm": 0.9439011216163635,
+      "learning_rate": 2e-05,
+      "loss": 0.0252577,
+      "step": 3472
+    },
+    {
+      "epoch": 6.946,
+      "grad_norm": 1.3101210594177246,
+      "learning_rate": 2e-05,
+      "loss": 0.02739098,
+      "step": 3473
+    },
+    {
+      "epoch": 6.948,
+      "grad_norm": 1.3791102170944214,
+      "learning_rate": 2e-05,
+      "loss": 0.03050887,
+      "step": 3474
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 1.8151313066482544,
+      "learning_rate": 2e-05,
+      "loss": 0.04466762,
+      "step": 3475
+    },
+    {
+      "epoch": 6.952,
+      "grad_norm": 1.7713474035263062,
+      "learning_rate": 2e-05,
+      "loss": 0.03592313,
+      "step": 3476
+    },
+    {
+      "epoch": 6.954,
+      "grad_norm": 1.4001665115356445,
+      "learning_rate": 2e-05,
+      "loss": 0.04347411,
+      "step": 3477
+    },
+    {
+      "epoch": 6.9559999999999995,
+      "grad_norm": 2.1821348667144775,
+      "learning_rate": 2e-05,
+      "loss": 0.04121158,
+      "step": 3478
+    },
+    {
+      "epoch": 6.958,
+      "grad_norm": 1.3329628705978394,
+      "learning_rate": 2e-05,
+      "loss": 0.0454905,
+      "step": 3479
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 2.2498857975006104,
+      "learning_rate": 2e-05,
+      "loss": 0.05597805,
+      "step": 3480
+    },
+    {
+      "epoch": 6.962,
+      "grad_norm": 1.236721396446228,
+      "learning_rate": 2e-05,
+      "loss": 0.03585602,
+      "step": 3481
+    },
+    {
+      "epoch": 6.964,
+      "grad_norm": 2.087568521499634,
+      "learning_rate": 2e-05,
+      "loss": 0.03135363,
+      "step": 3482
+    },
+    {
+      "epoch": 6.966,
+      "grad_norm": 2.696734666824341,
+      "learning_rate": 2e-05,
+      "loss": 0.05923796,
+      "step": 3483
+    },
+    {
+      "epoch": 6.968,
+      "grad_norm": 1.5622137784957886,
+      "learning_rate": 2e-05,
+      "loss": 0.04081835,
+      "step": 3484
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 1.6300699710845947,
+      "learning_rate": 2e-05,
+      "loss": 0.0384453,
+      "step": 3485
+    },
+    {
+      "epoch": 6.9719999999999995,
+      "grad_norm": 1.4294748306274414,
+      "learning_rate": 2e-05,
+      "loss": 0.05283351,
+      "step": 3486
+    },
+    {
+      "epoch": 6.974,
+      "grad_norm": 5.5012335777282715,
+      "learning_rate": 2e-05,
+      "loss": 0.04624298,
+      "step": 3487
+    },
+    {
+      "epoch": 6.976,
+      "grad_norm": 0.8324853181838989,
+      "learning_rate": 2e-05,
+      "loss": 0.0236811,
+      "step": 3488
+    },
+    {
+      "epoch": 6.978,
+      "grad_norm": 1.4482553005218506,
+      "learning_rate": 2e-05,
+      "loss": 0.04167052,
+      "step": 3489
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 1.1689647436141968,
+      "learning_rate": 2e-05,
+      "loss": 0.03899954,
+      "step": 3490
+    },
+    {
+      "epoch": 6.982,
+      "grad_norm": 1.237749695777893,
+      "learning_rate": 2e-05,
+      "loss": 0.03976803,
+      "step": 3491
+    },
+    {
+      "epoch": 6.984,
+      "grad_norm": 0.9023482799530029,
+      "learning_rate": 2e-05,
+      "loss": 0.02601608,
+      "step": 3492
+    },
+    {
+      "epoch": 6.986,
+      "grad_norm": 1.958407998085022,
+      "learning_rate": 2e-05,
+      "loss": 0.0403572,
+      "step": 3493
+    },
+    {
+      "epoch": 6.9879999999999995,
+      "grad_norm": 1.4088008403778076,
+      "learning_rate": 2e-05,
+      "loss": 0.03802837,
+      "step": 3494
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 1.656472086906433,
+      "learning_rate": 2e-05,
+      "loss": 0.04578206,
+      "step": 3495
+    },
+    {
+      "epoch": 6.992,
+      "grad_norm": 2.143359422683716,
+      "learning_rate": 2e-05,
+      "loss": 0.04628511,
+      "step": 3496
+    },
+    {
+      "epoch": 6.994,
+      "grad_norm": 1.4736688137054443,
+      "learning_rate": 2e-05,
+      "loss": 0.02740022,
+      "step": 3497
+    },
+    {
+      "epoch": 6.996,
+      "grad_norm": 1.2310930490493774,
+      "learning_rate": 2e-05,
+      "loss": 0.04610129,
+      "step": 3498
+    },
+    {
+      "epoch": 6.998,
+      "grad_norm": 1.3057410717010498,
+      "learning_rate": 2e-05,
+      "loss": 0.03868414,
+      "step": 3499
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 1.592368483543396,
+      "learning_rate": 2e-05,
+      "loss": 0.0320503,
+      "step": 3500
+    },
+    {
+      "epoch": 7.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.986,
+        "AngleClassification_2": 0.996,
+        "AngleClassification_3": 0.9021956087824351,
+        "Equal_1": 0.994,
+        "Equal_2": 0.9301397205588823,
+        "Equal_3": 0.7924151696606786,
+        "LineComparison_1": 0.998,
+        "LineComparison_2": 0.9960079840319361,
+        "LineComparison_3": 0.9920159680638723,
+        "Parallel_1": 0.9719438877755511,
+        "Parallel_2": 0.9919839679358717,
+        "Parallel_3": 0.988,
+        "Perpendicular_1": 0.982,
+        "Perpendicular_2": 0.74,
+        "Perpendicular_3": 0.3316633266533066,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9972666666666666,
+        "PointLiesOnCircle_3": 0.9936,
+        "PointLiesOnLine_1": 1.0,
+        "PointLiesOnLine_2": 0.9899799599198397,
+        "PointLiesOnLine_3": 0.8223552894211577
+      },
+      "eval_runtime": 319.1391,
+      "eval_samples_per_second": 32.901,
+      "eval_steps_per_second": 0.658,
+      "step": 3500
+    },
+    {
+      "epoch": 7.002,
+      "grad_norm": 2.186814069747925,
+      "learning_rate": 2e-05,
+      "loss": 0.06875495,
+      "step": 3501
+    },
+    {
+      "epoch": 7.004,
+      "grad_norm": 2.7851738929748535,
+      "learning_rate": 2e-05,
+      "loss": 0.07352186,
+      "step": 3502
+    },
+    {
+      "epoch": 7.006,
+      "grad_norm": 1.6021937131881714,
+      "learning_rate": 2e-05,
+      "loss": 0.04718731,
+      "step": 3503
+    },
+    {
+      "epoch": 7.008,
+      "grad_norm": 1.708878755569458,
+      "learning_rate": 2e-05,
+      "loss": 0.05183144,
+      "step": 3504
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 1.2704074382781982,
+      "learning_rate": 2e-05,
+      "loss": 0.04325004,
+      "step": 3505
+    },
+    {
+      "epoch": 7.012,
+      "grad_norm": 1.7846604585647583,
+      "learning_rate": 2e-05,
+      "loss": 0.04165163,
+      "step": 3506
+    },
+    {
+      "epoch": 7.014,
+      "grad_norm": 2.000779390335083,
+      "learning_rate": 2e-05,
+      "loss": 0.05959515,
+      "step": 3507
+    },
+    {
+      "epoch": 7.016,
+      "grad_norm": 1.2613239288330078,
+      "learning_rate": 2e-05,
+      "loss": 0.04999759,
+      "step": 3508
+    },
+    {
+      "epoch": 7.018,
+      "grad_norm": 1.419448971748352,
+      "learning_rate": 2e-05,
+      "loss": 0.05279196,
+      "step": 3509
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 1.1730986833572388,
+      "learning_rate": 2e-05,
+      "loss": 0.04814022,
+      "step": 3510
+    },
+    {
+      "epoch": 7.022,
+      "grad_norm": 1.3364038467407227,
+      "learning_rate": 2e-05,
+      "loss": 0.0524789,
+      "step": 3511
+    },
+    {
+      "epoch": 7.024,
+      "grad_norm": 1.4599989652633667,
+      "learning_rate": 2e-05,
+      "loss": 0.06457356,
+      "step": 3512
+    },
+    {
+      "epoch": 7.026,
+      "grad_norm": 1.2211647033691406,
+      "learning_rate": 2e-05,
+      "loss": 0.03781735,
+      "step": 3513
+    },
+    {
+      "epoch": 7.028,
+      "grad_norm": 1.916827917098999,
+      "learning_rate": 2e-05,
+      "loss": 0.04609166,
+      "step": 3514
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 1.3852880001068115,
+      "learning_rate": 2e-05,
+      "loss": 0.05185224,
+      "step": 3515
+    },
+    {
+      "epoch": 7.032,
+      "grad_norm": 1.9086310863494873,
+      "learning_rate": 2e-05,
+      "loss": 0.06317137,
+      "step": 3516
+    },
+    {
+      "epoch": 7.034,
+      "grad_norm": 1.3954685926437378,
+      "learning_rate": 2e-05,
+      "loss": 0.04755458,
+      "step": 3517
+    },
+    {
+      "epoch": 7.036,
+      "grad_norm": 1.2691612243652344,
+      "learning_rate": 2e-05,
+      "loss": 0.04614236,
+      "step": 3518
+    },
+    {
+      "epoch": 7.038,
+      "grad_norm": 1.438931941986084,
+      "learning_rate": 2e-05,
+      "loss": 0.06192727,
+      "step": 3519
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 1.505270004272461,
+      "learning_rate": 2e-05,
+      "loss": 0.04731307,
+      "step": 3520
+    },
+    {
+      "epoch": 7.042,
+      "grad_norm": 1.2275315523147583,
+      "learning_rate": 2e-05,
+      "loss": 0.05152331,
+      "step": 3521
+    },
+    {
+      "epoch": 7.044,
+      "grad_norm": 1.6780346632003784,
+      "learning_rate": 2e-05,
+      "loss": 0.04752352,
+      "step": 3522
+    },
+    {
+      "epoch": 7.046,
+      "grad_norm": 1.3278478384017944,
+      "learning_rate": 2e-05,
+      "loss": 0.03956543,
+      "step": 3523
+    },
+    {
+      "epoch": 7.048,
+      "grad_norm": 1.9214937686920166,
+      "learning_rate": 2e-05,
+      "loss": 0.05849922,
+      "step": 3524
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 2.2144150733947754,
+      "learning_rate": 2e-05,
+      "loss": 0.04467774,
+      "step": 3525
+    },
+    {
+      "epoch": 7.052,
+      "grad_norm": 1.5400785207748413,
+      "learning_rate": 2e-05,
+      "loss": 0.05098802,
+      "step": 3526
+    },
+    {
+      "epoch": 7.054,
+      "grad_norm": 2.3327882289886475,
+      "learning_rate": 2e-05,
+      "loss": 0.04657032,
+      "step": 3527
+    },
+    {
+      "epoch": 7.056,
+      "grad_norm": 1.3429843187332153,
+      "learning_rate": 2e-05,
+      "loss": 0.04105748,
+      "step": 3528
+    },
+    {
+      "epoch": 7.058,
+      "grad_norm": 1.845400094985962,
+      "learning_rate": 2e-05,
+      "loss": 0.04504231,
+      "step": 3529
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 2.3795065879821777,
+      "learning_rate": 2e-05,
+      "loss": 0.0596103,
+      "step": 3530
+    },
+    {
+      "epoch": 7.062,
+      "grad_norm": 1.7181967496871948,
+      "learning_rate": 2e-05,
+      "loss": 0.05508241,
+      "step": 3531
+    },
+    {
+      "epoch": 7.064,
+      "grad_norm": 2.2078285217285156,
+      "learning_rate": 2e-05,
+      "loss": 0.07361597,
+      "step": 3532
+    },
+    {
+      "epoch": 7.066,
+      "grad_norm": 1.3147023916244507,
+      "learning_rate": 2e-05,
+      "loss": 0.04743645,
+      "step": 3533
+    },
+    {
+      "epoch": 7.068,
+      "grad_norm": 1.294423222541809,
+      "learning_rate": 2e-05,
+      "loss": 0.03428509,
+      "step": 3534
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 1.5514196157455444,
+      "learning_rate": 2e-05,
+      "loss": 0.05675917,
+      "step": 3535
+    },
+    {
+      "epoch": 7.072,
+      "grad_norm": 2.4513447284698486,
+      "learning_rate": 2e-05,
+      "loss": 0.05515049,
+      "step": 3536
+    },
+    {
+      "epoch": 7.074,
+      "grad_norm": 1.598224401473999,
+      "learning_rate": 2e-05,
+      "loss": 0.07278877,
+      "step": 3537
+    },
+    {
+      "epoch": 7.076,
+      "grad_norm": 1.0842782258987427,
+      "learning_rate": 2e-05,
+      "loss": 0.04107867,
+      "step": 3538
+    },
+    {
+      "epoch": 7.078,
+      "grad_norm": 1.510959267616272,
+      "learning_rate": 2e-05,
+      "loss": 0.06507348,
+      "step": 3539
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 1.5507309436798096,
+      "learning_rate": 2e-05,
+      "loss": 0.03261951,
+      "step": 3540
+    },
+    {
+      "epoch": 7.082,
+      "grad_norm": 1.7355281114578247,
+      "learning_rate": 2e-05,
+      "loss": 0.06146067,
+      "step": 3541
+    },
+    {
+      "epoch": 7.084,
+      "grad_norm": 4.931440830230713,
+      "learning_rate": 2e-05,
+      "loss": 0.04650098,
+      "step": 3542
+    },
+    {
+      "epoch": 7.086,
+      "grad_norm": 2.3309950828552246,
+      "learning_rate": 2e-05,
+      "loss": 0.09169232,
+      "step": 3543
+    },
+    {
+      "epoch": 7.088,
+      "grad_norm": 2.2715747356414795,
+      "learning_rate": 2e-05,
+      "loss": 0.05420232,
+      "step": 3544
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 2.207563877105713,
+      "learning_rate": 2e-05,
+      "loss": 0.03945183,
+      "step": 3545
+    },
+    {
+      "epoch": 7.092,
+      "grad_norm": 1.6729447841644287,
+      "learning_rate": 2e-05,
+      "loss": 0.04845568,
+      "step": 3546
+    },
+    {
+      "epoch": 7.094,
+      "grad_norm": 1.796739101409912,
+      "learning_rate": 2e-05,
+      "loss": 0.06503285,
+      "step": 3547
+    },
+    {
+      "epoch": 7.096,
+      "grad_norm": 1.2983309030532837,
+      "learning_rate": 2e-05,
+      "loss": 0.04906647,
+      "step": 3548
+    },
+    {
+      "epoch": 7.098,
+      "grad_norm": 1.4331189393997192,
+      "learning_rate": 2e-05,
+      "loss": 0.05561874,
+      "step": 3549
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 1.5238709449768066,
+      "learning_rate": 2e-05,
+      "loss": 0.05148921,
+      "step": 3550
+    },
+    {
+      "epoch": 7.102,
+      "grad_norm": 1.2250235080718994,
+      "learning_rate": 2e-05,
+      "loss": 0.04760787,
+      "step": 3551
+    },
+    {
+      "epoch": 7.104,
+      "grad_norm": 1.3822916746139526,
+      "learning_rate": 2e-05,
+      "loss": 0.0511881,
+      "step": 3552
+    },
+    {
+      "epoch": 7.106,
+      "grad_norm": 2.5087249279022217,
+      "learning_rate": 2e-05,
+      "loss": 0.06286559,
+      "step": 3553
+    },
+    {
+      "epoch": 7.108,
+      "grad_norm": 2.8455286026000977,
+      "learning_rate": 2e-05,
+      "loss": 0.05309931,
+      "step": 3554
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 1.4749555587768555,
+      "learning_rate": 2e-05,
+      "loss": 0.05005356,
+      "step": 3555
+    },
+    {
+      "epoch": 7.112,
+      "grad_norm": 2.161928415298462,
+      "learning_rate": 2e-05,
+      "loss": 0.04809496,
+      "step": 3556
+    },
+    {
+      "epoch": 7.114,
+      "grad_norm": 1.3201868534088135,
+      "learning_rate": 2e-05,
+      "loss": 0.05066467,
+      "step": 3557
+    },
+    {
+      "epoch": 7.116,
+      "grad_norm": 1.2147846221923828,
+      "learning_rate": 2e-05,
+      "loss": 0.05296815,
+      "step": 3558
+    },
+    {
+      "epoch": 7.118,
+      "grad_norm": 1.3991172313690186,
+      "learning_rate": 2e-05,
+      "loss": 0.04677524,
+      "step": 3559
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 6.09710693359375,
+      "learning_rate": 2e-05,
+      "loss": 0.04200501,
+      "step": 3560
+    },
+    {
+      "epoch": 7.122,
+      "grad_norm": 1.4460575580596924,
+      "learning_rate": 2e-05,
+      "loss": 0.0484569,
+      "step": 3561
+    },
+    {
+      "epoch": 7.124,
+      "grad_norm": 1.2892142534255981,
+      "learning_rate": 2e-05,
+      "loss": 0.04489253,
+      "step": 3562
+    },
+    {
+      "epoch": 7.126,
+      "grad_norm": 2.244136333465576,
+      "learning_rate": 2e-05,
+      "loss": 0.05174838,
+      "step": 3563
+    },
+    {
+      "epoch": 7.128,
+      "grad_norm": 0.9121345281600952,
+      "learning_rate": 2e-05,
+      "loss": 0.02768136,
+      "step": 3564
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 1.626583218574524,
+      "learning_rate": 2e-05,
+      "loss": 0.05284789,
+      "step": 3565
+    },
+    {
+      "epoch": 7.132,
+      "grad_norm": 1.1587470769882202,
+      "learning_rate": 2e-05,
+      "loss": 0.04420138,
+      "step": 3566
+    },
+    {
+      "epoch": 7.134,
+      "grad_norm": 1.1360605955123901,
+      "learning_rate": 2e-05,
+      "loss": 0.03466136,
+      "step": 3567
+    },
+    {
+      "epoch": 7.136,
+      "grad_norm": 1.6288130283355713,
+      "learning_rate": 2e-05,
+      "loss": 0.05938084,
+      "step": 3568
+    },
+    {
+      "epoch": 7.138,
+      "grad_norm": 1.7451311349868774,
+      "learning_rate": 2e-05,
+      "loss": 0.07126134,
+      "step": 3569
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 1.0442672967910767,
+      "learning_rate": 2e-05,
+      "loss": 0.0229485,
+      "step": 3570
+    },
+    {
+      "epoch": 7.142,
+      "grad_norm": 0.9555102586746216,
+      "learning_rate": 2e-05,
+      "loss": 0.03555539,
+      "step": 3571
+    },
+    {
+      "epoch": 7.144,
+      "grad_norm": 1.9347809553146362,
+      "learning_rate": 2e-05,
+      "loss": 0.065361,
+      "step": 3572
+    },
+    {
+      "epoch": 7.146,
+      "grad_norm": 1.5342516899108887,
+      "learning_rate": 2e-05,
+      "loss": 0.05312759,
+      "step": 3573
+    },
+    {
+      "epoch": 7.148,
+      "grad_norm": 2.1047842502593994,
+      "learning_rate": 2e-05,
+      "loss": 0.04418172,
+      "step": 3574
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 1.7148213386535645,
+      "learning_rate": 2e-05,
+      "loss": 0.05144256,
+      "step": 3575
+    },
+    {
+      "epoch": 7.152,
+      "grad_norm": 2.749667167663574,
+      "learning_rate": 2e-05,
+      "loss": 0.07798346,
+      "step": 3576
+    },
+    {
+      "epoch": 7.154,
+      "grad_norm": 2.0609371662139893,
+      "learning_rate": 2e-05,
+      "loss": 0.04355539,
+      "step": 3577
+    },
+    {
+      "epoch": 7.156,
+      "grad_norm": 1.6791129112243652,
+      "learning_rate": 2e-05,
+      "loss": 0.05896265,
+      "step": 3578
+    },
+    {
+      "epoch": 7.158,
+      "grad_norm": 2.215697765350342,
+      "learning_rate": 2e-05,
+      "loss": 0.0662262,
+      "step": 3579
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 1.0952160358428955,
+      "learning_rate": 2e-05,
+      "loss": 0.03965465,
+      "step": 3580
+    },
+    {
+      "epoch": 7.162,
+      "grad_norm": 1.4443844556808472,
+      "learning_rate": 2e-05,
+      "loss": 0.041635,
+      "step": 3581
+    },
+    {
+      "epoch": 7.164,
+      "grad_norm": 3.1026885509490967,
+      "learning_rate": 2e-05,
+      "loss": 0.06438316,
+      "step": 3582
+    },
+    {
+      "epoch": 7.166,
+      "grad_norm": 1.489847183227539,
+      "learning_rate": 2e-05,
+      "loss": 0.04829738,
+      "step": 3583
+    },
+    {
+      "epoch": 7.168,
+      "grad_norm": 1.4437965154647827,
+      "learning_rate": 2e-05,
+      "loss": 0.05083038,
+      "step": 3584
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 1.0720726251602173,
+      "learning_rate": 2e-05,
+      "loss": 0.04749148,
+      "step": 3585
+    },
+    {
+      "epoch": 7.172,
+      "grad_norm": 1.5895586013793945,
+      "learning_rate": 2e-05,
+      "loss": 0.06001811,
+      "step": 3586
+    },
+    {
+      "epoch": 7.174,
+      "grad_norm": 1.490312933921814,
+      "learning_rate": 2e-05,
+      "loss": 0.03755351,
+      "step": 3587
+    },
+    {
+      "epoch": 7.176,
+      "grad_norm": 1.5626713037490845,
+      "learning_rate": 2e-05,
+      "loss": 0.05684235,
+      "step": 3588
+    },
+    {
+      "epoch": 7.178,
+      "grad_norm": 1.5374075174331665,
+      "learning_rate": 2e-05,
+      "loss": 0.05959694,
+      "step": 3589
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 1.7578250169754028,
+      "learning_rate": 2e-05,
+      "loss": 0.05503277,
+      "step": 3590
+    },
+    {
+      "epoch": 7.182,
+      "grad_norm": 2.2269232273101807,
+      "learning_rate": 2e-05,
+      "loss": 0.05974087,
+      "step": 3591
+    },
+    {
+      "epoch": 7.184,
+      "grad_norm": 2.0154407024383545,
+      "learning_rate": 2e-05,
+      "loss": 0.064426,
+      "step": 3592
+    },
+    {
+      "epoch": 7.186,
+      "grad_norm": 4.343257427215576,
+      "learning_rate": 2e-05,
+      "loss": 0.06492545,
+      "step": 3593
+    },
+    {
+      "epoch": 7.188,
+      "grad_norm": 1.5611995458602905,
+      "learning_rate": 2e-05,
+      "loss": 0.05464255,
+      "step": 3594
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 2.143812656402588,
+      "learning_rate": 2e-05,
+      "loss": 0.06049203,
+      "step": 3595
+    },
+    {
+      "epoch": 7.192,
+      "grad_norm": 1.4210190773010254,
+      "learning_rate": 2e-05,
+      "loss": 0.05442027,
+      "step": 3596
+    },
+    {
+      "epoch": 7.194,
+      "grad_norm": 1.7201104164123535,
+      "learning_rate": 2e-05,
+      "loss": 0.07043049,
+      "step": 3597
+    },
+    {
+      "epoch": 7.196,
+      "grad_norm": 1.361403226852417,
+      "learning_rate": 2e-05,
+      "loss": 0.04256427,
+      "step": 3598
+    },
+    {
+      "epoch": 7.198,
+      "grad_norm": 1.2954514026641846,
+      "learning_rate": 2e-05,
+      "loss": 0.04150357,
+      "step": 3599
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 1.6332778930664062,
+      "learning_rate": 2e-05,
+      "loss": 0.04461187,
+      "step": 3600
+    },
+    {
+      "epoch": 7.202,
+      "grad_norm": 1.5767091512680054,
+      "learning_rate": 2e-05,
+      "loss": 0.05819118,
+      "step": 3601
+    },
+    {
+      "epoch": 7.204,
+      "grad_norm": 1.3004631996154785,
+      "learning_rate": 2e-05,
+      "loss": 0.03637975,
+      "step": 3602
+    },
+    {
+      "epoch": 7.206,
+      "grad_norm": 1.3462961912155151,
+      "learning_rate": 2e-05,
+      "loss": 0.03405475,
+      "step": 3603
+    },
+    {
+      "epoch": 7.208,
+      "grad_norm": 1.1177797317504883,
+      "learning_rate": 2e-05,
+      "loss": 0.03819214,
+      "step": 3604
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 2.921354293823242,
+      "learning_rate": 2e-05,
+      "loss": 0.06285311,
+      "step": 3605
+    },
+    {
+      "epoch": 7.212,
+      "grad_norm": 1.4188235998153687,
+      "learning_rate": 2e-05,
+      "loss": 0.05607903,
+      "step": 3606
+    },
+    {
+      "epoch": 7.214,
+      "grad_norm": 2.9079253673553467,
+      "learning_rate": 2e-05,
+      "loss": 0.06680961,
+      "step": 3607
+    },
+    {
+      "epoch": 7.216,
+      "grad_norm": 1.4562344551086426,
+      "learning_rate": 2e-05,
+      "loss": 0.05460964,
+      "step": 3608
+    },
+    {
+      "epoch": 7.218,
+      "grad_norm": 1.5972447395324707,
+      "learning_rate": 2e-05,
+      "loss": 0.04806063,
+      "step": 3609
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 1.8971936702728271,
+      "learning_rate": 2e-05,
+      "loss": 0.05641398,
+      "step": 3610
+    },
+    {
+      "epoch": 7.222,
+      "grad_norm": 1.4135507345199585,
+      "learning_rate": 2e-05,
+      "loss": 0.04675736,
+      "step": 3611
+    },
+    {
+      "epoch": 7.224,
+      "grad_norm": 1.3104100227355957,
+      "learning_rate": 2e-05,
+      "loss": 0.04616308,
+      "step": 3612
+    },
+    {
+      "epoch": 7.226,
+      "grad_norm": 1.6768646240234375,
+      "learning_rate": 2e-05,
+      "loss": 0.05637186,
+      "step": 3613
+    },
+    {
+      "epoch": 7.228,
+      "grad_norm": 1.6318180561065674,
+      "learning_rate": 2e-05,
+      "loss": 0.06836526,
+      "step": 3614
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 1.0449095964431763,
+      "learning_rate": 2e-05,
+      "loss": 0.03408219,
+      "step": 3615
+    },
+    {
+      "epoch": 7.232,
+      "grad_norm": 1.6176615953445435,
+      "learning_rate": 2e-05,
+      "loss": 0.06785093,
+      "step": 3616
+    },
+    {
+      "epoch": 7.234,
+      "grad_norm": 1.3260332345962524,
+      "learning_rate": 2e-05,
+      "loss": 0.04694654,
+      "step": 3617
+    },
+    {
+      "epoch": 7.236,
+      "grad_norm": 1.478721261024475,
+      "learning_rate": 2e-05,
+      "loss": 0.04473628,
+      "step": 3618
+    },
+    {
+      "epoch": 7.2379999999999995,
+      "grad_norm": 1.9737378358840942,
+      "learning_rate": 2e-05,
+      "loss": 0.06412191,
+      "step": 3619
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 1.4525940418243408,
+      "learning_rate": 2e-05,
+      "loss": 0.04360288,
+      "step": 3620
+    },
+    {
+      "epoch": 7.242,
+      "grad_norm": 2.22672963142395,
+      "learning_rate": 2e-05,
+      "loss": 0.06661811,
+      "step": 3621
+    },
+    {
+      "epoch": 7.244,
+      "grad_norm": 1.2235952615737915,
+      "learning_rate": 2e-05,
+      "loss": 0.0464549,
+      "step": 3622
+    },
+    {
+      "epoch": 7.246,
+      "grad_norm": 1.7983734607696533,
+      "learning_rate": 2e-05,
+      "loss": 0.04726644,
+      "step": 3623
+    },
+    {
+      "epoch": 7.248,
+      "grad_norm": 1.316263198852539,
+      "learning_rate": 2e-05,
+      "loss": 0.0438324,
+      "step": 3624
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 1.3271807432174683,
+      "learning_rate": 2e-05,
+      "loss": 0.04459179,
+      "step": 3625
+    },
+    {
+      "epoch": 7.252,
+      "grad_norm": 1.7631975412368774,
+      "learning_rate": 2e-05,
+      "loss": 0.04965796,
+      "step": 3626
+    },
+    {
+      "epoch": 7.254,
+      "grad_norm": 2.7661640644073486,
+      "learning_rate": 2e-05,
+      "loss": 0.06493346,
+      "step": 3627
+    },
+    {
+      "epoch": 7.256,
+      "grad_norm": 1.1280755996704102,
+      "learning_rate": 2e-05,
+      "loss": 0.03161623,
+      "step": 3628
+    },
+    {
+      "epoch": 7.258,
+      "grad_norm": 1.7222224473953247,
+      "learning_rate": 2e-05,
+      "loss": 0.0412074,
+      "step": 3629
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 1.4843904972076416,
+      "learning_rate": 2e-05,
+      "loss": 0.04300556,
+      "step": 3630
+    },
+    {
+      "epoch": 7.2620000000000005,
+      "grad_norm": 1.4301111698150635,
+      "learning_rate": 2e-05,
+      "loss": 0.05478829,
+      "step": 3631
+    },
+    {
+      "epoch": 7.264,
+      "grad_norm": 1.6331700086593628,
+      "learning_rate": 2e-05,
+      "loss": 0.04211323,
+      "step": 3632
+    },
+    {
+      "epoch": 7.266,
+      "grad_norm": 1.3613964319229126,
+      "learning_rate": 2e-05,
+      "loss": 0.04794502,
+      "step": 3633
+    },
+    {
+      "epoch": 7.268,
+      "grad_norm": 1.2767889499664307,
+      "learning_rate": 2e-05,
+      "loss": 0.04734127,
+      "step": 3634
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 1.6509748697280884,
+      "learning_rate": 2e-05,
+      "loss": 0.05143131,
+      "step": 3635
+    },
+    {
+      "epoch": 7.272,
+      "grad_norm": 1.3792498111724854,
+      "learning_rate": 2e-05,
+      "loss": 0.04537631,
+      "step": 3636
+    },
+    {
+      "epoch": 7.274,
+      "grad_norm": 1.7637391090393066,
+      "learning_rate": 2e-05,
+      "loss": 0.04865746,
+      "step": 3637
+    },
+    {
+      "epoch": 7.276,
+      "grad_norm": 1.275904655456543,
+      "learning_rate": 2e-05,
+      "loss": 0.03691653,
+      "step": 3638
+    },
+    {
+      "epoch": 7.2780000000000005,
+      "grad_norm": 1.575995922088623,
+      "learning_rate": 2e-05,
+      "loss": 0.05559878,
+      "step": 3639
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 2.00274920463562,
+      "learning_rate": 2e-05,
+      "loss": 0.06520915,
+      "step": 3640
+    },
+    {
+      "epoch": 7.282,
+      "grad_norm": 1.6603018045425415,
+      "learning_rate": 2e-05,
+      "loss": 0.04692579,
+      "step": 3641
+    },
+    {
+      "epoch": 7.284,
+      "grad_norm": 1.4534879922866821,
+      "learning_rate": 2e-05,
+      "loss": 0.05134547,
+      "step": 3642
+    },
+    {
+      "epoch": 7.286,
+      "grad_norm": 1.8674389123916626,
+      "learning_rate": 2e-05,
+      "loss": 0.06286973,
+      "step": 3643
+    },
+    {
+      "epoch": 7.288,
+      "grad_norm": 1.2336664199829102,
+      "learning_rate": 2e-05,
+      "loss": 0.04018402,
+      "step": 3644
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 1.0961309671401978,
+      "learning_rate": 2e-05,
+      "loss": 0.03630316,
+      "step": 3645
+    },
+    {
+      "epoch": 7.292,
+      "grad_norm": 1.4161972999572754,
+      "learning_rate": 2e-05,
+      "loss": 0.04280121,
+      "step": 3646
+    },
+    {
+      "epoch": 7.294,
+      "grad_norm": 1.8841363191604614,
+      "learning_rate": 2e-05,
+      "loss": 0.06651107,
+      "step": 3647
+    },
+    {
+      "epoch": 7.296,
+      "grad_norm": 1.3848674297332764,
+      "learning_rate": 2e-05,
+      "loss": 0.05075488,
+      "step": 3648
+    },
+    {
+      "epoch": 7.298,
+      "grad_norm": 2.675525665283203,
+      "learning_rate": 2e-05,
+      "loss": 0.0590144,
+      "step": 3649
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 2.8043618202209473,
+      "learning_rate": 2e-05,
+      "loss": 0.05341516,
+      "step": 3650
+    },
+    {
+      "epoch": 7.302,
+      "grad_norm": 1.16990065574646,
+      "learning_rate": 2e-05,
+      "loss": 0.03649558,
+      "step": 3651
+    },
+    {
+      "epoch": 7.304,
+      "grad_norm": 1.3103454113006592,
+      "learning_rate": 2e-05,
+      "loss": 0.04651901,
+      "step": 3652
+    },
+    {
+      "epoch": 7.306,
+      "grad_norm": 1.39406156539917,
+      "learning_rate": 2e-05,
+      "loss": 0.04839561,
+      "step": 3653
+    },
+    {
+      "epoch": 7.308,
+      "grad_norm": 1.6239453554153442,
+      "learning_rate": 2e-05,
+      "loss": 0.05581132,
+      "step": 3654
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 1.9287298917770386,
+      "learning_rate": 2e-05,
+      "loss": 0.05162586,
+      "step": 3655
+    },
+    {
+      "epoch": 7.312,
+      "grad_norm": 1.1683954000473022,
+      "learning_rate": 2e-05,
+      "loss": 0.04523386,
+      "step": 3656
+    },
+    {
+      "epoch": 7.314,
+      "grad_norm": 1.9765461683273315,
+      "learning_rate": 2e-05,
+      "loss": 0.05748572,
+      "step": 3657
+    },
+    {
+      "epoch": 7.316,
+      "grad_norm": 1.5653101205825806,
+      "learning_rate": 2e-05,
+      "loss": 0.05275439,
+      "step": 3658
+    },
+    {
+      "epoch": 7.318,
+      "grad_norm": 1.9025769233703613,
+      "learning_rate": 2e-05,
+      "loss": 0.0513374,
+      "step": 3659
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 0.9924256801605225,
+      "learning_rate": 2e-05,
+      "loss": 0.02910082,
+      "step": 3660
+    },
+    {
+      "epoch": 7.322,
+      "grad_norm": 1.058456301689148,
+      "learning_rate": 2e-05,
+      "loss": 0.03298526,
+      "step": 3661
+    },
+    {
+      "epoch": 7.324,
+      "grad_norm": 1.0151259899139404,
+      "learning_rate": 2e-05,
+      "loss": 0.03118435,
+      "step": 3662
+    },
+    {
+      "epoch": 7.326,
+      "grad_norm": 2.0813844203948975,
+      "learning_rate": 2e-05,
+      "loss": 0.05353644,
+      "step": 3663
+    },
+    {
+      "epoch": 7.328,
+      "grad_norm": 2.497664213180542,
+      "learning_rate": 2e-05,
+      "loss": 0.07998416,
+      "step": 3664
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 1.1892470121383667,
+      "learning_rate": 2e-05,
+      "loss": 0.04133311,
+      "step": 3665
+    },
+    {
+      "epoch": 7.332,
+      "grad_norm": 1.0578644275665283,
+      "learning_rate": 2e-05,
+      "loss": 0.03139057,
+      "step": 3666
+    },
+    {
+      "epoch": 7.334,
+      "grad_norm": 1.7718684673309326,
+      "learning_rate": 2e-05,
+      "loss": 0.06689329,
+      "step": 3667
+    },
+    {
+      "epoch": 7.336,
+      "grad_norm": 2.739551305770874,
+      "learning_rate": 2e-05,
+      "loss": 0.03666856,
+      "step": 3668
+    },
+    {
+      "epoch": 7.338,
+      "grad_norm": 1.403628945350647,
+      "learning_rate": 2e-05,
+      "loss": 0.0506552,
+      "step": 3669
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 1.8418684005737305,
+      "learning_rate": 2e-05,
+      "loss": 0.05253516,
+      "step": 3670
+    },
+    {
+      "epoch": 7.342,
+      "grad_norm": 1.2462759017944336,
+      "learning_rate": 2e-05,
+      "loss": 0.04325264,
+      "step": 3671
+    },
+    {
+      "epoch": 7.344,
+      "grad_norm": 3.195005178451538,
+      "learning_rate": 2e-05,
+      "loss": 0.05133444,
+      "step": 3672
+    },
+    {
+      "epoch": 7.346,
+      "grad_norm": 1.543524146080017,
+      "learning_rate": 2e-05,
+      "loss": 0.04906403,
+      "step": 3673
+    },
+    {
+      "epoch": 7.348,
+      "grad_norm": 1.4386686086654663,
+      "learning_rate": 2e-05,
+      "loss": 0.04781891,
+      "step": 3674
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 1.7783071994781494,
+      "learning_rate": 2e-05,
+      "loss": 0.0641686,
+      "step": 3675
+    },
+    {
+      "epoch": 7.352,
+      "grad_norm": 1.090715765953064,
+      "learning_rate": 2e-05,
+      "loss": 0.03242114,
+      "step": 3676
+    },
+    {
+      "epoch": 7.354,
+      "grad_norm": 1.3116369247436523,
+      "learning_rate": 2e-05,
+      "loss": 0.03984781,
+      "step": 3677
+    },
+    {
+      "epoch": 7.356,
+      "grad_norm": 1.894168734550476,
+      "learning_rate": 2e-05,
+      "loss": 0.03056693,
+      "step": 3678
+    },
+    {
+      "epoch": 7.358,
+      "grad_norm": 1.7085542678833008,
+      "learning_rate": 2e-05,
+      "loss": 0.05443097,
+      "step": 3679
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 1.401723027229309,
+      "learning_rate": 2e-05,
+      "loss": 0.04774553,
+      "step": 3680
+    },
+    {
+      "epoch": 7.362,
+      "grad_norm": 1.4581307172775269,
+      "learning_rate": 2e-05,
+      "loss": 0.04450407,
+      "step": 3681
+    },
+    {
+      "epoch": 7.364,
+      "grad_norm": 1.6069996356964111,
+      "learning_rate": 2e-05,
+      "loss": 0.06170261,
+      "step": 3682
+    },
+    {
+      "epoch": 7.366,
+      "grad_norm": 1.324460506439209,
+      "learning_rate": 2e-05,
+      "loss": 0.04895044,
+      "step": 3683
+    },
+    {
+      "epoch": 7.368,
+      "grad_norm": 1.3643289804458618,
+      "learning_rate": 2e-05,
+      "loss": 0.04545024,
+      "step": 3684
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 1.6585203409194946,
+      "learning_rate": 2e-05,
+      "loss": 0.04801652,
+      "step": 3685
+    },
+    {
+      "epoch": 7.372,
+      "grad_norm": 1.4287261962890625,
+      "learning_rate": 2e-05,
+      "loss": 0.0614093,
+      "step": 3686
+    },
+    {
+      "epoch": 7.374,
+      "grad_norm": 1.5790996551513672,
+      "learning_rate": 2e-05,
+      "loss": 0.05381008,
+      "step": 3687
+    },
+    {
+      "epoch": 7.376,
+      "grad_norm": 2.0287680625915527,
+      "learning_rate": 2e-05,
+      "loss": 0.06601522,
+      "step": 3688
+    },
+    {
+      "epoch": 7.378,
+      "grad_norm": 1.171452283859253,
+      "learning_rate": 2e-05,
+      "loss": 0.03029294,
+      "step": 3689
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 1.8826758861541748,
+      "learning_rate": 2e-05,
+      "loss": 0.0547618,
+      "step": 3690
+    },
+    {
+      "epoch": 7.382,
+      "grad_norm": 1.565005898475647,
+      "learning_rate": 2e-05,
+      "loss": 0.0449411,
+      "step": 3691
+    },
+    {
+      "epoch": 7.384,
+      "grad_norm": 1.5869972705841064,
+      "learning_rate": 2e-05,
+      "loss": 0.04129549,
+      "step": 3692
+    },
+    {
+      "epoch": 7.386,
+      "grad_norm": 1.52235746383667,
+      "learning_rate": 2e-05,
+      "loss": 0.05102809,
+      "step": 3693
+    },
+    {
+      "epoch": 7.388,
+      "grad_norm": 1.656127691268921,
+      "learning_rate": 2e-05,
+      "loss": 0.04343271,
+      "step": 3694
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 1.1580047607421875,
+      "learning_rate": 2e-05,
+      "loss": 0.04137683,
+      "step": 3695
+    },
+    {
+      "epoch": 7.392,
+      "grad_norm": 1.3034822940826416,
+      "learning_rate": 2e-05,
+      "loss": 0.04663299,
+      "step": 3696
+    },
+    {
+      "epoch": 7.394,
+      "grad_norm": 1.391483187675476,
+      "learning_rate": 2e-05,
+      "loss": 0.04537113,
+      "step": 3697
+    },
+    {
+      "epoch": 7.396,
+      "grad_norm": 1.4313862323760986,
+      "learning_rate": 2e-05,
+      "loss": 0.04555268,
+      "step": 3698
+    },
+    {
+      "epoch": 7.398,
+      "grad_norm": 3.1390175819396973,
+      "learning_rate": 2e-05,
+      "loss": 0.06643859,
+      "step": 3699
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 1.6369024515151978,
+      "learning_rate": 2e-05,
+      "loss": 0.05245568,
+      "step": 3700
+    },
+    {
+      "epoch": 7.402,
+      "grad_norm": 1.2686548233032227,
+      "learning_rate": 2e-05,
+      "loss": 0.04974839,
+      "step": 3701
+    },
+    {
+      "epoch": 7.404,
+      "grad_norm": 1.2958829402923584,
+      "learning_rate": 2e-05,
+      "loss": 0.04743432,
+      "step": 3702
+    },
+    {
+      "epoch": 7.406,
+      "grad_norm": 1.9894742965698242,
+      "learning_rate": 2e-05,
+      "loss": 0.04571943,
+      "step": 3703
+    },
+    {
+      "epoch": 7.408,
+      "grad_norm": 1.0856343507766724,
+      "learning_rate": 2e-05,
+      "loss": 0.03500437,
+      "step": 3704
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 1.949939250946045,
+      "learning_rate": 2e-05,
+      "loss": 0.05323801,
+      "step": 3705
+    },
+    {
+      "epoch": 7.412,
+      "grad_norm": 1.4476184844970703,
+      "learning_rate": 2e-05,
+      "loss": 0.04785695,
+      "step": 3706
+    },
+    {
+      "epoch": 7.414,
+      "grad_norm": 3.4324965476989746,
+      "learning_rate": 2e-05,
+      "loss": 0.06573795,
+      "step": 3707
+    },
+    {
+      "epoch": 7.416,
+      "grad_norm": 1.3931694030761719,
+      "learning_rate": 2e-05,
+      "loss": 0.04984997,
+      "step": 3708
+    },
+    {
+      "epoch": 7.418,
+      "grad_norm": 1.56455659866333,
+      "learning_rate": 2e-05,
+      "loss": 0.05046746,
+      "step": 3709
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 1.200805425643921,
+      "learning_rate": 2e-05,
+      "loss": 0.03915904,
+      "step": 3710
+    },
+    {
+      "epoch": 7.422,
+      "grad_norm": 1.366477608680725,
+      "learning_rate": 2e-05,
+      "loss": 0.04656648,
+      "step": 3711
+    },
+    {
+      "epoch": 7.424,
+      "grad_norm": 1.8359090089797974,
+      "learning_rate": 2e-05,
+      "loss": 0.05103552,
+      "step": 3712
+    },
+    {
+      "epoch": 7.426,
+      "grad_norm": 1.4380766153335571,
+      "learning_rate": 2e-05,
+      "loss": 0.04777982,
+      "step": 3713
+    },
+    {
+      "epoch": 7.428,
+      "grad_norm": 1.0222777128219604,
+      "learning_rate": 2e-05,
+      "loss": 0.03150903,
+      "step": 3714
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 2.749195098876953,
+      "learning_rate": 2e-05,
+      "loss": 0.05602058,
+      "step": 3715
+    },
+    {
+      "epoch": 7.432,
+      "grad_norm": 1.7111470699310303,
+      "learning_rate": 2e-05,
+      "loss": 0.04583771,
+      "step": 3716
+    },
+    {
+      "epoch": 7.434,
+      "grad_norm": 1.3404757976531982,
+      "learning_rate": 2e-05,
+      "loss": 0.04275534,
+      "step": 3717
+    },
+    {
+      "epoch": 7.436,
+      "grad_norm": 1.210221529006958,
+      "learning_rate": 2e-05,
+      "loss": 0.04058505,
+      "step": 3718
+    },
+    {
+      "epoch": 7.438,
+      "grad_norm": 2.068676471710205,
+      "learning_rate": 2e-05,
+      "loss": 0.04987239,
+      "step": 3719
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 1.6463974714279175,
+      "learning_rate": 2e-05,
+      "loss": 0.0489587,
+      "step": 3720
+    },
+    {
+      "epoch": 7.442,
+      "grad_norm": 0.9529186487197876,
+      "learning_rate": 2e-05,
+      "loss": 0.02814676,
+      "step": 3721
+    },
+    {
+      "epoch": 7.444,
+      "grad_norm": 1.097665548324585,
+      "learning_rate": 2e-05,
+      "loss": 0.0485144,
+      "step": 3722
+    },
+    {
+      "epoch": 7.446,
+      "grad_norm": 1.5147879123687744,
+      "learning_rate": 2e-05,
+      "loss": 0.04698938,
+      "step": 3723
+    },
+    {
+      "epoch": 7.448,
+      "grad_norm": 1.2705267667770386,
+      "learning_rate": 2e-05,
+      "loss": 0.03390349,
+      "step": 3724
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 2.0715646743774414,
+      "learning_rate": 2e-05,
+      "loss": 0.07898147,
+      "step": 3725
+    },
+    {
+      "epoch": 7.452,
+      "grad_norm": 1.3505516052246094,
+      "learning_rate": 2e-05,
+      "loss": 0.05181169,
+      "step": 3726
+    },
+    {
+      "epoch": 7.454,
+      "grad_norm": 1.6015242338180542,
+      "learning_rate": 2e-05,
+      "loss": 0.05964314,
+      "step": 3727
+    },
+    {
+      "epoch": 7.456,
+      "grad_norm": 1.358168601989746,
+      "learning_rate": 2e-05,
+      "loss": 0.04774554,
+      "step": 3728
+    },
+    {
+      "epoch": 7.458,
+      "grad_norm": 1.53998863697052,
+      "learning_rate": 2e-05,
+      "loss": 0.04986886,
+      "step": 3729
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 1.2373335361480713,
+      "learning_rate": 2e-05,
+      "loss": 0.05145576,
+      "step": 3730
+    },
+    {
+      "epoch": 7.462,
+      "grad_norm": 1.2821953296661377,
+      "learning_rate": 2e-05,
+      "loss": 0.05637913,
+      "step": 3731
+    },
+    {
+      "epoch": 7.464,
+      "grad_norm": 1.3105919361114502,
+      "learning_rate": 2e-05,
+      "loss": 0.04199533,
+      "step": 3732
+    },
+    {
+      "epoch": 7.466,
+      "grad_norm": 1.5130268335342407,
+      "learning_rate": 2e-05,
+      "loss": 0.04699155,
+      "step": 3733
+    },
+    {
+      "epoch": 7.468,
+      "grad_norm": 1.737265706062317,
+      "learning_rate": 2e-05,
+      "loss": 0.03969675,
+      "step": 3734
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 1.4570841789245605,
+      "learning_rate": 2e-05,
+      "loss": 0.04642572,
+      "step": 3735
+    },
+    {
+      "epoch": 7.4719999999999995,
+      "grad_norm": 0.9005123376846313,
+      "learning_rate": 2e-05,
+      "loss": 0.03132126,
+      "step": 3736
+    },
+    {
+      "epoch": 7.474,
+      "grad_norm": 1.973960041999817,
+      "learning_rate": 2e-05,
+      "loss": 0.05362177,
+      "step": 3737
+    },
+    {
+      "epoch": 7.476,
+      "grad_norm": 2.2350363731384277,
+      "learning_rate": 2e-05,
+      "loss": 0.06357549,
+      "step": 3738
+    },
+    {
+      "epoch": 7.478,
+      "grad_norm": 1.4672166109085083,
+      "learning_rate": 2e-05,
+      "loss": 0.04843113,
+      "step": 3739
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 1.9889647960662842,
+      "learning_rate": 2e-05,
+      "loss": 0.04717405,
+      "step": 3740
+    },
+    {
+      "epoch": 7.482,
+      "grad_norm": 1.1756142377853394,
+      "learning_rate": 2e-05,
+      "loss": 0.04274727,
+      "step": 3741
+    },
+    {
+      "epoch": 7.484,
+      "grad_norm": 1.004916787147522,
+      "learning_rate": 2e-05,
+      "loss": 0.03753413,
+      "step": 3742
+    },
+    {
+      "epoch": 7.486,
+      "grad_norm": 1.277570128440857,
+      "learning_rate": 2e-05,
+      "loss": 0.04001706,
+      "step": 3743
+    },
+    {
+      "epoch": 7.4879999999999995,
+      "grad_norm": 1.0165624618530273,
+      "learning_rate": 2e-05,
+      "loss": 0.03563813,
+      "step": 3744
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 1.3447015285491943,
+      "learning_rate": 2e-05,
+      "loss": 0.04556485,
+      "step": 3745
+    },
+    {
+      "epoch": 7.492,
+      "grad_norm": 1.9407835006713867,
+      "learning_rate": 2e-05,
+      "loss": 0.05292122,
+      "step": 3746
+    },
+    {
+      "epoch": 7.494,
+      "grad_norm": 4.274045467376709,
+      "learning_rate": 2e-05,
+      "loss": 0.04551305,
+      "step": 3747
+    },
+    {
+      "epoch": 7.496,
+      "grad_norm": 1.2199167013168335,
+      "learning_rate": 2e-05,
+      "loss": 0.04236592,
+      "step": 3748
+    },
+    {
+      "epoch": 7.498,
+      "grad_norm": 1.2471674680709839,
+      "learning_rate": 2e-05,
+      "loss": 0.03323042,
+      "step": 3749
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 1.1750595569610596,
+      "learning_rate": 2e-05,
+      "loss": 0.0397836,
+      "step": 3750
+    },
+    {
+      "epoch": 7.502,
+      "grad_norm": 1.7069101333618164,
+      "learning_rate": 2e-05,
+      "loss": 0.04767216,
+      "step": 3751
+    },
+    {
+      "epoch": 7.504,
+      "grad_norm": 1.3713217973709106,
+      "learning_rate": 2e-05,
+      "loss": 0.04113061,
+      "step": 3752
+    },
+    {
+      "epoch": 7.506,
+      "grad_norm": 1.2698354721069336,
+      "learning_rate": 2e-05,
+      "loss": 0.04726923,
+      "step": 3753
+    },
+    {
+      "epoch": 7.508,
+      "grad_norm": 1.3195385932922363,
+      "learning_rate": 2e-05,
+      "loss": 0.04144061,
+      "step": 3754
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 1.405106782913208,
+      "learning_rate": 2e-05,
+      "loss": 0.04036184,
+      "step": 3755
+    },
+    {
+      "epoch": 7.5120000000000005,
+      "grad_norm": 1.3161296844482422,
+      "learning_rate": 2e-05,
+      "loss": 0.03146423,
+      "step": 3756
+    },
+    {
+      "epoch": 7.514,
+      "grad_norm": 1.8844631910324097,
+      "learning_rate": 2e-05,
+      "loss": 0.06133197,
+      "step": 3757
+    },
+    {
+      "epoch": 7.516,
+      "grad_norm": 2.0802526473999023,
+      "learning_rate": 2e-05,
+      "loss": 0.04892286,
+      "step": 3758
+    },
+    {
+      "epoch": 7.518,
+      "grad_norm": 3.128467082977295,
+      "learning_rate": 2e-05,
+      "loss": 0.06032344,
+      "step": 3759
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 2.1159846782684326,
+      "learning_rate": 2e-05,
+      "loss": 0.07681625,
+      "step": 3760
+    },
+    {
+      "epoch": 7.522,
+      "grad_norm": 1.6027671098709106,
+      "learning_rate": 2e-05,
+      "loss": 0.05662168,
+      "step": 3761
+    },
+    {
+      "epoch": 7.524,
+      "grad_norm": 2.217183828353882,
+      "learning_rate": 2e-05,
+      "loss": 0.06822868,
+      "step": 3762
+    },
+    {
+      "epoch": 7.526,
+      "grad_norm": 1.5272791385650635,
+      "learning_rate": 2e-05,
+      "loss": 0.04064973,
+      "step": 3763
+    },
+    {
+      "epoch": 7.5280000000000005,
+      "grad_norm": 1.4808307886123657,
+      "learning_rate": 2e-05,
+      "loss": 0.04885202,
+      "step": 3764
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 1.6249582767486572,
+      "learning_rate": 2e-05,
+      "loss": 0.05506381,
+      "step": 3765
+    },
+    {
+      "epoch": 7.532,
+      "grad_norm": 1.6478004455566406,
+      "learning_rate": 2e-05,
+      "loss": 0.04652187,
+      "step": 3766
+    },
+    {
+      "epoch": 7.534,
+      "grad_norm": 2.208526134490967,
+      "learning_rate": 2e-05,
+      "loss": 0.049362,
+      "step": 3767
+    },
+    {
+      "epoch": 7.536,
+      "grad_norm": 1.3688409328460693,
+      "learning_rate": 2e-05,
+      "loss": 0.04084139,
+      "step": 3768
+    },
+    {
+      "epoch": 7.538,
+      "grad_norm": 1.5067732334136963,
+      "learning_rate": 2e-05,
+      "loss": 0.04833493,
+      "step": 3769
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 1.175787091255188,
+      "learning_rate": 2e-05,
+      "loss": 0.04301163,
+      "step": 3770
+    },
+    {
+      "epoch": 7.542,
+      "grad_norm": 2.0285022258758545,
+      "learning_rate": 2e-05,
+      "loss": 0.06193438,
+      "step": 3771
+    },
+    {
+      "epoch": 7.5440000000000005,
+      "grad_norm": 1.2851672172546387,
+      "learning_rate": 2e-05,
+      "loss": 0.05366874,
+      "step": 3772
+    },
+    {
+      "epoch": 7.546,
+      "grad_norm": 1.5538355112075806,
+      "learning_rate": 2e-05,
+      "loss": 0.06452001,
+      "step": 3773
+    },
+    {
+      "epoch": 7.548,
+      "grad_norm": 1.1925280094146729,
+      "learning_rate": 2e-05,
+      "loss": 0.04862928,
+      "step": 3774
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 1.3164457082748413,
+      "learning_rate": 2e-05,
+      "loss": 0.04869618,
+      "step": 3775
+    },
+    {
+      "epoch": 7.552,
+      "grad_norm": 1.8876863718032837,
+      "learning_rate": 2e-05,
+      "loss": 0.04987069,
+      "step": 3776
+    },
+    {
+      "epoch": 7.554,
+      "grad_norm": 1.348827600479126,
+      "learning_rate": 2e-05,
+      "loss": 0.05645898,
+      "step": 3777
+    },
+    {
+      "epoch": 7.556,
+      "grad_norm": 1.2659227848052979,
+      "learning_rate": 2e-05,
+      "loss": 0.04560403,
+      "step": 3778
+    },
+    {
+      "epoch": 7.558,
+      "grad_norm": 1.553835391998291,
+      "learning_rate": 2e-05,
+      "loss": 0.05577955,
+      "step": 3779
+    },
+    {
+      "epoch": 7.5600000000000005,
+      "grad_norm": 2.5734472274780273,
+      "learning_rate": 2e-05,
+      "loss": 0.05894909,
+      "step": 3780
+    },
+    {
+      "epoch": 7.562,
+      "grad_norm": 1.9130885601043701,
+      "learning_rate": 2e-05,
+      "loss": 0.04794927,
+      "step": 3781
+    },
+    {
+      "epoch": 7.564,
+      "grad_norm": 1.092248797416687,
+      "learning_rate": 2e-05,
+      "loss": 0.03762339,
+      "step": 3782
+    },
+    {
+      "epoch": 7.566,
+      "grad_norm": 1.438640832901001,
+      "learning_rate": 2e-05,
+      "loss": 0.03907945,
+      "step": 3783
+    },
+    {
+      "epoch": 7.568,
+      "grad_norm": 1.047423005104065,
+      "learning_rate": 2e-05,
+      "loss": 0.0316382,
+      "step": 3784
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 1.7063649892807007,
+      "learning_rate": 2e-05,
+      "loss": 0.04632139,
+      "step": 3785
+    },
+    {
+      "epoch": 7.572,
+      "grad_norm": 1.954330325126648,
+      "learning_rate": 2e-05,
+      "loss": 0.05917243,
+      "step": 3786
+    },
+    {
+      "epoch": 7.574,
+      "grad_norm": 1.586054801940918,
+      "learning_rate": 2e-05,
+      "loss": 0.06359352,
+      "step": 3787
+    },
+    {
+      "epoch": 7.576,
+      "grad_norm": 1.6239789724349976,
+      "learning_rate": 2e-05,
+      "loss": 0.04376218,
+      "step": 3788
+    },
+    {
+      "epoch": 7.578,
+      "grad_norm": 1.7396554946899414,
+      "learning_rate": 2e-05,
+      "loss": 0.04864378,
+      "step": 3789
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 1.1192086935043335,
+      "learning_rate": 2e-05,
+      "loss": 0.03764372,
+      "step": 3790
+    },
+    {
+      "epoch": 7.582,
+      "grad_norm": 2.047725200653076,
+      "learning_rate": 2e-05,
+      "loss": 0.05432349,
+      "step": 3791
+    },
+    {
+      "epoch": 7.584,
+      "grad_norm": 1.4690004587173462,
+      "learning_rate": 2e-05,
+      "loss": 0.05431195,
+      "step": 3792
+    },
+    {
+      "epoch": 7.586,
+      "grad_norm": 1.6998225450515747,
+      "learning_rate": 2e-05,
+      "loss": 0.05252321,
+      "step": 3793
+    },
+    {
+      "epoch": 7.588,
+      "grad_norm": 1.0173852443695068,
+      "learning_rate": 2e-05,
+      "loss": 0.03292936,
+      "step": 3794
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 1.2898446321487427,
+      "learning_rate": 2e-05,
+      "loss": 0.04756307,
+      "step": 3795
+    },
+    {
+      "epoch": 7.592,
+      "grad_norm": 1.3239659070968628,
+      "learning_rate": 2e-05,
+      "loss": 0.03519905,
+      "step": 3796
+    },
+    {
+      "epoch": 7.594,
+      "grad_norm": 0.9740890860557556,
+      "learning_rate": 2e-05,
+      "loss": 0.03270172,
+      "step": 3797
+    },
+    {
+      "epoch": 7.596,
+      "grad_norm": 1.0018112659454346,
+      "learning_rate": 2e-05,
+      "loss": 0.03196118,
+      "step": 3798
+    },
+    {
+      "epoch": 7.598,
+      "grad_norm": 1.5058925151824951,
+      "learning_rate": 2e-05,
+      "loss": 0.04523263,
+      "step": 3799
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 1.1637154817581177,
+      "learning_rate": 2e-05,
+      "loss": 0.03923118,
+      "step": 3800
+    },
+    {
+      "epoch": 7.602,
+      "grad_norm": 1.1870934963226318,
+      "learning_rate": 2e-05,
+      "loss": 0.04084285,
+      "step": 3801
+    },
+    {
+      "epoch": 7.604,
+      "grad_norm": 2.0652079582214355,
+      "learning_rate": 2e-05,
+      "loss": 0.05630117,
+      "step": 3802
+    },
+    {
+      "epoch": 7.606,
+      "grad_norm": 2.6154839992523193,
+      "learning_rate": 2e-05,
+      "loss": 0.06760383,
+      "step": 3803
+    },
+    {
+      "epoch": 7.608,
+      "grad_norm": 1.4074623584747314,
+      "learning_rate": 2e-05,
+      "loss": 0.04475513,
+      "step": 3804
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 1.3313047885894775,
+      "learning_rate": 2e-05,
+      "loss": 0.04672603,
+      "step": 3805
+    },
+    {
+      "epoch": 7.612,
+      "grad_norm": 1.1976416110992432,
+      "learning_rate": 2e-05,
+      "loss": 0.04390628,
+      "step": 3806
+    },
+    {
+      "epoch": 7.614,
+      "grad_norm": 2.146451950073242,
+      "learning_rate": 2e-05,
+      "loss": 0.05690853,
+      "step": 3807
+    },
+    {
+      "epoch": 7.616,
+      "grad_norm": 1.3076081275939941,
+      "learning_rate": 2e-05,
+      "loss": 0.05246904,
+      "step": 3808
+    },
+    {
+      "epoch": 7.618,
+      "grad_norm": 1.1059523820877075,
+      "learning_rate": 2e-05,
+      "loss": 0.02809783,
+      "step": 3809
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 1.2201027870178223,
+      "learning_rate": 2e-05,
+      "loss": 0.03628596,
+      "step": 3810
+    },
+    {
+      "epoch": 7.622,
+      "grad_norm": 1.1609594821929932,
+      "learning_rate": 2e-05,
+      "loss": 0.03147845,
+      "step": 3811
+    },
+    {
+      "epoch": 7.624,
+      "grad_norm": 1.2720181941986084,
+      "learning_rate": 2e-05,
+      "loss": 0.04290503,
+      "step": 3812
+    },
+    {
+      "epoch": 7.626,
+      "grad_norm": 1.812412142753601,
+      "learning_rate": 2e-05,
+      "loss": 0.07085785,
+      "step": 3813
+    },
+    {
+      "epoch": 7.628,
+      "grad_norm": 1.4819214344024658,
+      "learning_rate": 2e-05,
+      "loss": 0.05656476,
+      "step": 3814
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 1.9143245220184326,
+      "learning_rate": 2e-05,
+      "loss": 0.08178715,
+      "step": 3815
+    },
+    {
+      "epoch": 7.632,
+      "grad_norm": 1.3047770261764526,
+      "learning_rate": 2e-05,
+      "loss": 0.05004913,
+      "step": 3816
+    },
+    {
+      "epoch": 7.634,
+      "grad_norm": 1.793925404548645,
+      "learning_rate": 2e-05,
+      "loss": 0.04365033,
+      "step": 3817
+    },
+    {
+      "epoch": 7.636,
+      "grad_norm": 1.9158854484558105,
+      "learning_rate": 2e-05,
+      "loss": 0.05776561,
+      "step": 3818
+    },
+    {
+      "epoch": 7.638,
+      "grad_norm": 1.4316856861114502,
+      "learning_rate": 2e-05,
+      "loss": 0.04811051,
+      "step": 3819
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 1.6033645868301392,
+      "learning_rate": 2e-05,
+      "loss": 0.04948676,
+      "step": 3820
+    },
+    {
+      "epoch": 7.642,
+      "grad_norm": 1.9721709489822388,
+      "learning_rate": 2e-05,
+      "loss": 0.07113369,
+      "step": 3821
+    },
+    {
+      "epoch": 7.644,
+      "grad_norm": 2.0091919898986816,
+      "learning_rate": 2e-05,
+      "loss": 0.04521521,
+      "step": 3822
+    },
+    {
+      "epoch": 7.646,
+      "grad_norm": 2.2790145874023438,
+      "learning_rate": 2e-05,
+      "loss": 0.06836452,
+      "step": 3823
+    },
+    {
+      "epoch": 7.648,
+      "grad_norm": 1.6919300556182861,
+      "learning_rate": 2e-05,
+      "loss": 0.04315101,
+      "step": 3824
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 1.6368211507797241,
+      "learning_rate": 2e-05,
+      "loss": 0.06151176,
+      "step": 3825
+    },
+    {
+      "epoch": 7.652,
+      "grad_norm": 1.644728660583496,
+      "learning_rate": 2e-05,
+      "loss": 0.05243739,
+      "step": 3826
+    },
+    {
+      "epoch": 7.654,
+      "grad_norm": 2.1954734325408936,
+      "learning_rate": 2e-05,
+      "loss": 0.04884458,
+      "step": 3827
+    },
+    {
+      "epoch": 7.656,
+      "grad_norm": 1.4791172742843628,
+      "learning_rate": 2e-05,
+      "loss": 0.04870691,
+      "step": 3828
+    },
+    {
+      "epoch": 7.658,
+      "grad_norm": 1.6743173599243164,
+      "learning_rate": 2e-05,
+      "loss": 0.06012721,
+      "step": 3829
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 1.8005884885787964,
+      "learning_rate": 2e-05,
+      "loss": 0.05182242,
+      "step": 3830
+    },
+    {
+      "epoch": 7.662,
+      "grad_norm": 1.710728645324707,
+      "learning_rate": 2e-05,
+      "loss": 0.05671708,
+      "step": 3831
+    },
+    {
+      "epoch": 7.664,
+      "grad_norm": 1.4567813873291016,
+      "learning_rate": 2e-05,
+      "loss": 0.05406286,
+      "step": 3832
+    },
+    {
+      "epoch": 7.666,
+      "grad_norm": 1.37957763671875,
+      "learning_rate": 2e-05,
+      "loss": 0.05166141,
+      "step": 3833
+    },
+    {
+      "epoch": 7.668,
+      "grad_norm": 1.3298771381378174,
+      "learning_rate": 2e-05,
+      "loss": 0.04657075,
+      "step": 3834
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 1.678098440170288,
+      "learning_rate": 2e-05,
+      "loss": 0.06359571,
+      "step": 3835
+    },
+    {
+      "epoch": 7.672,
+      "grad_norm": 1.3091044425964355,
+      "learning_rate": 2e-05,
+      "loss": 0.04521374,
+      "step": 3836
+    },
+    {
+      "epoch": 7.674,
+      "grad_norm": 1.1188842058181763,
+      "learning_rate": 2e-05,
+      "loss": 0.05468755,
+      "step": 3837
+    },
+    {
+      "epoch": 7.676,
+      "grad_norm": 1.4298003911972046,
+      "learning_rate": 2e-05,
+      "loss": 0.05605362,
+      "step": 3838
+    },
+    {
+      "epoch": 7.678,
+      "grad_norm": 2.1561481952667236,
+      "learning_rate": 2e-05,
+      "loss": 0.05341341,
+      "step": 3839
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 1.3226423263549805,
+      "learning_rate": 2e-05,
+      "loss": 0.04852953,
+      "step": 3840
+    },
+    {
+      "epoch": 7.682,
+      "grad_norm": 1.267012596130371,
+      "learning_rate": 2e-05,
+      "loss": 0.05608751,
+      "step": 3841
+    },
+    {
+      "epoch": 7.684,
+      "grad_norm": 1.670479416847229,
+      "learning_rate": 2e-05,
+      "loss": 0.06202754,
+      "step": 3842
+    },
+    {
+      "epoch": 7.686,
+      "grad_norm": 1.5169838666915894,
+      "learning_rate": 2e-05,
+      "loss": 0.04942085,
+      "step": 3843
+    },
+    {
+      "epoch": 7.688,
+      "grad_norm": 1.5922930240631104,
+      "learning_rate": 2e-05,
+      "loss": 0.03477084,
+      "step": 3844
+    },
+    {
+      "epoch": 7.6899999999999995,
+      "grad_norm": 1.939310908317566,
+      "learning_rate": 2e-05,
+      "loss": 0.03845644,
+      "step": 3845
+    },
+    {
+      "epoch": 7.692,
+      "grad_norm": 1.3163906335830688,
+      "learning_rate": 2e-05,
+      "loss": 0.04649363,
+      "step": 3846
+    },
+    {
+      "epoch": 7.694,
+      "grad_norm": 2.3533008098602295,
+      "learning_rate": 2e-05,
+      "loss": 0.05440918,
+      "step": 3847
+    },
+    {
+      "epoch": 7.696,
+      "grad_norm": 2.0161502361297607,
+      "learning_rate": 2e-05,
+      "loss": 0.04734223,
+      "step": 3848
+    },
+    {
+      "epoch": 7.698,
+      "grad_norm": 1.932399868965149,
+      "learning_rate": 2e-05,
+      "loss": 0.04660235,
+      "step": 3849
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 1.0067873001098633,
+      "learning_rate": 2e-05,
+      "loss": 0.03094147,
+      "step": 3850
+    },
+    {
+      "epoch": 7.702,
+      "grad_norm": 1.2816112041473389,
+      "learning_rate": 2e-05,
+      "loss": 0.04377981,
+      "step": 3851
+    },
+    {
+      "epoch": 7.704,
+      "grad_norm": 1.6109353303909302,
+      "learning_rate": 2e-05,
+      "loss": 0.04601968,
+      "step": 3852
+    },
+    {
+      "epoch": 7.7059999999999995,
+      "grad_norm": 1.514350175857544,
+      "learning_rate": 2e-05,
+      "loss": 0.06082528,
+      "step": 3853
+    },
+    {
+      "epoch": 7.708,
+      "grad_norm": 1.3705843687057495,
+      "learning_rate": 2e-05,
+      "loss": 0.04245031,
+      "step": 3854
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 1.340019941329956,
+      "learning_rate": 2e-05,
+      "loss": 0.03780312,
+      "step": 3855
+    },
+    {
+      "epoch": 7.712,
+      "grad_norm": 1.9258254766464233,
+      "learning_rate": 2e-05,
+      "loss": 0.04252723,
+      "step": 3856
+    },
+    {
+      "epoch": 7.714,
+      "grad_norm": 2.0864787101745605,
+      "learning_rate": 2e-05,
+      "loss": 0.05551865,
+      "step": 3857
+    },
+    {
+      "epoch": 7.716,
+      "grad_norm": 1.5121456384658813,
+      "learning_rate": 2e-05,
+      "loss": 0.04484479,
+      "step": 3858
+    },
+    {
+      "epoch": 7.718,
+      "grad_norm": 2.2497289180755615,
+      "learning_rate": 2e-05,
+      "loss": 0.06071467,
+      "step": 3859
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 2.1290595531463623,
+      "learning_rate": 2e-05,
+      "loss": 0.05557271,
+      "step": 3860
+    },
+    {
+      "epoch": 7.7219999999999995,
+      "grad_norm": 1.1759511232376099,
+      "learning_rate": 2e-05,
+      "loss": 0.03121856,
+      "step": 3861
+    },
+    {
+      "epoch": 7.724,
+      "grad_norm": 1.8583542108535767,
+      "learning_rate": 2e-05,
+      "loss": 0.04486027,
+      "step": 3862
+    },
+    {
+      "epoch": 7.726,
+      "grad_norm": 1.753166913986206,
+      "learning_rate": 2e-05,
+      "loss": 0.05814328,
+      "step": 3863
+    },
+    {
+      "epoch": 7.728,
+      "grad_norm": 1.2085702419281006,
+      "learning_rate": 2e-05,
+      "loss": 0.04606169,
+      "step": 3864
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 0.9439897537231445,
+      "learning_rate": 2e-05,
+      "loss": 0.03185405,
+      "step": 3865
+    },
+    {
+      "epoch": 7.732,
+      "grad_norm": 1.1782246828079224,
+      "learning_rate": 2e-05,
+      "loss": 0.038643,
+      "step": 3866
+    },
+    {
+      "epoch": 7.734,
+      "grad_norm": 2.3399665355682373,
+      "learning_rate": 2e-05,
+      "loss": 0.04473191,
+      "step": 3867
+    },
+    {
+      "epoch": 7.736,
+      "grad_norm": 1.0290220975875854,
+      "learning_rate": 2e-05,
+      "loss": 0.03318851,
+      "step": 3868
+    },
+    {
+      "epoch": 7.7379999999999995,
+      "grad_norm": 2.829364061355591,
+      "learning_rate": 2e-05,
+      "loss": 0.05109379,
+      "step": 3869
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 1.3606927394866943,
+      "learning_rate": 2e-05,
+      "loss": 0.05341338,
+      "step": 3870
+    },
+    {
+      "epoch": 7.742,
+      "grad_norm": 1.5313520431518555,
+      "learning_rate": 2e-05,
+      "loss": 0.04662801,
+      "step": 3871
+    },
+    {
+      "epoch": 7.744,
+      "grad_norm": 1.2733118534088135,
+      "learning_rate": 2e-05,
+      "loss": 0.04062459,
+      "step": 3872
+    },
+    {
+      "epoch": 7.746,
+      "grad_norm": 1.3704501390457153,
+      "learning_rate": 2e-05,
+      "loss": 0.05422856,
+      "step": 3873
+    },
+    {
+      "epoch": 7.748,
+      "grad_norm": 1.1714882850646973,
+      "learning_rate": 2e-05,
+      "loss": 0.04571648,
+      "step": 3874
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 1.7015575170516968,
+      "learning_rate": 2e-05,
+      "loss": 0.05158704,
+      "step": 3875
+    },
+    {
+      "epoch": 7.752,
+      "grad_norm": 1.004583716392517,
+      "learning_rate": 2e-05,
+      "loss": 0.0300518,
+      "step": 3876
+    },
+    {
+      "epoch": 7.754,
+      "grad_norm": 1.118895173072815,
+      "learning_rate": 2e-05,
+      "loss": 0.03498417,
+      "step": 3877
+    },
+    {
+      "epoch": 7.756,
+      "grad_norm": 1.5967800617218018,
+      "learning_rate": 2e-05,
+      "loss": 0.05754827,
+      "step": 3878
+    },
+    {
+      "epoch": 7.758,
+      "grad_norm": 2.8003337383270264,
+      "learning_rate": 2e-05,
+      "loss": 0.06484474,
+      "step": 3879
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 1.5037521123886108,
+      "learning_rate": 2e-05,
+      "loss": 0.05595599,
+      "step": 3880
+    },
+    {
+      "epoch": 7.7620000000000005,
+      "grad_norm": 1.328269362449646,
+      "learning_rate": 2e-05,
+      "loss": 0.03779745,
+      "step": 3881
+    },
+    {
+      "epoch": 7.764,
+      "grad_norm": 1.3670969009399414,
+      "learning_rate": 2e-05,
+      "loss": 0.04492682,
+      "step": 3882
+    },
+    {
+      "epoch": 7.766,
+      "grad_norm": 1.441186785697937,
+      "learning_rate": 2e-05,
+      "loss": 0.04883204,
+      "step": 3883
+    },
+    {
+      "epoch": 7.768,
+      "grad_norm": 1.1069203615188599,
+      "learning_rate": 2e-05,
+      "loss": 0.03243296,
+      "step": 3884
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 1.1207300424575806,
+      "learning_rate": 2e-05,
+      "loss": 0.0498887,
+      "step": 3885
+    },
+    {
+      "epoch": 7.772,
+      "grad_norm": 1.2130826711654663,
+      "learning_rate": 2e-05,
+      "loss": 0.03474583,
+      "step": 3886
+    },
+    {
+      "epoch": 7.774,
+      "grad_norm": 1.099841833114624,
+      "learning_rate": 2e-05,
+      "loss": 0.04760575,
+      "step": 3887
+    },
+    {
+      "epoch": 7.776,
+      "grad_norm": 1.401013731956482,
+      "learning_rate": 2e-05,
+      "loss": 0.04593216,
+      "step": 3888
+    },
+    {
+      "epoch": 7.7780000000000005,
+      "grad_norm": 0.9707529544830322,
+      "learning_rate": 2e-05,
+      "loss": 0.03742143,
+      "step": 3889
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 1.4640308618545532,
+      "learning_rate": 2e-05,
+      "loss": 0.05783337,
+      "step": 3890
+    },
+    {
+      "epoch": 7.782,
+      "grad_norm": 1.4882322549819946,
+      "learning_rate": 2e-05,
+      "loss": 0.0567808,
+      "step": 3891
+    },
+    {
+      "epoch": 7.784,
+      "grad_norm": 1.206899642944336,
+      "learning_rate": 2e-05,
+      "loss": 0.0442479,
+      "step": 3892
+    },
+    {
+      "epoch": 7.786,
+      "grad_norm": 1.2967495918273926,
+      "learning_rate": 2e-05,
+      "loss": 0.04167457,
+      "step": 3893
+    },
+    {
+      "epoch": 7.788,
+      "grad_norm": 1.8896660804748535,
+      "learning_rate": 2e-05,
+      "loss": 0.04580644,
+      "step": 3894
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 2.820158004760742,
+      "learning_rate": 2e-05,
+      "loss": 0.0647872,
+      "step": 3895
+    },
+    {
+      "epoch": 7.792,
+      "grad_norm": 2.5546226501464844,
+      "learning_rate": 2e-05,
+      "loss": 0.07351266,
+      "step": 3896
+    },
+    {
+      "epoch": 7.7940000000000005,
+      "grad_norm": 2.217388868331909,
+      "learning_rate": 2e-05,
+      "loss": 0.03897737,
+      "step": 3897
+    },
+    {
+      "epoch": 7.796,
+      "grad_norm": 1.7939376831054688,
+      "learning_rate": 2e-05,
+      "loss": 0.06104586,
+      "step": 3898
+    },
+    {
+      "epoch": 7.798,
+      "grad_norm": 1.0939199924468994,
+      "learning_rate": 2e-05,
+      "loss": 0.03686985,
+      "step": 3899
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 1.504327416419983,
+      "learning_rate": 2e-05,
+      "loss": 0.05772873,
+      "step": 3900
+    },
+    {
+      "epoch": 7.802,
+      "grad_norm": 1.402275562286377,
+      "learning_rate": 2e-05,
+      "loss": 0.05212471,
+      "step": 3901
+    },
+    {
+      "epoch": 7.804,
+      "grad_norm": 1.3643393516540527,
+      "learning_rate": 2e-05,
+      "loss": 0.04227594,
+      "step": 3902
+    },
+    {
+      "epoch": 7.806,
+      "grad_norm": 1.3468226194381714,
+      "learning_rate": 2e-05,
+      "loss": 0.04262753,
+      "step": 3903
+    },
+    {
+      "epoch": 7.808,
+      "grad_norm": 1.6918001174926758,
+      "learning_rate": 2e-05,
+      "loss": 0.04145945,
+      "step": 3904
+    },
+    {
+      "epoch": 7.8100000000000005,
+      "grad_norm": 1.0832053422927856,
+      "learning_rate": 2e-05,
+      "loss": 0.03417937,
+      "step": 3905
+    },
+    {
+      "epoch": 7.812,
+      "grad_norm": 2.516186237335205,
+      "learning_rate": 2e-05,
+      "loss": 0.07497337,
+      "step": 3906
+    },
+    {
+      "epoch": 7.814,
+      "grad_norm": 3.0091376304626465,
+      "learning_rate": 2e-05,
+      "loss": 0.05103327,
+      "step": 3907
+    },
+    {
+      "epoch": 7.816,
+      "grad_norm": 1.4199695587158203,
+      "learning_rate": 2e-05,
+      "loss": 0.04106236,
+      "step": 3908
+    },
+    {
+      "epoch": 7.818,
+      "grad_norm": 2.151015281677246,
+      "learning_rate": 2e-05,
+      "loss": 0.05017766,
+      "step": 3909
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 1.086867094039917,
+      "learning_rate": 2e-05,
+      "loss": 0.04102978,
+      "step": 3910
+    },
+    {
+      "epoch": 7.822,
+      "grad_norm": 1.1980141401290894,
+      "learning_rate": 2e-05,
+      "loss": 0.04687583,
+      "step": 3911
+    },
+    {
+      "epoch": 7.824,
+      "grad_norm": 1.7949175834655762,
+      "learning_rate": 2e-05,
+      "loss": 0.04523139,
+      "step": 3912
+    },
+    {
+      "epoch": 7.826,
+      "grad_norm": 1.5633015632629395,
+      "learning_rate": 2e-05,
+      "loss": 0.0565387,
+      "step": 3913
+    },
+    {
+      "epoch": 7.828,
+      "grad_norm": 1.9414931535720825,
+      "learning_rate": 2e-05,
+      "loss": 0.03930401,
+      "step": 3914
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 1.1001695394515991,
+      "learning_rate": 2e-05,
+      "loss": 0.03971751,
+      "step": 3915
+    },
+    {
+      "epoch": 7.832,
+      "grad_norm": 1.6626832485198975,
+      "learning_rate": 2e-05,
+      "loss": 0.06389378,
+      "step": 3916
+    },
+    {
+      "epoch": 7.834,
+      "grad_norm": 1.166463017463684,
+      "learning_rate": 2e-05,
+      "loss": 0.03996446,
+      "step": 3917
+    },
+    {
+      "epoch": 7.836,
+      "grad_norm": 1.091837763786316,
+      "learning_rate": 2e-05,
+      "loss": 0.03501406,
+      "step": 3918
+    },
+    {
+      "epoch": 7.838,
+      "grad_norm": 1.5070854425430298,
+      "learning_rate": 2e-05,
+      "loss": 0.04527948,
+      "step": 3919
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 2.192519187927246,
+      "learning_rate": 2e-05,
+      "loss": 0.05735584,
+      "step": 3920
+    },
+    {
+      "epoch": 7.842,
+      "grad_norm": 3.073923349380493,
+      "learning_rate": 2e-05,
+      "loss": 0.06089948,
+      "step": 3921
+    },
+    {
+      "epoch": 7.844,
+      "grad_norm": 2.4988508224487305,
+      "learning_rate": 2e-05,
+      "loss": 0.06159104,
+      "step": 3922
+    },
+    {
+      "epoch": 7.846,
+      "grad_norm": 1.6986947059631348,
+      "learning_rate": 2e-05,
+      "loss": 0.06097404,
+      "step": 3923
+    },
+    {
+      "epoch": 7.848,
+      "grad_norm": 1.2336899042129517,
+      "learning_rate": 2e-05,
+      "loss": 0.0403936,
+      "step": 3924
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 3.7979280948638916,
+      "learning_rate": 2e-05,
+      "loss": 0.05501074,
+      "step": 3925
+    },
+    {
+      "epoch": 7.852,
+      "grad_norm": 1.620132565498352,
+      "learning_rate": 2e-05,
+      "loss": 0.05394637,
+      "step": 3926
+    },
+    {
+      "epoch": 7.854,
+      "grad_norm": 1.056098222732544,
+      "learning_rate": 2e-05,
+      "loss": 0.02777421,
+      "step": 3927
+    },
+    {
+      "epoch": 7.856,
+      "grad_norm": 1.5885734558105469,
+      "learning_rate": 2e-05,
+      "loss": 0.05239529,
+      "step": 3928
+    },
+    {
+      "epoch": 7.858,
+      "grad_norm": 1.4689480066299438,
+      "learning_rate": 2e-05,
+      "loss": 0.05046121,
+      "step": 3929
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 2.288593053817749,
+      "learning_rate": 2e-05,
+      "loss": 0.04710715,
+      "step": 3930
+    },
+    {
+      "epoch": 7.862,
+      "grad_norm": 1.4412630796432495,
+      "learning_rate": 2e-05,
+      "loss": 0.04159814,
+      "step": 3931
+    },
+    {
+      "epoch": 7.864,
+      "grad_norm": 1.4466181993484497,
+      "learning_rate": 2e-05,
+      "loss": 0.03699116,
+      "step": 3932
+    },
+    {
+      "epoch": 7.866,
+      "grad_norm": 3.294027328491211,
+      "learning_rate": 2e-05,
+      "loss": 0.05937823,
+      "step": 3933
+    },
+    {
+      "epoch": 7.868,
+      "grad_norm": 2.1902272701263428,
+      "learning_rate": 2e-05,
+      "loss": 0.06694526,
+      "step": 3934
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 2.0902066230773926,
+      "learning_rate": 2e-05,
+      "loss": 0.04954188,
+      "step": 3935
+    },
+    {
+      "epoch": 7.872,
+      "grad_norm": 1.534563422203064,
+      "learning_rate": 2e-05,
+      "loss": 0.05196944,
+      "step": 3936
+    },
+    {
+      "epoch": 7.874,
+      "grad_norm": 1.241812825202942,
+      "learning_rate": 2e-05,
+      "loss": 0.04576861,
+      "step": 3937
+    },
+    {
+      "epoch": 7.876,
+      "grad_norm": 2.344264507293701,
+      "learning_rate": 2e-05,
+      "loss": 0.04945996,
+      "step": 3938
+    },
+    {
+      "epoch": 7.878,
+      "grad_norm": 3.0257856845855713,
+      "learning_rate": 2e-05,
+      "loss": 0.05597422,
+      "step": 3939
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 1.6022652387619019,
+      "learning_rate": 2e-05,
+      "loss": 0.05853411,
+      "step": 3940
+    },
+    {
+      "epoch": 7.882,
+      "grad_norm": 1.321745753288269,
+      "learning_rate": 2e-05,
+      "loss": 0.05200193,
+      "step": 3941
+    },
+    {
+      "epoch": 7.884,
+      "grad_norm": 1.6581352949142456,
+      "learning_rate": 2e-05,
+      "loss": 0.05210207,
+      "step": 3942
+    },
+    {
+      "epoch": 7.886,
+      "grad_norm": 2.1290030479431152,
+      "learning_rate": 2e-05,
+      "loss": 0.04898274,
+      "step": 3943
+    },
+    {
+      "epoch": 7.888,
+      "grad_norm": 1.2903791666030884,
+      "learning_rate": 2e-05,
+      "loss": 0.03680386,
+      "step": 3944
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 1.1676714420318604,
+      "learning_rate": 2e-05,
+      "loss": 0.03573669,
+      "step": 3945
+    },
+    {
+      "epoch": 7.892,
+      "grad_norm": 3.1287758350372314,
+      "learning_rate": 2e-05,
+      "loss": 0.05708075,
+      "step": 3946
+    },
+    {
+      "epoch": 7.894,
+      "grad_norm": 1.6967209577560425,
+      "learning_rate": 2e-05,
+      "loss": 0.0523774,
+      "step": 3947
+    },
+    {
+      "epoch": 7.896,
+      "grad_norm": 1.3352283239364624,
+      "learning_rate": 2e-05,
+      "loss": 0.05093125,
+      "step": 3948
+    },
+    {
+      "epoch": 7.898,
+      "grad_norm": 1.4991921186447144,
+      "learning_rate": 2e-05,
+      "loss": 0.04498458,
+      "step": 3949
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 1.7871036529541016,
+      "learning_rate": 2e-05,
+      "loss": 0.05441068,
+      "step": 3950
+    },
+    {
+      "epoch": 7.902,
+      "grad_norm": 1.8313108682632446,
+      "learning_rate": 2e-05,
+      "loss": 0.04300974,
+      "step": 3951
+    },
+    {
+      "epoch": 7.904,
+      "grad_norm": 2.545856475830078,
+      "learning_rate": 2e-05,
+      "loss": 0.0370265,
+      "step": 3952
+    },
+    {
+      "epoch": 7.906,
+      "grad_norm": 1.4351707696914673,
+      "learning_rate": 2e-05,
+      "loss": 0.04847009,
+      "step": 3953
+    },
+    {
+      "epoch": 7.908,
+      "grad_norm": 1.5962963104248047,
+      "learning_rate": 2e-05,
+      "loss": 0.05077235,
+      "step": 3954
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 1.3633484840393066,
+      "learning_rate": 2e-05,
+      "loss": 0.04732682,
+      "step": 3955
+    },
+    {
+      "epoch": 7.912,
+      "grad_norm": 1.209276556968689,
+      "learning_rate": 2e-05,
+      "loss": 0.03878414,
+      "step": 3956
+    },
+    {
+      "epoch": 7.914,
+      "grad_norm": 1.957251787185669,
+      "learning_rate": 2e-05,
+      "loss": 0.07924426,
+      "step": 3957
+    },
+    {
+      "epoch": 7.916,
+      "grad_norm": 3.1658568382263184,
+      "learning_rate": 2e-05,
+      "loss": 0.05959687,
+      "step": 3958
+    },
+    {
+      "epoch": 7.918,
+      "grad_norm": 0.9807082414627075,
+      "learning_rate": 2e-05,
+      "loss": 0.03472539,
+      "step": 3959
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 2.003063917160034,
+      "learning_rate": 2e-05,
+      "loss": 0.05765334,
+      "step": 3960
+    },
+    {
+      "epoch": 7.922,
+      "grad_norm": 1.3743501901626587,
+      "learning_rate": 2e-05,
+      "loss": 0.048442,
+      "step": 3961
+    },
+    {
+      "epoch": 7.924,
+      "grad_norm": 1.9723905324935913,
+      "learning_rate": 2e-05,
+      "loss": 0.06652151,
+      "step": 3962
+    },
+    {
+      "epoch": 7.926,
+      "grad_norm": 1.4909707307815552,
+      "learning_rate": 2e-05,
+      "loss": 0.03205038,
+      "step": 3963
+    },
+    {
+      "epoch": 7.928,
+      "grad_norm": 1.1835213899612427,
+      "learning_rate": 2e-05,
+      "loss": 0.03661142,
+      "step": 3964
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 1.435006856918335,
+      "learning_rate": 2e-05,
+      "loss": 0.05666805,
+      "step": 3965
+    },
+    {
+      "epoch": 7.932,
+      "grad_norm": 2.22232723236084,
+      "learning_rate": 2e-05,
+      "loss": 0.03182552,
+      "step": 3966
+    },
+    {
+      "epoch": 7.934,
+      "grad_norm": 1.0161634683609009,
+      "learning_rate": 2e-05,
+      "loss": 0.03687178,
+      "step": 3967
+    },
+    {
+      "epoch": 7.936,
+      "grad_norm": 1.145479679107666,
+      "learning_rate": 2e-05,
+      "loss": 0.04139556,
+      "step": 3968
+    },
+    {
+      "epoch": 7.938,
+      "grad_norm": 1.3257615566253662,
+      "learning_rate": 2e-05,
+      "loss": 0.03712116,
+      "step": 3969
+    },
+    {
+      "epoch": 7.9399999999999995,
+      "grad_norm": 1.5132187604904175,
+      "learning_rate": 2e-05,
+      "loss": 0.0525584,
+      "step": 3970
+    },
+    {
+      "epoch": 7.942,
+      "grad_norm": 1.7809244394302368,
+      "learning_rate": 2e-05,
+      "loss": 0.05465652,
+      "step": 3971
+    },
+    {
+      "epoch": 7.944,
+      "grad_norm": 1.6416901350021362,
+      "learning_rate": 2e-05,
+      "loss": 0.05641666,
+      "step": 3972
+    },
+    {
+      "epoch": 7.946,
+      "grad_norm": 1.5238450765609741,
+      "learning_rate": 2e-05,
+      "loss": 0.03642607,
+      "step": 3973
+    },
+    {
+      "epoch": 7.948,
+      "grad_norm": 2.732297420501709,
+      "learning_rate": 2e-05,
+      "loss": 0.05012231,
+      "step": 3974
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 1.3529413938522339,
+      "learning_rate": 2e-05,
+      "loss": 0.03961353,
+      "step": 3975
+    },
+    {
+      "epoch": 7.952,
+      "grad_norm": 3.372781753540039,
+      "learning_rate": 2e-05,
+      "loss": 0.04853031,
+      "step": 3976
+    },
+    {
+      "epoch": 7.954,
+      "grad_norm": 2.362414836883545,
+      "learning_rate": 2e-05,
+      "loss": 0.0558218,
+      "step": 3977
+    },
+    {
+      "epoch": 7.9559999999999995,
+      "grad_norm": 1.9768744707107544,
+      "learning_rate": 2e-05,
+      "loss": 0.04110438,
+      "step": 3978
+    },
+    {
+      "epoch": 7.958,
+      "grad_norm": 1.477582335472107,
+      "learning_rate": 2e-05,
+      "loss": 0.04463062,
+      "step": 3979
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 1.62968111038208,
+      "learning_rate": 2e-05,
+      "loss": 0.04137983,
+      "step": 3980
+    },
+    {
+      "epoch": 7.962,
+      "grad_norm": 2.0086047649383545,
+      "learning_rate": 2e-05,
+      "loss": 0.05147458,
+      "step": 3981
+    },
+    {
+      "epoch": 7.964,
+      "grad_norm": 2.941561222076416,
+      "learning_rate": 2e-05,
+      "loss": 0.05373877,
+      "step": 3982
+    },
+    {
+      "epoch": 7.966,
+      "grad_norm": 1.2433993816375732,
+      "learning_rate": 2e-05,
+      "loss": 0.03581764,
+      "step": 3983
+    },
+    {
+      "epoch": 7.968,
+      "grad_norm": 1.5138049125671387,
+      "learning_rate": 2e-05,
+      "loss": 0.04050705,
+      "step": 3984
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 1.289192795753479,
+      "learning_rate": 2e-05,
+      "loss": 0.04422689,
+      "step": 3985
+    },
+    {
+      "epoch": 7.9719999999999995,
+      "grad_norm": 1.3445667028427124,
+      "learning_rate": 2e-05,
+      "loss": 0.03943577,
+      "step": 3986
+    },
+    {
+      "epoch": 7.974,
+      "grad_norm": 1.3681025505065918,
+      "learning_rate": 2e-05,
+      "loss": 0.04766712,
+      "step": 3987
+    },
+    {
+      "epoch": 7.976,
+      "grad_norm": 1.488904356956482,
+      "learning_rate": 2e-05,
+      "loss": 0.04435727,
+      "step": 3988
+    },
+    {
+      "epoch": 7.978,
+      "grad_norm": 1.3543517589569092,
+      "learning_rate": 2e-05,
+      "loss": 0.05017151,
+      "step": 3989
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 1.1544119119644165,
+      "learning_rate": 2e-05,
+      "loss": 0.04686942,
+      "step": 3990
+    },
+    {
+      "epoch": 7.982,
+      "grad_norm": 1.188092589378357,
+      "learning_rate": 2e-05,
+      "loss": 0.0426802,
+      "step": 3991
+    },
+    {
+      "epoch": 7.984,
+      "grad_norm": 1.4243437051773071,
+      "learning_rate": 2e-05,
+      "loss": 0.06280725,
+      "step": 3992
+    },
+    {
+      "epoch": 7.986,
+      "grad_norm": 1.6010160446166992,
+      "learning_rate": 2e-05,
+      "loss": 0.0427033,
+      "step": 3993
+    },
+    {
+      "epoch": 7.9879999999999995,
+      "grad_norm": 1.1920783519744873,
+      "learning_rate": 2e-05,
+      "loss": 0.04067779,
+      "step": 3994
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 1.6925147771835327,
+      "learning_rate": 2e-05,
+      "loss": 0.05137225,
+      "step": 3995
+    },
+    {
+      "epoch": 7.992,
+      "grad_norm": 1.1469167470932007,
+      "learning_rate": 2e-05,
+      "loss": 0.0329942,
+      "step": 3996
+    },
+    {
+      "epoch": 7.994,
+      "grad_norm": 1.2153685092926025,
+      "learning_rate": 2e-05,
+      "loss": 0.0425185,
+      "step": 3997
+    },
+    {
+      "epoch": 7.996,
+      "grad_norm": 1.2573164701461792,
+      "learning_rate": 2e-05,
+      "loss": 0.0418978,
+      "step": 3998
+    },
+    {
+      "epoch": 7.998,
+      "grad_norm": 1.892161250114441,
+      "learning_rate": 2e-05,
+      "loss": 0.04587626,
+      "step": 3999
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.3999619483947754,
+      "learning_rate": 2e-05,
+      "loss": 0.0356279,
+      "step": 4000
+    },
+    {
+      "epoch": 8.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.998,
+        "AngleClassification_2": 0.994,
+        "AngleClassification_3": 0.9241516966067864,
+        "Equal_1": 0.992,
+        "Equal_2": 0.9421157684630739,
+        "Equal_3": 0.8323353293413174,
+        "LineComparison_1": 0.996,
+        "LineComparison_2": 0.9960079840319361,
+        "LineComparison_3": 0.9760479041916168,
+        "Parallel_1": 0.9839679358717435,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.988,
+        "Perpendicular_1": 0.98,
+        "Perpendicular_2": 0.838,
+        "Perpendicular_3": 0.5160320641282565,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9996666666666667,
+        "PointLiesOnCircle_3": 0.9848666666666667,
+        "PointLiesOnLine_1": 0.9919839679358717,
+        "PointLiesOnLine_2": 0.9899799599198397,
+        "PointLiesOnLine_3": 0.8702594810379242
+      },
+      "eval_runtime": 320.1064,
+      "eval_samples_per_second": 32.802,
+      "eval_steps_per_second": 0.656,
+      "step": 4000
+    },
+    {
+      "epoch": 8.002,
+      "grad_norm": 1.324432611465454,
+      "learning_rate": 2e-05,
+      "loss": 0.05207051,
+      "step": 4001
+    },
+    {
+      "epoch": 8.004,
+      "grad_norm": 1.0732799768447876,
+      "learning_rate": 2e-05,
+      "loss": 0.0246641,
+      "step": 4002
+    },
+    {
+      "epoch": 8.006,
+      "grad_norm": 2.0701658725738525,
+      "learning_rate": 2e-05,
+      "loss": 0.04946818,
+      "step": 4003
+    },
+    {
+      "epoch": 8.008,
+      "grad_norm": 1.8197208642959595,
+      "learning_rate": 2e-05,
+      "loss": 0.04853348,
+      "step": 4004
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 1.2051986455917358,
+      "learning_rate": 2e-05,
+      "loss": 0.03389344,
+      "step": 4005
+    },
+    {
+      "epoch": 8.012,
+      "grad_norm": 1.289626121520996,
+      "learning_rate": 2e-05,
+      "loss": 0.04067516,
+      "step": 4006
+    },
+    {
+      "epoch": 8.014,
+      "grad_norm": 3.4996466636657715,
+      "learning_rate": 2e-05,
+      "loss": 0.07795639,
+      "step": 4007
+    },
+    {
+      "epoch": 8.016,
+      "grad_norm": 1.8532557487487793,
+      "learning_rate": 2e-05,
+      "loss": 0.04180763,
+      "step": 4008
+    },
+    {
+      "epoch": 8.018,
+      "grad_norm": 1.4671931266784668,
+      "learning_rate": 2e-05,
+      "loss": 0.04495455,
+      "step": 4009
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 1.7721261978149414,
+      "learning_rate": 2e-05,
+      "loss": 0.04306563,
+      "step": 4010
+    },
+    {
+      "epoch": 8.022,
+      "grad_norm": 1.5724709033966064,
+      "learning_rate": 2e-05,
+      "loss": 0.03561451,
+      "step": 4011
+    },
+    {
+      "epoch": 8.024,
+      "grad_norm": 1.0541595220565796,
+      "learning_rate": 2e-05,
+      "loss": 0.03057979,
+      "step": 4012
+    },
+    {
+      "epoch": 8.026,
+      "grad_norm": 1.8422211408615112,
+      "learning_rate": 2e-05,
+      "loss": 0.05177515,
+      "step": 4013
+    },
+    {
+      "epoch": 8.028,
+      "grad_norm": 1.4579343795776367,
+      "learning_rate": 2e-05,
+      "loss": 0.04849966,
+      "step": 4014
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 1.6735895872116089,
+      "learning_rate": 2e-05,
+      "loss": 0.03308116,
+      "step": 4015
+    },
+    {
+      "epoch": 8.032,
+      "grad_norm": 1.21261465549469,
+      "learning_rate": 2e-05,
+      "loss": 0.03874951,
+      "step": 4016
+    },
+    {
+      "epoch": 8.034,
+      "grad_norm": 3.218160390853882,
+      "learning_rate": 2e-05,
+      "loss": 0.03598247,
+      "step": 4017
+    },
+    {
+      "epoch": 8.036,
+      "grad_norm": 1.264754056930542,
+      "learning_rate": 2e-05,
+      "loss": 0.03381562,
+      "step": 4018
+    },
+    {
+      "epoch": 8.038,
+      "grad_norm": 2.2951626777648926,
+      "learning_rate": 2e-05,
+      "loss": 0.05828879,
+      "step": 4019
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 1.1480064392089844,
+      "learning_rate": 2e-05,
+      "loss": 0.03656226,
+      "step": 4020
+    },
+    {
+      "epoch": 8.042,
+      "grad_norm": 1.7028883695602417,
+      "learning_rate": 2e-05,
+      "loss": 0.0495864,
+      "step": 4021
+    },
+    {
+      "epoch": 8.044,
+      "grad_norm": 2.4957919120788574,
+      "learning_rate": 2e-05,
+      "loss": 0.06445777,
+      "step": 4022
+    },
+    {
+      "epoch": 8.046,
+      "grad_norm": 2.0943565368652344,
+      "learning_rate": 2e-05,
+      "loss": 0.05261732,
+      "step": 4023
+    },
+    {
+      "epoch": 8.048,
+      "grad_norm": 2.1373465061187744,
+      "learning_rate": 2e-05,
+      "loss": 0.06218139,
+      "step": 4024
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 1.4479154348373413,
+      "learning_rate": 2e-05,
+      "loss": 0.05500597,
+      "step": 4025
+    },
+    {
+      "epoch": 8.052,
+      "grad_norm": 1.107658863067627,
+      "learning_rate": 2e-05,
+      "loss": 0.04318401,
+      "step": 4026
+    },
+    {
+      "epoch": 8.054,
+      "grad_norm": 1.5696206092834473,
+      "learning_rate": 2e-05,
+      "loss": 0.04423702,
+      "step": 4027
+    },
+    {
+      "epoch": 8.056,
+      "grad_norm": 2.1380436420440674,
+      "learning_rate": 2e-05,
+      "loss": 0.04484853,
+      "step": 4028
+    },
+    {
+      "epoch": 8.058,
+      "grad_norm": 1.4376975297927856,
+      "learning_rate": 2e-05,
+      "loss": 0.03389053,
+      "step": 4029
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 1.9726535081863403,
+      "learning_rate": 2e-05,
+      "loss": 0.05158387,
+      "step": 4030
+    },
+    {
+      "epoch": 8.062,
+      "grad_norm": 1.9487265348434448,
+      "learning_rate": 2e-05,
+      "loss": 0.05221076,
+      "step": 4031
+    },
+    {
+      "epoch": 8.064,
+      "grad_norm": 1.4403632879257202,
+      "learning_rate": 2e-05,
+      "loss": 0.04184157,
+      "step": 4032
+    },
+    {
+      "epoch": 8.066,
+      "grad_norm": 1.2656970024108887,
+      "learning_rate": 2e-05,
+      "loss": 0.04085121,
+      "step": 4033
+    },
+    {
+      "epoch": 8.068,
+      "grad_norm": 1.3084996938705444,
+      "learning_rate": 2e-05,
+      "loss": 0.03328889,
+      "step": 4034
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 1.760668396949768,
+      "learning_rate": 2e-05,
+      "loss": 0.04869121,
+      "step": 4035
+    },
+    {
+      "epoch": 8.072,
+      "grad_norm": 1.6372207403182983,
+      "learning_rate": 2e-05,
+      "loss": 0.04539852,
+      "step": 4036
+    },
+    {
+      "epoch": 8.074,
+      "grad_norm": 1.2221434116363525,
+      "learning_rate": 2e-05,
+      "loss": 0.03623684,
+      "step": 4037
+    },
+    {
+      "epoch": 8.076,
+      "grad_norm": 1.080573558807373,
+      "learning_rate": 2e-05,
+      "loss": 0.03736069,
+      "step": 4038
+    },
+    {
+      "epoch": 8.078,
+      "grad_norm": 1.135930061340332,
+      "learning_rate": 2e-05,
+      "loss": 0.02960359,
+      "step": 4039
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 1.2680519819259644,
+      "learning_rate": 2e-05,
+      "loss": 0.04433823,
+      "step": 4040
+    },
+    {
+      "epoch": 8.082,
+      "grad_norm": 1.3253871202468872,
+      "learning_rate": 2e-05,
+      "loss": 0.03741252,
+      "step": 4041
+    },
+    {
+      "epoch": 8.084,
+      "grad_norm": 1.1967227458953857,
+      "learning_rate": 2e-05,
+      "loss": 0.03461279,
+      "step": 4042
+    },
+    {
+      "epoch": 8.086,
+      "grad_norm": 1.4370568990707397,
+      "learning_rate": 2e-05,
+      "loss": 0.05240814,
+      "step": 4043
+    },
+    {
+      "epoch": 8.088,
+      "grad_norm": 2.1426353454589844,
+      "learning_rate": 2e-05,
+      "loss": 0.04719769,
+      "step": 4044
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 1.9252382516860962,
+      "learning_rate": 2e-05,
+      "loss": 0.05830365,
+      "step": 4045
+    },
+    {
+      "epoch": 8.092,
+      "grad_norm": 1.4431771039962769,
+      "learning_rate": 2e-05,
+      "loss": 0.05209182,
+      "step": 4046
+    },
+    {
+      "epoch": 8.094,
+      "grad_norm": 1.4242539405822754,
+      "learning_rate": 2e-05,
+      "loss": 0.04281589,
+      "step": 4047
+    },
+    {
+      "epoch": 8.096,
+      "grad_norm": 2.547609567642212,
+      "learning_rate": 2e-05,
+      "loss": 0.04535336,
+      "step": 4048
+    },
+    {
+      "epoch": 8.098,
+      "grad_norm": 5.5258965492248535,
+      "learning_rate": 2e-05,
+      "loss": 0.06601989,
+      "step": 4049
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 1.6703463792800903,
+      "learning_rate": 2e-05,
+      "loss": 0.05124013,
+      "step": 4050
+    },
+    {
+      "epoch": 8.102,
+      "grad_norm": 1.6066875457763672,
+      "learning_rate": 2e-05,
+      "loss": 0.05647477,
+      "step": 4051
+    },
+    {
+      "epoch": 8.104,
+      "grad_norm": 1.0216000080108643,
+      "learning_rate": 2e-05,
+      "loss": 0.03139941,
+      "step": 4052
+    },
+    {
+      "epoch": 8.106,
+      "grad_norm": 1.099047303199768,
+      "learning_rate": 2e-05,
+      "loss": 0.0338769,
+      "step": 4053
+    },
+    {
+      "epoch": 8.108,
+      "grad_norm": 1.3956516981124878,
+      "learning_rate": 2e-05,
+      "loss": 0.04579592,
+      "step": 4054
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 1.361737608909607,
+      "learning_rate": 2e-05,
+      "loss": 0.04546849,
+      "step": 4055
+    },
+    {
+      "epoch": 8.112,
+      "grad_norm": 1.3461675643920898,
+      "learning_rate": 2e-05,
+      "loss": 0.05312871,
+      "step": 4056
+    },
+    {
+      "epoch": 8.114,
+      "grad_norm": 1.6900047063827515,
+      "learning_rate": 2e-05,
+      "loss": 0.03799757,
+      "step": 4057
+    },
+    {
+      "epoch": 8.116,
+      "grad_norm": 1.488451361656189,
+      "learning_rate": 2e-05,
+      "loss": 0.04809422,
+      "step": 4058
+    },
+    {
+      "epoch": 8.118,
+      "grad_norm": 1.6791332960128784,
+      "learning_rate": 2e-05,
+      "loss": 0.05416831,
+      "step": 4059
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 1.0899131298065186,
+      "learning_rate": 2e-05,
+      "loss": 0.03874314,
+      "step": 4060
+    },
+    {
+      "epoch": 8.122,
+      "grad_norm": 1.2762742042541504,
+      "learning_rate": 2e-05,
+      "loss": 0.04447027,
+      "step": 4061
+    },
+    {
+      "epoch": 8.124,
+      "grad_norm": 1.5424968004226685,
+      "learning_rate": 2e-05,
+      "loss": 0.04579141,
+      "step": 4062
+    },
+    {
+      "epoch": 8.126,
+      "grad_norm": 1.8162665367126465,
+      "learning_rate": 2e-05,
+      "loss": 0.04598666,
+      "step": 4063
+    },
+    {
+      "epoch": 8.128,
+      "grad_norm": 1.428122639656067,
+      "learning_rate": 2e-05,
+      "loss": 0.05174505,
+      "step": 4064
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 1.538621187210083,
+      "learning_rate": 2e-05,
+      "loss": 0.03250303,
+      "step": 4065
+    },
+    {
+      "epoch": 8.132,
+      "grad_norm": 1.3809401988983154,
+      "learning_rate": 2e-05,
+      "loss": 0.03839519,
+      "step": 4066
+    },
+    {
+      "epoch": 8.134,
+      "grad_norm": 1.4454106092453003,
+      "learning_rate": 2e-05,
+      "loss": 0.0449599,
+      "step": 4067
+    },
+    {
+      "epoch": 8.136,
+      "grad_norm": 1.3159528970718384,
+      "learning_rate": 2e-05,
+      "loss": 0.04408343,
+      "step": 4068
+    },
+    {
+      "epoch": 8.138,
+      "grad_norm": 1.5110152959823608,
+      "learning_rate": 2e-05,
+      "loss": 0.03739053,
+      "step": 4069
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 1.500827431678772,
+      "learning_rate": 2e-05,
+      "loss": 0.04702216,
+      "step": 4070
+    },
+    {
+      "epoch": 8.142,
+      "grad_norm": 1.5902290344238281,
+      "learning_rate": 2e-05,
+      "loss": 0.04143539,
+      "step": 4071
+    },
+    {
+      "epoch": 8.144,
+      "grad_norm": 1.9934486150741577,
+      "learning_rate": 2e-05,
+      "loss": 0.03529008,
+      "step": 4072
+    },
+    {
+      "epoch": 8.146,
+      "grad_norm": 1.8927077054977417,
+      "learning_rate": 2e-05,
+      "loss": 0.04800845,
+      "step": 4073
+    },
+    {
+      "epoch": 8.148,
+      "grad_norm": 1.6216754913330078,
+      "learning_rate": 2e-05,
+      "loss": 0.05250569,
+      "step": 4074
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 1.7913944721221924,
+      "learning_rate": 2e-05,
+      "loss": 0.04616375,
+      "step": 4075
+    },
+    {
+      "epoch": 8.152,
+      "grad_norm": 1.5655014514923096,
+      "learning_rate": 2e-05,
+      "loss": 0.05108138,
+      "step": 4076
+    },
+    {
+      "epoch": 8.154,
+      "grad_norm": 1.6830389499664307,
+      "learning_rate": 2e-05,
+      "loss": 0.05072855,
+      "step": 4077
+    },
+    {
+      "epoch": 8.156,
+      "grad_norm": 1.5033944845199585,
+      "learning_rate": 2e-05,
+      "loss": 0.04459769,
+      "step": 4078
+    },
+    {
+      "epoch": 8.158,
+      "grad_norm": 1.5763124227523804,
+      "learning_rate": 2e-05,
+      "loss": 0.05702509,
+      "step": 4079
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 1.5542982816696167,
+      "learning_rate": 2e-05,
+      "loss": 0.05888487,
+      "step": 4080
+    },
+    {
+      "epoch": 8.162,
+      "grad_norm": 1.8763879537582397,
+      "learning_rate": 2e-05,
+      "loss": 0.03832364,
+      "step": 4081
+    },
+    {
+      "epoch": 8.164,
+      "grad_norm": 2.169970989227295,
+      "learning_rate": 2e-05,
+      "loss": 0.06408481,
+      "step": 4082
+    },
+    {
+      "epoch": 8.166,
+      "grad_norm": 1.4158964157104492,
+      "learning_rate": 2e-05,
+      "loss": 0.0494574,
+      "step": 4083
+    },
+    {
+      "epoch": 8.168,
+      "grad_norm": 1.2305322885513306,
+      "learning_rate": 2e-05,
+      "loss": 0.04329508,
+      "step": 4084
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 2.0390779972076416,
+      "learning_rate": 2e-05,
+      "loss": 0.03159539,
+      "step": 4085
+    },
+    {
+      "epoch": 8.172,
+      "grad_norm": 1.616227149963379,
+      "learning_rate": 2e-05,
+      "loss": 0.08061463,
+      "step": 4086
+    },
+    {
+      "epoch": 8.174,
+      "grad_norm": 2.116055965423584,
+      "learning_rate": 2e-05,
+      "loss": 0.05509023,
+      "step": 4087
+    },
+    {
+      "epoch": 8.176,
+      "grad_norm": 1.04578697681427,
+      "learning_rate": 2e-05,
+      "loss": 0.03397291,
+      "step": 4088
+    },
+    {
+      "epoch": 8.178,
+      "grad_norm": 1.4719500541687012,
+      "learning_rate": 2e-05,
+      "loss": 0.05373418,
+      "step": 4089
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 1.9806996583938599,
+      "learning_rate": 2e-05,
+      "loss": 0.07111174,
+      "step": 4090
+    },
+    {
+      "epoch": 8.182,
+      "grad_norm": 1.30508291721344,
+      "learning_rate": 2e-05,
+      "loss": 0.04480116,
+      "step": 4091
+    },
+    {
+      "epoch": 8.184,
+      "grad_norm": 0.9421553015708923,
+      "learning_rate": 2e-05,
+      "loss": 0.03321157,
+      "step": 4092
+    },
+    {
+      "epoch": 8.186,
+      "grad_norm": 1.0021604299545288,
+      "learning_rate": 2e-05,
+      "loss": 0.04082393,
+      "step": 4093
+    },
+    {
+      "epoch": 8.188,
+      "grad_norm": 1.6946673393249512,
+      "learning_rate": 2e-05,
+      "loss": 0.05384293,
+      "step": 4094
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 1.0426790714263916,
+      "learning_rate": 2e-05,
+      "loss": 0.03759111,
+      "step": 4095
+    },
+    {
+      "epoch": 8.192,
+      "grad_norm": 2.0568313598632812,
+      "learning_rate": 2e-05,
+      "loss": 0.05517369,
+      "step": 4096
+    },
+    {
+      "epoch": 8.194,
+      "grad_norm": 1.5960932970046997,
+      "learning_rate": 2e-05,
+      "loss": 0.06000062,
+      "step": 4097
+    },
+    {
+      "epoch": 8.196,
+      "grad_norm": 1.7500308752059937,
+      "learning_rate": 2e-05,
+      "loss": 0.05646428,
+      "step": 4098
+    },
+    {
+      "epoch": 8.198,
+      "grad_norm": 1.4026926755905151,
+      "learning_rate": 2e-05,
+      "loss": 0.03786077,
+      "step": 4099
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 1.0574769973754883,
+      "learning_rate": 2e-05,
+      "loss": 0.03918152,
+      "step": 4100
+    },
+    {
+      "epoch": 8.202,
+      "grad_norm": 1.5338131189346313,
+      "learning_rate": 2e-05,
+      "loss": 0.04909813,
+      "step": 4101
+    },
+    {
+      "epoch": 8.204,
+      "grad_norm": 1.2247587442398071,
+      "learning_rate": 2e-05,
+      "loss": 0.03550359,
+      "step": 4102
+    },
+    {
+      "epoch": 8.206,
+      "grad_norm": 0.9067367315292358,
+      "learning_rate": 2e-05,
+      "loss": 0.02538646,
+      "step": 4103
+    },
+    {
+      "epoch": 8.208,
+      "grad_norm": 1.0645413398742676,
+      "learning_rate": 2e-05,
+      "loss": 0.02908769,
+      "step": 4104
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 1.0075726509094238,
+      "learning_rate": 2e-05,
+      "loss": 0.03003458,
+      "step": 4105
+    },
+    {
+      "epoch": 8.212,
+      "grad_norm": 1.2275352478027344,
+      "learning_rate": 2e-05,
+      "loss": 0.051549,
+      "step": 4106
+    },
+    {
+      "epoch": 8.214,
+      "grad_norm": 1.427625298500061,
+      "learning_rate": 2e-05,
+      "loss": 0.04538945,
+      "step": 4107
+    },
+    {
+      "epoch": 8.216,
+      "grad_norm": 1.29180908203125,
+      "learning_rate": 2e-05,
+      "loss": 0.04527157,
+      "step": 4108
+    },
+    {
+      "epoch": 8.218,
+      "grad_norm": 1.678622841835022,
+      "learning_rate": 2e-05,
+      "loss": 0.03252411,
+      "step": 4109
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 2.289098024368286,
+      "learning_rate": 2e-05,
+      "loss": 0.0732346,
+      "step": 4110
+    },
+    {
+      "epoch": 8.222,
+      "grad_norm": 1.325034499168396,
+      "learning_rate": 2e-05,
+      "loss": 0.05234051,
+      "step": 4111
+    },
+    {
+      "epoch": 8.224,
+      "grad_norm": 1.978644847869873,
+      "learning_rate": 2e-05,
+      "loss": 0.05532134,
+      "step": 4112
+    },
+    {
+      "epoch": 8.226,
+      "grad_norm": 1.7182114124298096,
+      "learning_rate": 2e-05,
+      "loss": 0.0578098,
+      "step": 4113
+    },
+    {
+      "epoch": 8.228,
+      "grad_norm": 1.4846113920211792,
+      "learning_rate": 2e-05,
+      "loss": 0.04360331,
+      "step": 4114
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 1.358889102935791,
+      "learning_rate": 2e-05,
+      "loss": 0.05047805,
+      "step": 4115
+    },
+    {
+      "epoch": 8.232,
+      "grad_norm": 1.1359084844589233,
+      "learning_rate": 2e-05,
+      "loss": 0.04568867,
+      "step": 4116
+    },
+    {
+      "epoch": 8.234,
+      "grad_norm": 1.419180154800415,
+      "learning_rate": 2e-05,
+      "loss": 0.03961842,
+      "step": 4117
+    },
+    {
+      "epoch": 8.236,
+      "grad_norm": 1.3813687562942505,
+      "learning_rate": 2e-05,
+      "loss": 0.05298204,
+      "step": 4118
+    },
+    {
+      "epoch": 8.238,
+      "grad_norm": 1.8170478343963623,
+      "learning_rate": 2e-05,
+      "loss": 0.04838009,
+      "step": 4119
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 2.8637313842773438,
+      "learning_rate": 2e-05,
+      "loss": 0.06887215,
+      "step": 4120
+    },
+    {
+      "epoch": 8.242,
+      "grad_norm": 1.631001591682434,
+      "learning_rate": 2e-05,
+      "loss": 0.04441953,
+      "step": 4121
+    },
+    {
+      "epoch": 8.244,
+      "grad_norm": 1.4597910642623901,
+      "learning_rate": 2e-05,
+      "loss": 0.04484698,
+      "step": 4122
+    },
+    {
+      "epoch": 8.246,
+      "grad_norm": 0.9799979329109192,
+      "learning_rate": 2e-05,
+      "loss": 0.02647844,
+      "step": 4123
+    },
+    {
+      "epoch": 8.248,
+      "grad_norm": 1.7152873277664185,
+      "learning_rate": 2e-05,
+      "loss": 0.06567133,
+      "step": 4124
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 1.425673007965088,
+      "learning_rate": 2e-05,
+      "loss": 0.04011939,
+      "step": 4125
+    },
+    {
+      "epoch": 8.252,
+      "grad_norm": 2.331421375274658,
+      "learning_rate": 2e-05,
+      "loss": 0.06116243,
+      "step": 4126
+    },
+    {
+      "epoch": 8.254,
+      "grad_norm": 3.033942461013794,
+      "learning_rate": 2e-05,
+      "loss": 0.04823023,
+      "step": 4127
+    },
+    {
+      "epoch": 8.256,
+      "grad_norm": 1.2400274276733398,
+      "learning_rate": 2e-05,
+      "loss": 0.03877744,
+      "step": 4128
+    },
+    {
+      "epoch": 8.258,
+      "grad_norm": 1.145862340927124,
+      "learning_rate": 2e-05,
+      "loss": 0.04107977,
+      "step": 4129
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 1.4702391624450684,
+      "learning_rate": 2e-05,
+      "loss": 0.06288863,
+      "step": 4130
+    },
+    {
+      "epoch": 8.262,
+      "grad_norm": 1.4819300174713135,
+      "learning_rate": 2e-05,
+      "loss": 0.05648782,
+      "step": 4131
+    },
+    {
+      "epoch": 8.264,
+      "grad_norm": 1.4082889556884766,
+      "learning_rate": 2e-05,
+      "loss": 0.05002226,
+      "step": 4132
+    },
+    {
+      "epoch": 8.266,
+      "grad_norm": 1.2530863285064697,
+      "learning_rate": 2e-05,
+      "loss": 0.04050343,
+      "step": 4133
+    },
+    {
+      "epoch": 8.268,
+      "grad_norm": 1.1162209510803223,
+      "learning_rate": 2e-05,
+      "loss": 0.03798089,
+      "step": 4134
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 1.6518158912658691,
+      "learning_rate": 2e-05,
+      "loss": 0.04391007,
+      "step": 4135
+    },
+    {
+      "epoch": 8.272,
+      "grad_norm": 1.1009466648101807,
+      "learning_rate": 2e-05,
+      "loss": 0.04458437,
+      "step": 4136
+    },
+    {
+      "epoch": 8.274000000000001,
+      "grad_norm": 1.3596596717834473,
+      "learning_rate": 2e-05,
+      "loss": 0.0591647,
+      "step": 4137
+    },
+    {
+      "epoch": 8.276,
+      "grad_norm": 1.1786128282546997,
+      "learning_rate": 2e-05,
+      "loss": 0.03626633,
+      "step": 4138
+    },
+    {
+      "epoch": 8.278,
+      "grad_norm": 1.3034263849258423,
+      "learning_rate": 2e-05,
+      "loss": 0.03502483,
+      "step": 4139
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 1.229113221168518,
+      "learning_rate": 2e-05,
+      "loss": 0.05113623,
+      "step": 4140
+    },
+    {
+      "epoch": 8.282,
+      "grad_norm": 1.813858985900879,
+      "learning_rate": 2e-05,
+      "loss": 0.04342901,
+      "step": 4141
+    },
+    {
+      "epoch": 8.284,
+      "grad_norm": 1.5317081212997437,
+      "learning_rate": 2e-05,
+      "loss": 0.05725561,
+      "step": 4142
+    },
+    {
+      "epoch": 8.286,
+      "grad_norm": 1.4534281492233276,
+      "learning_rate": 2e-05,
+      "loss": 0.04506959,
+      "step": 4143
+    },
+    {
+      "epoch": 8.288,
+      "grad_norm": 1.4005870819091797,
+      "learning_rate": 2e-05,
+      "loss": 0.05307934,
+      "step": 4144
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 1.480741262435913,
+      "learning_rate": 2e-05,
+      "loss": 0.0603156,
+      "step": 4145
+    },
+    {
+      "epoch": 8.292,
+      "grad_norm": 1.7137929201126099,
+      "learning_rate": 2e-05,
+      "loss": 0.06312002,
+      "step": 4146
+    },
+    {
+      "epoch": 8.294,
+      "grad_norm": 1.4770913124084473,
+      "learning_rate": 2e-05,
+      "loss": 0.05373389,
+      "step": 4147
+    },
+    {
+      "epoch": 8.296,
+      "grad_norm": 1.1751221418380737,
+      "learning_rate": 2e-05,
+      "loss": 0.03505269,
+      "step": 4148
+    },
+    {
+      "epoch": 8.298,
+      "grad_norm": 1.420811653137207,
+      "learning_rate": 2e-05,
+      "loss": 0.03696169,
+      "step": 4149
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 1.4480655193328857,
+      "learning_rate": 2e-05,
+      "loss": 0.04662473,
+      "step": 4150
+    },
+    {
+      "epoch": 8.302,
+      "grad_norm": 2.5621485710144043,
+      "learning_rate": 2e-05,
+      "loss": 0.04032537,
+      "step": 4151
+    },
+    {
+      "epoch": 8.304,
+      "grad_norm": 1.8456149101257324,
+      "learning_rate": 2e-05,
+      "loss": 0.05323803,
+      "step": 4152
+    },
+    {
+      "epoch": 8.306,
+      "grad_norm": 1.150632619857788,
+      "learning_rate": 2e-05,
+      "loss": 0.03429832,
+      "step": 4153
+    },
+    {
+      "epoch": 8.308,
+      "grad_norm": 1.3768160343170166,
+      "learning_rate": 2e-05,
+      "loss": 0.04700703,
+      "step": 4154
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 3.1976876258850098,
+      "learning_rate": 2e-05,
+      "loss": 0.0506537,
+      "step": 4155
+    },
+    {
+      "epoch": 8.312,
+      "grad_norm": 1.144446611404419,
+      "learning_rate": 2e-05,
+      "loss": 0.04074134,
+      "step": 4156
+    },
+    {
+      "epoch": 8.314,
+      "grad_norm": 1.0326695442199707,
+      "learning_rate": 2e-05,
+      "loss": 0.03758462,
+      "step": 4157
+    },
+    {
+      "epoch": 8.316,
+      "grad_norm": 1.5255597829818726,
+      "learning_rate": 2e-05,
+      "loss": 0.05389597,
+      "step": 4158
+    },
+    {
+      "epoch": 8.318,
+      "grad_norm": 1.6066386699676514,
+      "learning_rate": 2e-05,
+      "loss": 0.04561181,
+      "step": 4159
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 1.2478289604187012,
+      "learning_rate": 2e-05,
+      "loss": 0.05142663,
+      "step": 4160
+    },
+    {
+      "epoch": 8.322,
+      "grad_norm": 1.8503996133804321,
+      "learning_rate": 2e-05,
+      "loss": 0.04205081,
+      "step": 4161
+    },
+    {
+      "epoch": 8.324,
+      "grad_norm": 2.421180486679077,
+      "learning_rate": 2e-05,
+      "loss": 0.03882204,
+      "step": 4162
+    },
+    {
+      "epoch": 8.326,
+      "grad_norm": 1.386807918548584,
+      "learning_rate": 2e-05,
+      "loss": 0.05854031,
+      "step": 4163
+    },
+    {
+      "epoch": 8.328,
+      "grad_norm": 2.2929842472076416,
+      "learning_rate": 2e-05,
+      "loss": 0.04718957,
+      "step": 4164
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 0.9168434739112854,
+      "learning_rate": 2e-05,
+      "loss": 0.02453806,
+      "step": 4165
+    },
+    {
+      "epoch": 8.332,
+      "grad_norm": 6.381651401519775,
+      "learning_rate": 2e-05,
+      "loss": 0.06762838,
+      "step": 4166
+    },
+    {
+      "epoch": 8.334,
+      "grad_norm": 1.249552607536316,
+      "learning_rate": 2e-05,
+      "loss": 0.0523265,
+      "step": 4167
+    },
+    {
+      "epoch": 8.336,
+      "grad_norm": 1.3933881521224976,
+      "learning_rate": 2e-05,
+      "loss": 0.04636804,
+      "step": 4168
+    },
+    {
+      "epoch": 8.338,
+      "grad_norm": 1.6516380310058594,
+      "learning_rate": 2e-05,
+      "loss": 0.04308794,
+      "step": 4169
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 1.5074326992034912,
+      "learning_rate": 2e-05,
+      "loss": 0.03881112,
+      "step": 4170
+    },
+    {
+      "epoch": 8.342,
+      "grad_norm": 1.099740743637085,
+      "learning_rate": 2e-05,
+      "loss": 0.04978018,
+      "step": 4171
+    },
+    {
+      "epoch": 8.344,
+      "grad_norm": 1.2606548070907593,
+      "learning_rate": 2e-05,
+      "loss": 0.04098994,
+      "step": 4172
+    },
+    {
+      "epoch": 8.346,
+      "grad_norm": 1.293941617012024,
+      "learning_rate": 2e-05,
+      "loss": 0.04166168,
+      "step": 4173
+    },
+    {
+      "epoch": 8.348,
+      "grad_norm": 1.2298349142074585,
+      "learning_rate": 2e-05,
+      "loss": 0.04704496,
+      "step": 4174
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 1.1148796081542969,
+      "learning_rate": 2e-05,
+      "loss": 0.04025009,
+      "step": 4175
+    },
+    {
+      "epoch": 8.352,
+      "grad_norm": 1.6456156969070435,
+      "learning_rate": 2e-05,
+      "loss": 0.04465715,
+      "step": 4176
+    },
+    {
+      "epoch": 8.354,
+      "grad_norm": 1.3357329368591309,
+      "learning_rate": 2e-05,
+      "loss": 0.04234711,
+      "step": 4177
+    },
+    {
+      "epoch": 8.356,
+      "grad_norm": 1.0840578079223633,
+      "learning_rate": 2e-05,
+      "loss": 0.0413487,
+      "step": 4178
+    },
+    {
+      "epoch": 8.358,
+      "grad_norm": 1.1986706256866455,
+      "learning_rate": 2e-05,
+      "loss": 0.04364356,
+      "step": 4179
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 1.2560433149337769,
+      "learning_rate": 2e-05,
+      "loss": 0.04892932,
+      "step": 4180
+    },
+    {
+      "epoch": 8.362,
+      "grad_norm": 1.2587891817092896,
+      "learning_rate": 2e-05,
+      "loss": 0.04414427,
+      "step": 4181
+    },
+    {
+      "epoch": 8.364,
+      "grad_norm": 1.3355039358139038,
+      "learning_rate": 2e-05,
+      "loss": 0.05022869,
+      "step": 4182
+    },
+    {
+      "epoch": 8.366,
+      "grad_norm": 1.277706503868103,
+      "learning_rate": 2e-05,
+      "loss": 0.04096631,
+      "step": 4183
+    },
+    {
+      "epoch": 8.368,
+      "grad_norm": 1.781320333480835,
+      "learning_rate": 2e-05,
+      "loss": 0.0590826,
+      "step": 4184
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 1.9013862609863281,
+      "learning_rate": 2e-05,
+      "loss": 0.04787229,
+      "step": 4185
+    },
+    {
+      "epoch": 8.372,
+      "grad_norm": 1.6217079162597656,
+      "learning_rate": 2e-05,
+      "loss": 0.05203458,
+      "step": 4186
+    },
+    {
+      "epoch": 8.374,
+      "grad_norm": 1.394450068473816,
+      "learning_rate": 2e-05,
+      "loss": 0.05752437,
+      "step": 4187
+    },
+    {
+      "epoch": 8.376,
+      "grad_norm": 0.8604865670204163,
+      "learning_rate": 2e-05,
+      "loss": 0.02740677,
+      "step": 4188
+    },
+    {
+      "epoch": 8.378,
+      "grad_norm": 1.4956645965576172,
+      "learning_rate": 2e-05,
+      "loss": 0.0441338,
+      "step": 4189
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 1.9489855766296387,
+      "learning_rate": 2e-05,
+      "loss": 0.07236366,
+      "step": 4190
+    },
+    {
+      "epoch": 8.382,
+      "grad_norm": 1.3766798973083496,
+      "learning_rate": 2e-05,
+      "loss": 0.04797775,
+      "step": 4191
+    },
+    {
+      "epoch": 8.384,
+      "grad_norm": 1.082892894744873,
+      "learning_rate": 2e-05,
+      "loss": 0.04361612,
+      "step": 4192
+    },
+    {
+      "epoch": 8.386,
+      "grad_norm": 1.0579158067703247,
+      "learning_rate": 2e-05,
+      "loss": 0.04247576,
+      "step": 4193
+    },
+    {
+      "epoch": 8.388,
+      "grad_norm": 2.016787052154541,
+      "learning_rate": 2e-05,
+      "loss": 0.05003262,
+      "step": 4194
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 1.2439075708389282,
+      "learning_rate": 2e-05,
+      "loss": 0.04027151,
+      "step": 4195
+    },
+    {
+      "epoch": 8.392,
+      "grad_norm": 1.218614935874939,
+      "learning_rate": 2e-05,
+      "loss": 0.04031168,
+      "step": 4196
+    },
+    {
+      "epoch": 8.394,
+      "grad_norm": 1.2276456356048584,
+      "learning_rate": 2e-05,
+      "loss": 0.04959612,
+      "step": 4197
+    },
+    {
+      "epoch": 8.396,
+      "grad_norm": 1.965421199798584,
+      "learning_rate": 2e-05,
+      "loss": 0.04630712,
+      "step": 4198
+    },
+    {
+      "epoch": 8.398,
+      "grad_norm": 1.103851318359375,
+      "learning_rate": 2e-05,
+      "loss": 0.03305928,
+      "step": 4199
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 1.183467149734497,
+      "learning_rate": 2e-05,
+      "loss": 0.04819242,
+      "step": 4200
+    },
+    {
+      "epoch": 8.402,
+      "grad_norm": 1.2708202600479126,
+      "learning_rate": 2e-05,
+      "loss": 0.04293144,
+      "step": 4201
+    },
+    {
+      "epoch": 8.404,
+      "grad_norm": 1.534544587135315,
+      "learning_rate": 2e-05,
+      "loss": 0.04298072,
+      "step": 4202
+    },
+    {
+      "epoch": 8.406,
+      "grad_norm": 1.9671140909194946,
+      "learning_rate": 2e-05,
+      "loss": 0.05531413,
+      "step": 4203
+    },
+    {
+      "epoch": 8.408,
+      "grad_norm": 1.3386366367340088,
+      "learning_rate": 2e-05,
+      "loss": 0.03748547,
+      "step": 4204
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 2.237844467163086,
+      "learning_rate": 2e-05,
+      "loss": 0.05199295,
+      "step": 4205
+    },
+    {
+      "epoch": 8.412,
+      "grad_norm": 1.5903010368347168,
+      "learning_rate": 2e-05,
+      "loss": 0.04232503,
+      "step": 4206
+    },
+    {
+      "epoch": 8.414,
+      "grad_norm": 1.1368825435638428,
+      "learning_rate": 2e-05,
+      "loss": 0.03277536,
+      "step": 4207
+    },
+    {
+      "epoch": 8.416,
+      "grad_norm": 1.1275688409805298,
+      "learning_rate": 2e-05,
+      "loss": 0.04298157,
+      "step": 4208
+    },
+    {
+      "epoch": 8.418,
+      "grad_norm": 1.7428966760635376,
+      "learning_rate": 2e-05,
+      "loss": 0.05383879,
+      "step": 4209
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 1.735756516456604,
+      "learning_rate": 2e-05,
+      "loss": 0.06661352,
+      "step": 4210
+    },
+    {
+      "epoch": 8.422,
+      "grad_norm": 1.376869797706604,
+      "learning_rate": 2e-05,
+      "loss": 0.04623829,
+      "step": 4211
+    },
+    {
+      "epoch": 8.424,
+      "grad_norm": 2.886859178543091,
+      "learning_rate": 2e-05,
+      "loss": 0.06156854,
+      "step": 4212
+    },
+    {
+      "epoch": 8.426,
+      "grad_norm": 1.3723516464233398,
+      "learning_rate": 2e-05,
+      "loss": 0.04914469,
+      "step": 4213
+    },
+    {
+      "epoch": 8.428,
+      "grad_norm": 1.2038379907608032,
+      "learning_rate": 2e-05,
+      "loss": 0.0332229,
+      "step": 4214
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 1.1281368732452393,
+      "learning_rate": 2e-05,
+      "loss": 0.03839248,
+      "step": 4215
+    },
+    {
+      "epoch": 8.432,
+      "grad_norm": 1.2697054147720337,
+      "learning_rate": 2e-05,
+      "loss": 0.03925816,
+      "step": 4216
+    },
+    {
+      "epoch": 8.434,
+      "grad_norm": 1.1074013710021973,
+      "learning_rate": 2e-05,
+      "loss": 0.04592735,
+      "step": 4217
+    },
+    {
+      "epoch": 8.436,
+      "grad_norm": 1.722346305847168,
+      "learning_rate": 2e-05,
+      "loss": 0.05641073,
+      "step": 4218
+    },
+    {
+      "epoch": 8.438,
+      "grad_norm": 1.3256508111953735,
+      "learning_rate": 2e-05,
+      "loss": 0.04843621,
+      "step": 4219
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 1.3698394298553467,
+      "learning_rate": 2e-05,
+      "loss": 0.04274657,
+      "step": 4220
+    },
+    {
+      "epoch": 8.442,
+      "grad_norm": 1.7315605878829956,
+      "learning_rate": 2e-05,
+      "loss": 0.05456549,
+      "step": 4221
+    },
+    {
+      "epoch": 8.444,
+      "grad_norm": 1.2529323101043701,
+      "learning_rate": 2e-05,
+      "loss": 0.02203387,
+      "step": 4222
+    },
+    {
+      "epoch": 8.446,
+      "grad_norm": 2.7959108352661133,
+      "learning_rate": 2e-05,
+      "loss": 0.05525438,
+      "step": 4223
+    },
+    {
+      "epoch": 8.448,
+      "grad_norm": 1.3126109838485718,
+      "learning_rate": 2e-05,
+      "loss": 0.0481948,
+      "step": 4224
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 1.4437806606292725,
+      "learning_rate": 2e-05,
+      "loss": 0.05020737,
+      "step": 4225
+    },
+    {
+      "epoch": 8.452,
+      "grad_norm": 1.2649277448654175,
+      "learning_rate": 2e-05,
+      "loss": 0.04435945,
+      "step": 4226
+    },
+    {
+      "epoch": 8.454,
+      "grad_norm": 1.9862298965454102,
+      "learning_rate": 2e-05,
+      "loss": 0.04354966,
+      "step": 4227
+    },
+    {
+      "epoch": 8.456,
+      "grad_norm": 1.2959977388381958,
+      "learning_rate": 2e-05,
+      "loss": 0.04303737,
+      "step": 4228
+    },
+    {
+      "epoch": 8.458,
+      "grad_norm": 1.9561591148376465,
+      "learning_rate": 2e-05,
+      "loss": 0.04295287,
+      "step": 4229
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 1.8894590139389038,
+      "learning_rate": 2e-05,
+      "loss": 0.04403112,
+      "step": 4230
+    },
+    {
+      "epoch": 8.462,
+      "grad_norm": 2.1889703273773193,
+      "learning_rate": 2e-05,
+      "loss": 0.06404384,
+      "step": 4231
+    },
+    {
+      "epoch": 8.464,
+      "grad_norm": 1.6557297706604004,
+      "learning_rate": 2e-05,
+      "loss": 0.05207235,
+      "step": 4232
+    },
+    {
+      "epoch": 8.466,
+      "grad_norm": 1.6295475959777832,
+      "learning_rate": 2e-05,
+      "loss": 0.04500223,
+      "step": 4233
+    },
+    {
+      "epoch": 8.468,
+      "grad_norm": 1.411564588546753,
+      "learning_rate": 2e-05,
+      "loss": 0.04487611,
+      "step": 4234
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 1.886080026626587,
+      "learning_rate": 2e-05,
+      "loss": 0.03916345,
+      "step": 4235
+    },
+    {
+      "epoch": 8.472,
+      "grad_norm": 1.2690297365188599,
+      "learning_rate": 2e-05,
+      "loss": 0.03984193,
+      "step": 4236
+    },
+    {
+      "epoch": 8.474,
+      "grad_norm": 1.3404802083969116,
+      "learning_rate": 2e-05,
+      "loss": 0.05036929,
+      "step": 4237
+    },
+    {
+      "epoch": 8.475999999999999,
+      "grad_norm": 1.5249813795089722,
+      "learning_rate": 2e-05,
+      "loss": 0.04390936,
+      "step": 4238
+    },
+    {
+      "epoch": 8.478,
+      "grad_norm": 1.550538182258606,
+      "learning_rate": 2e-05,
+      "loss": 0.06574559,
+      "step": 4239
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 1.3451995849609375,
+      "learning_rate": 2e-05,
+      "loss": 0.04779359,
+      "step": 4240
+    },
+    {
+      "epoch": 8.482,
+      "grad_norm": 1.6042060852050781,
+      "learning_rate": 2e-05,
+      "loss": 0.04354705,
+      "step": 4241
+    },
+    {
+      "epoch": 8.484,
+      "grad_norm": 1.3204594850540161,
+      "learning_rate": 2e-05,
+      "loss": 0.03058997,
+      "step": 4242
+    },
+    {
+      "epoch": 8.486,
+      "grad_norm": 1.1289502382278442,
+      "learning_rate": 2e-05,
+      "loss": 0.04202674,
+      "step": 4243
+    },
+    {
+      "epoch": 8.488,
+      "grad_norm": 1.6240679025650024,
+      "learning_rate": 2e-05,
+      "loss": 0.05241815,
+      "step": 4244
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 1.0163280963897705,
+      "learning_rate": 2e-05,
+      "loss": 0.03246521,
+      "step": 4245
+    },
+    {
+      "epoch": 8.492,
+      "grad_norm": 1.3230700492858887,
+      "learning_rate": 2e-05,
+      "loss": 0.04476965,
+      "step": 4246
+    },
+    {
+      "epoch": 8.494,
+      "grad_norm": 1.686160683631897,
+      "learning_rate": 2e-05,
+      "loss": 0.05470574,
+      "step": 4247
+    },
+    {
+      "epoch": 8.496,
+      "grad_norm": 1.419310450553894,
+      "learning_rate": 2e-05,
+      "loss": 0.04859785,
+      "step": 4248
+    },
+    {
+      "epoch": 8.498,
+      "grad_norm": 1.431846261024475,
+      "learning_rate": 2e-05,
+      "loss": 0.04167376,
+      "step": 4249
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 1.4465125799179077,
+      "learning_rate": 2e-05,
+      "loss": 0.06550825,
+      "step": 4250
+    },
+    {
+      "epoch": 8.502,
+      "grad_norm": 2.1822566986083984,
+      "learning_rate": 2e-05,
+      "loss": 0.05803905,
+      "step": 4251
+    },
+    {
+      "epoch": 8.504,
+      "grad_norm": 1.2648488283157349,
+      "learning_rate": 2e-05,
+      "loss": 0.05157655,
+      "step": 4252
+    },
+    {
+      "epoch": 8.506,
+      "grad_norm": 1.1298472881317139,
+      "learning_rate": 2e-05,
+      "loss": 0.04487451,
+      "step": 4253
+    },
+    {
+      "epoch": 8.508,
+      "grad_norm": 1.696395754814148,
+      "learning_rate": 2e-05,
+      "loss": 0.04066026,
+      "step": 4254
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 1.068587064743042,
+      "learning_rate": 2e-05,
+      "loss": 0.02926126,
+      "step": 4255
+    },
+    {
+      "epoch": 8.512,
+      "grad_norm": 1.2515802383422852,
+      "learning_rate": 2e-05,
+      "loss": 0.04963365,
+      "step": 4256
+    },
+    {
+      "epoch": 8.514,
+      "grad_norm": 1.1776715517044067,
+      "learning_rate": 2e-05,
+      "loss": 0.03471005,
+      "step": 4257
+    },
+    {
+      "epoch": 8.516,
+      "grad_norm": 1.47025728225708,
+      "learning_rate": 2e-05,
+      "loss": 0.04580873,
+      "step": 4258
+    },
+    {
+      "epoch": 8.518,
+      "grad_norm": 1.3387584686279297,
+      "learning_rate": 2e-05,
+      "loss": 0.0392997,
+      "step": 4259
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 1.7431012392044067,
+      "learning_rate": 2e-05,
+      "loss": 0.04956745,
+      "step": 4260
+    },
+    {
+      "epoch": 8.522,
+      "grad_norm": 2.1029186248779297,
+      "learning_rate": 2e-05,
+      "loss": 0.0650689,
+      "step": 4261
+    },
+    {
+      "epoch": 8.524000000000001,
+      "grad_norm": 1.218745470046997,
+      "learning_rate": 2e-05,
+      "loss": 0.03940692,
+      "step": 4262
+    },
+    {
+      "epoch": 8.526,
+      "grad_norm": 1.8055042028427124,
+      "learning_rate": 2e-05,
+      "loss": 0.04726121,
+      "step": 4263
+    },
+    {
+      "epoch": 8.528,
+      "grad_norm": 1.2838687896728516,
+      "learning_rate": 2e-05,
+      "loss": 0.04223323,
+      "step": 4264
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 1.1817346811294556,
+      "learning_rate": 2e-05,
+      "loss": 0.03910241,
+      "step": 4265
+    },
+    {
+      "epoch": 8.532,
+      "grad_norm": 2.022214651107788,
+      "learning_rate": 2e-05,
+      "loss": 0.05597316,
+      "step": 4266
+    },
+    {
+      "epoch": 8.534,
+      "grad_norm": 2.186800479888916,
+      "learning_rate": 2e-05,
+      "loss": 0.03829915,
+      "step": 4267
+    },
+    {
+      "epoch": 8.536,
+      "grad_norm": 1.6917003393173218,
+      "learning_rate": 2e-05,
+      "loss": 0.05600104,
+      "step": 4268
+    },
+    {
+      "epoch": 8.538,
+      "grad_norm": 1.4837727546691895,
+      "learning_rate": 2e-05,
+      "loss": 0.04096905,
+      "step": 4269
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 1.8344814777374268,
+      "learning_rate": 2e-05,
+      "loss": 0.0464284,
+      "step": 4270
+    },
+    {
+      "epoch": 8.542,
+      "grad_norm": 1.7132585048675537,
+      "learning_rate": 2e-05,
+      "loss": 0.04327628,
+      "step": 4271
+    },
+    {
+      "epoch": 8.544,
+      "grad_norm": 1.674782395362854,
+      "learning_rate": 2e-05,
+      "loss": 0.04533742,
+      "step": 4272
+    },
+    {
+      "epoch": 8.546,
+      "grad_norm": 1.6211055517196655,
+      "learning_rate": 2e-05,
+      "loss": 0.04555016,
+      "step": 4273
+    },
+    {
+      "epoch": 8.548,
+      "grad_norm": 2.098546266555786,
+      "learning_rate": 2e-05,
+      "loss": 0.06424834,
+      "step": 4274
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 1.812697172164917,
+      "learning_rate": 2e-05,
+      "loss": 0.05120952,
+      "step": 4275
+    },
+    {
+      "epoch": 8.552,
+      "grad_norm": 1.0520689487457275,
+      "learning_rate": 2e-05,
+      "loss": 0.03506008,
+      "step": 4276
+    },
+    {
+      "epoch": 8.554,
+      "grad_norm": 1.5413808822631836,
+      "learning_rate": 2e-05,
+      "loss": 0.03506286,
+      "step": 4277
+    },
+    {
+      "epoch": 8.556000000000001,
+      "grad_norm": 0.9895492196083069,
+      "learning_rate": 2e-05,
+      "loss": 0.03194591,
+      "step": 4278
+    },
+    {
+      "epoch": 8.558,
+      "grad_norm": 2.4871461391448975,
+      "learning_rate": 2e-05,
+      "loss": 0.04312088,
+      "step": 4279
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 2.135133743286133,
+      "learning_rate": 2e-05,
+      "loss": 0.05813865,
+      "step": 4280
+    },
+    {
+      "epoch": 8.562,
+      "grad_norm": 1.7616969347000122,
+      "learning_rate": 2e-05,
+      "loss": 0.05905205,
+      "step": 4281
+    },
+    {
+      "epoch": 8.564,
+      "grad_norm": 1.1986794471740723,
+      "learning_rate": 2e-05,
+      "loss": 0.04172123,
+      "step": 4282
+    },
+    {
+      "epoch": 8.566,
+      "grad_norm": 1.0787440538406372,
+      "learning_rate": 2e-05,
+      "loss": 0.03688566,
+      "step": 4283
+    },
+    {
+      "epoch": 8.568,
+      "grad_norm": 1.3387644290924072,
+      "learning_rate": 2e-05,
+      "loss": 0.04537062,
+      "step": 4284
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 2.0228538513183594,
+      "learning_rate": 2e-05,
+      "loss": 0.06465685,
+      "step": 4285
+    },
+    {
+      "epoch": 8.572,
+      "grad_norm": 1.7300091981887817,
+      "learning_rate": 2e-05,
+      "loss": 0.05473095,
+      "step": 4286
+    },
+    {
+      "epoch": 8.574,
+      "grad_norm": 1.032915711402893,
+      "learning_rate": 2e-05,
+      "loss": 0.0316976,
+      "step": 4287
+    },
+    {
+      "epoch": 8.576,
+      "grad_norm": 1.168629765510559,
+      "learning_rate": 2e-05,
+      "loss": 0.0534174,
+      "step": 4288
+    },
+    {
+      "epoch": 8.578,
+      "grad_norm": 1.6948527097702026,
+      "learning_rate": 2e-05,
+      "loss": 0.05644149,
+      "step": 4289
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 1.5725301504135132,
+      "learning_rate": 2e-05,
+      "loss": 0.04966125,
+      "step": 4290
+    },
+    {
+      "epoch": 8.582,
+      "grad_norm": 1.6310228109359741,
+      "learning_rate": 2e-05,
+      "loss": 0.04760915,
+      "step": 4291
+    },
+    {
+      "epoch": 8.584,
+      "grad_norm": 1.232686996459961,
+      "learning_rate": 2e-05,
+      "loss": 0.03924007,
+      "step": 4292
+    },
+    {
+      "epoch": 8.586,
+      "grad_norm": 1.2287647724151611,
+      "learning_rate": 2e-05,
+      "loss": 0.03555786,
+      "step": 4293
+    },
+    {
+      "epoch": 8.588,
+      "grad_norm": 2.1337671279907227,
+      "learning_rate": 2e-05,
+      "loss": 0.03892121,
+      "step": 4294
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 0.8796424865722656,
+      "learning_rate": 2e-05,
+      "loss": 0.02620396,
+      "step": 4295
+    },
+    {
+      "epoch": 8.592,
+      "grad_norm": 1.5610501766204834,
+      "learning_rate": 2e-05,
+      "loss": 0.05278179,
+      "step": 4296
+    },
+    {
+      "epoch": 8.594,
+      "grad_norm": 1.3259083032608032,
+      "learning_rate": 2e-05,
+      "loss": 0.05059385,
+      "step": 4297
+    },
+    {
+      "epoch": 8.596,
+      "grad_norm": 1.2965830564498901,
+      "learning_rate": 2e-05,
+      "loss": 0.04437011,
+      "step": 4298
+    },
+    {
+      "epoch": 8.598,
+      "grad_norm": 1.28528892993927,
+      "learning_rate": 2e-05,
+      "loss": 0.04673284,
+      "step": 4299
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 1.567898154258728,
+      "learning_rate": 2e-05,
+      "loss": 0.05069918,
+      "step": 4300
+    },
+    {
+      "epoch": 8.602,
+      "grad_norm": 1.5278816223144531,
+      "learning_rate": 2e-05,
+      "loss": 0.04407776,
+      "step": 4301
+    },
+    {
+      "epoch": 8.604,
+      "grad_norm": 2.5078423023223877,
+      "learning_rate": 2e-05,
+      "loss": 0.04832874,
+      "step": 4302
+    },
+    {
+      "epoch": 8.606,
+      "grad_norm": 3.015632390975952,
+      "learning_rate": 2e-05,
+      "loss": 0.05268471,
+      "step": 4303
+    },
+    {
+      "epoch": 8.608,
+      "grad_norm": 1.8364429473876953,
+      "learning_rate": 2e-05,
+      "loss": 0.04438996,
+      "step": 4304
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 2.421682119369507,
+      "learning_rate": 2e-05,
+      "loss": 0.05156955,
+      "step": 4305
+    },
+    {
+      "epoch": 8.612,
+      "grad_norm": 1.4038487672805786,
+      "learning_rate": 2e-05,
+      "loss": 0.03900445,
+      "step": 4306
+    },
+    {
+      "epoch": 8.614,
+      "grad_norm": 1.6436742544174194,
+      "learning_rate": 2e-05,
+      "loss": 0.05502493,
+      "step": 4307
+    },
+    {
+      "epoch": 8.616,
+      "grad_norm": 1.551708459854126,
+      "learning_rate": 2e-05,
+      "loss": 0.052692,
+      "step": 4308
+    },
+    {
+      "epoch": 8.618,
+      "grad_norm": 1.148949384689331,
+      "learning_rate": 2e-05,
+      "loss": 0.04152766,
+      "step": 4309
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 1.9535950422286987,
+      "learning_rate": 2e-05,
+      "loss": 0.04875577,
+      "step": 4310
+    },
+    {
+      "epoch": 8.622,
+      "grad_norm": 2.445028066635132,
+      "learning_rate": 2e-05,
+      "loss": 0.05650291,
+      "step": 4311
+    },
+    {
+      "epoch": 8.624,
+      "grad_norm": 1.3438160419464111,
+      "learning_rate": 2e-05,
+      "loss": 0.04274925,
+      "step": 4312
+    },
+    {
+      "epoch": 8.626,
+      "grad_norm": 1.1872248649597168,
+      "learning_rate": 2e-05,
+      "loss": 0.05045023,
+      "step": 4313
+    },
+    {
+      "epoch": 8.628,
+      "grad_norm": 0.948291540145874,
+      "learning_rate": 2e-05,
+      "loss": 0.03525672,
+      "step": 4314
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 1.2028381824493408,
+      "learning_rate": 2e-05,
+      "loss": 0.03516944,
+      "step": 4315
+    },
+    {
+      "epoch": 8.632,
+      "grad_norm": 1.463492512702942,
+      "learning_rate": 2e-05,
+      "loss": 0.05510323,
+      "step": 4316
+    },
+    {
+      "epoch": 8.634,
+      "grad_norm": 1.7272472381591797,
+      "learning_rate": 2e-05,
+      "loss": 0.04611889,
+      "step": 4317
+    },
+    {
+      "epoch": 8.636,
+      "grad_norm": 1.8609179258346558,
+      "learning_rate": 2e-05,
+      "loss": 0.05178494,
+      "step": 4318
+    },
+    {
+      "epoch": 8.638,
+      "grad_norm": 1.6433526277542114,
+      "learning_rate": 2e-05,
+      "loss": 0.05495635,
+      "step": 4319
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 2.0367109775543213,
+      "learning_rate": 2e-05,
+      "loss": 0.05452634,
+      "step": 4320
+    },
+    {
+      "epoch": 8.642,
+      "grad_norm": 1.3078205585479736,
+      "learning_rate": 2e-05,
+      "loss": 0.04912869,
+      "step": 4321
+    },
+    {
+      "epoch": 8.644,
+      "grad_norm": 1.7824225425720215,
+      "learning_rate": 2e-05,
+      "loss": 0.06498539,
+      "step": 4322
+    },
+    {
+      "epoch": 8.646,
+      "grad_norm": 1.2698032855987549,
+      "learning_rate": 2e-05,
+      "loss": 0.03695104,
+      "step": 4323
+    },
+    {
+      "epoch": 8.648,
+      "grad_norm": 1.456852912902832,
+      "learning_rate": 2e-05,
+      "loss": 0.04406907,
+      "step": 4324
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 1.1602133512496948,
+      "learning_rate": 2e-05,
+      "loss": 0.03222954,
+      "step": 4325
+    },
+    {
+      "epoch": 8.652,
+      "grad_norm": 1.945662021636963,
+      "learning_rate": 2e-05,
+      "loss": 0.05972052,
+      "step": 4326
+    },
+    {
+      "epoch": 8.654,
+      "grad_norm": 1.0064737796783447,
+      "learning_rate": 2e-05,
+      "loss": 0.03110239,
+      "step": 4327
+    },
+    {
+      "epoch": 8.656,
+      "grad_norm": 1.220845103263855,
+      "learning_rate": 2e-05,
+      "loss": 0.04152022,
+      "step": 4328
+    },
+    {
+      "epoch": 8.658,
+      "grad_norm": 1.2207328081130981,
+      "learning_rate": 2e-05,
+      "loss": 0.0476977,
+      "step": 4329
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 2.3454084396362305,
+      "learning_rate": 2e-05,
+      "loss": 0.05494661,
+      "step": 4330
+    },
+    {
+      "epoch": 8.662,
+      "grad_norm": 1.1374601125717163,
+      "learning_rate": 2e-05,
+      "loss": 0.04281973,
+      "step": 4331
+    },
+    {
+      "epoch": 8.664,
+      "grad_norm": 1.3726214170455933,
+      "learning_rate": 2e-05,
+      "loss": 0.0457747,
+      "step": 4332
+    },
+    {
+      "epoch": 8.666,
+      "grad_norm": 2.3275725841522217,
+      "learning_rate": 2e-05,
+      "loss": 0.0412515,
+      "step": 4333
+    },
+    {
+      "epoch": 8.668,
+      "grad_norm": 1.3664522171020508,
+      "learning_rate": 2e-05,
+      "loss": 0.04096291,
+      "step": 4334
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 1.359914779663086,
+      "learning_rate": 2e-05,
+      "loss": 0.04645447,
+      "step": 4335
+    },
+    {
+      "epoch": 8.672,
+      "grad_norm": 1.2743496894836426,
+      "learning_rate": 2e-05,
+      "loss": 0.03728715,
+      "step": 4336
+    },
+    {
+      "epoch": 8.674,
+      "grad_norm": 1.1057415008544922,
+      "learning_rate": 2e-05,
+      "loss": 0.03883886,
+      "step": 4337
+    },
+    {
+      "epoch": 8.676,
+      "grad_norm": 2.738429546356201,
+      "learning_rate": 2e-05,
+      "loss": 0.05373093,
+      "step": 4338
+    },
+    {
+      "epoch": 8.678,
+      "grad_norm": 1.3785386085510254,
+      "learning_rate": 2e-05,
+      "loss": 0.04660262,
+      "step": 4339
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 1.4276515245437622,
+      "learning_rate": 2e-05,
+      "loss": 0.04423345,
+      "step": 4340
+    },
+    {
+      "epoch": 8.682,
+      "grad_norm": 1.684853196144104,
+      "learning_rate": 2e-05,
+      "loss": 0.05505417,
+      "step": 4341
+    },
+    {
+      "epoch": 8.684,
+      "grad_norm": 1.370588779449463,
+      "learning_rate": 2e-05,
+      "loss": 0.04044452,
+      "step": 4342
+    },
+    {
+      "epoch": 8.686,
+      "grad_norm": 1.3224636316299438,
+      "learning_rate": 2e-05,
+      "loss": 0.06086275,
+      "step": 4343
+    },
+    {
+      "epoch": 8.688,
+      "grad_norm": 1.181963324546814,
+      "learning_rate": 2e-05,
+      "loss": 0.04808854,
+      "step": 4344
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 1.409883975982666,
+      "learning_rate": 2e-05,
+      "loss": 0.05268676,
+      "step": 4345
+    },
+    {
+      "epoch": 8.692,
+      "grad_norm": 5.802072525024414,
+      "learning_rate": 2e-05,
+      "loss": 0.05748282,
+      "step": 4346
+    },
+    {
+      "epoch": 8.693999999999999,
+      "grad_norm": 1.2217000722885132,
+      "learning_rate": 2e-05,
+      "loss": 0.03751257,
+      "step": 4347
+    },
+    {
+      "epoch": 8.696,
+      "grad_norm": 1.3667843341827393,
+      "learning_rate": 2e-05,
+      "loss": 0.04032121,
+      "step": 4348
+    },
+    {
+      "epoch": 8.698,
+      "grad_norm": 1.1752148866653442,
+      "learning_rate": 2e-05,
+      "loss": 0.03708064,
+      "step": 4349
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 0.9877530336380005,
+      "learning_rate": 2e-05,
+      "loss": 0.02924716,
+      "step": 4350
+    },
+    {
+      "epoch": 8.702,
+      "grad_norm": 1.4895782470703125,
+      "learning_rate": 2e-05,
+      "loss": 0.04259634,
+      "step": 4351
+    },
+    {
+      "epoch": 8.704,
+      "grad_norm": 1.0186421871185303,
+      "learning_rate": 2e-05,
+      "loss": 0.04007004,
+      "step": 4352
+    },
+    {
+      "epoch": 8.706,
+      "grad_norm": 1.8200626373291016,
+      "learning_rate": 2e-05,
+      "loss": 0.05589747,
+      "step": 4353
+    },
+    {
+      "epoch": 8.708,
+      "grad_norm": 1.385703444480896,
+      "learning_rate": 2e-05,
+      "loss": 0.04419235,
+      "step": 4354
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 1.624614953994751,
+      "learning_rate": 2e-05,
+      "loss": 0.03844646,
+      "step": 4355
+    },
+    {
+      "epoch": 8.712,
+      "grad_norm": 1.4218086004257202,
+      "learning_rate": 2e-05,
+      "loss": 0.05011544,
+      "step": 4356
+    },
+    {
+      "epoch": 8.714,
+      "grad_norm": 1.4951738119125366,
+      "learning_rate": 2e-05,
+      "loss": 0.05101977,
+      "step": 4357
+    },
+    {
+      "epoch": 8.716,
+      "grad_norm": 1.324513554573059,
+      "learning_rate": 2e-05,
+      "loss": 0.04258473,
+      "step": 4358
+    },
+    {
+      "epoch": 8.718,
+      "grad_norm": 1.4158241748809814,
+      "learning_rate": 2e-05,
+      "loss": 0.06014551,
+      "step": 4359
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 1.1224169731140137,
+      "learning_rate": 2e-05,
+      "loss": 0.04492978,
+      "step": 4360
+    },
+    {
+      "epoch": 8.722,
+      "grad_norm": 1.1736798286437988,
+      "learning_rate": 2e-05,
+      "loss": 0.04360911,
+      "step": 4361
+    },
+    {
+      "epoch": 8.724,
+      "grad_norm": 1.0621728897094727,
+      "learning_rate": 2e-05,
+      "loss": 0.03657208,
+      "step": 4362
+    },
+    {
+      "epoch": 8.725999999999999,
+      "grad_norm": 1.2689882516860962,
+      "learning_rate": 2e-05,
+      "loss": 0.05324752,
+      "step": 4363
+    },
+    {
+      "epoch": 8.728,
+      "grad_norm": 1.0777446031570435,
+      "learning_rate": 2e-05,
+      "loss": 0.0410941,
+      "step": 4364
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 1.4247167110443115,
+      "learning_rate": 2e-05,
+      "loss": 0.04764949,
+      "step": 4365
+    },
+    {
+      "epoch": 8.732,
+      "grad_norm": 1.272257685661316,
+      "learning_rate": 2e-05,
+      "loss": 0.03613374,
+      "step": 4366
+    },
+    {
+      "epoch": 8.734,
+      "grad_norm": 0.9077991843223572,
+      "learning_rate": 2e-05,
+      "loss": 0.03062226,
+      "step": 4367
+    },
+    {
+      "epoch": 8.736,
+      "grad_norm": 1.2713913917541504,
+      "learning_rate": 2e-05,
+      "loss": 0.05430157,
+      "step": 4368
+    },
+    {
+      "epoch": 8.738,
+      "grad_norm": 1.4644341468811035,
+      "learning_rate": 2e-05,
+      "loss": 0.05187561,
+      "step": 4369
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 1.0588116645812988,
+      "learning_rate": 2e-05,
+      "loss": 0.03304412,
+      "step": 4370
+    },
+    {
+      "epoch": 8.742,
+      "grad_norm": 2.1608660221099854,
+      "learning_rate": 2e-05,
+      "loss": 0.05963966,
+      "step": 4371
+    },
+    {
+      "epoch": 8.744,
+      "grad_norm": 1.0907118320465088,
+      "learning_rate": 2e-05,
+      "loss": 0.03733838,
+      "step": 4372
+    },
+    {
+      "epoch": 8.746,
+      "grad_norm": 2.0541343688964844,
+      "learning_rate": 2e-05,
+      "loss": 0.04198984,
+      "step": 4373
+    },
+    {
+      "epoch": 8.748,
+      "grad_norm": 1.6834027767181396,
+      "learning_rate": 2e-05,
+      "loss": 0.02988002,
+      "step": 4374
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 1.2698630094528198,
+      "learning_rate": 2e-05,
+      "loss": 0.04798315,
+      "step": 4375
+    },
+    {
+      "epoch": 8.752,
+      "grad_norm": 2.5312836170196533,
+      "learning_rate": 2e-05,
+      "loss": 0.0558813,
+      "step": 4376
+    },
+    {
+      "epoch": 8.754,
+      "grad_norm": 1.4217891693115234,
+      "learning_rate": 2e-05,
+      "loss": 0.05483246,
+      "step": 4377
+    },
+    {
+      "epoch": 8.756,
+      "grad_norm": 1.6704591512680054,
+      "learning_rate": 2e-05,
+      "loss": 0.0550387,
+      "step": 4378
+    },
+    {
+      "epoch": 8.758,
+      "grad_norm": 1.5293974876403809,
+      "learning_rate": 2e-05,
+      "loss": 0.05219758,
+      "step": 4379
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 1.3517143726348877,
+      "learning_rate": 2e-05,
+      "loss": 0.04489984,
+      "step": 4380
+    },
+    {
+      "epoch": 8.762,
+      "grad_norm": 1.3553781509399414,
+      "learning_rate": 2e-05,
+      "loss": 0.04696585,
+      "step": 4381
+    },
+    {
+      "epoch": 8.764,
+      "grad_norm": 1.3556764125823975,
+      "learning_rate": 2e-05,
+      "loss": 0.03715477,
+      "step": 4382
+    },
+    {
+      "epoch": 8.766,
+      "grad_norm": 2.0850911140441895,
+      "learning_rate": 2e-05,
+      "loss": 0.05471366,
+      "step": 4383
+    },
+    {
+      "epoch": 8.768,
+      "grad_norm": 1.2767047882080078,
+      "learning_rate": 2e-05,
+      "loss": 0.03740011,
+      "step": 4384
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 1.2590597867965698,
+      "learning_rate": 2e-05,
+      "loss": 0.0351716,
+      "step": 4385
+    },
+    {
+      "epoch": 8.772,
+      "grad_norm": 1.1498363018035889,
+      "learning_rate": 2e-05,
+      "loss": 0.04321297,
+      "step": 4386
+    },
+    {
+      "epoch": 8.774000000000001,
+      "grad_norm": 1.0147833824157715,
+      "learning_rate": 2e-05,
+      "loss": 0.03920402,
+      "step": 4387
+    },
+    {
+      "epoch": 8.776,
+      "grad_norm": 1.6818196773529053,
+      "learning_rate": 2e-05,
+      "loss": 0.05098328,
+      "step": 4388
+    },
+    {
+      "epoch": 8.778,
+      "grad_norm": 1.4138433933258057,
+      "learning_rate": 2e-05,
+      "loss": 0.03823352,
+      "step": 4389
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 1.0351346731185913,
+      "learning_rate": 2e-05,
+      "loss": 0.03335971,
+      "step": 4390
+    },
+    {
+      "epoch": 8.782,
+      "grad_norm": 1.7980780601501465,
+      "learning_rate": 2e-05,
+      "loss": 0.07383931,
+      "step": 4391
+    },
+    {
+      "epoch": 8.784,
+      "grad_norm": 1.1801741123199463,
+      "learning_rate": 2e-05,
+      "loss": 0.04325704,
+      "step": 4392
+    },
+    {
+      "epoch": 8.786,
+      "grad_norm": 0.9786828756332397,
+      "learning_rate": 2e-05,
+      "loss": 0.03433499,
+      "step": 4393
+    },
+    {
+      "epoch": 8.788,
+      "grad_norm": 1.6378285884857178,
+      "learning_rate": 2e-05,
+      "loss": 0.03292269,
+      "step": 4394
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 1.5976024866104126,
+      "learning_rate": 2e-05,
+      "loss": 0.05588251,
+      "step": 4395
+    },
+    {
+      "epoch": 8.792,
+      "grad_norm": 4.473025798797607,
+      "learning_rate": 2e-05,
+      "loss": 0.07918499,
+      "step": 4396
+    },
+    {
+      "epoch": 8.794,
+      "grad_norm": 1.5586071014404297,
+      "learning_rate": 2e-05,
+      "loss": 0.05898192,
+      "step": 4397
+    },
+    {
+      "epoch": 8.796,
+      "grad_norm": 2.6317224502563477,
+      "learning_rate": 2e-05,
+      "loss": 0.05829602,
+      "step": 4398
+    },
+    {
+      "epoch": 8.798,
+      "grad_norm": 1.3767268657684326,
+      "learning_rate": 2e-05,
+      "loss": 0.05414031,
+      "step": 4399
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.9226973652839661,
+      "learning_rate": 2e-05,
+      "loss": 0.02986227,
+      "step": 4400
+    },
+    {
+      "epoch": 8.802,
+      "grad_norm": 2.223020315170288,
+      "learning_rate": 2e-05,
+      "loss": 0.05309587,
+      "step": 4401
+    },
+    {
+      "epoch": 8.804,
+      "grad_norm": 1.2311885356903076,
+      "learning_rate": 2e-05,
+      "loss": 0.04153631,
+      "step": 4402
+    },
+    {
+      "epoch": 8.806000000000001,
+      "grad_norm": 1.4580053091049194,
+      "learning_rate": 2e-05,
+      "loss": 0.0463781,
+      "step": 4403
+    },
+    {
+      "epoch": 8.808,
+      "grad_norm": 1.253326416015625,
+      "learning_rate": 2e-05,
+      "loss": 0.03931756,
+      "step": 4404
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 1.2515385150909424,
+      "learning_rate": 2e-05,
+      "loss": 0.04841346,
+      "step": 4405
+    },
+    {
+      "epoch": 8.812,
+      "grad_norm": 1.8925061225891113,
+      "learning_rate": 2e-05,
+      "loss": 0.04482786,
+      "step": 4406
+    },
+    {
+      "epoch": 8.814,
+      "grad_norm": 1.552807331085205,
+      "learning_rate": 2e-05,
+      "loss": 0.05065498,
+      "step": 4407
+    },
+    {
+      "epoch": 8.816,
+      "grad_norm": 1.1485422849655151,
+      "learning_rate": 2e-05,
+      "loss": 0.04279656,
+      "step": 4408
+    },
+    {
+      "epoch": 8.818,
+      "grad_norm": 1.3679656982421875,
+      "learning_rate": 2e-05,
+      "loss": 0.04369261,
+      "step": 4409
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 1.7776272296905518,
+      "learning_rate": 2e-05,
+      "loss": 0.03488866,
+      "step": 4410
+    },
+    {
+      "epoch": 8.822,
+      "grad_norm": 1.3123103380203247,
+      "learning_rate": 2e-05,
+      "loss": 0.0491695,
+      "step": 4411
+    },
+    {
+      "epoch": 8.824,
+      "grad_norm": 1.5892049074172974,
+      "learning_rate": 2e-05,
+      "loss": 0.04135371,
+      "step": 4412
+    },
+    {
+      "epoch": 8.826,
+      "grad_norm": 1.5765873193740845,
+      "learning_rate": 2e-05,
+      "loss": 0.04748004,
+      "step": 4413
+    },
+    {
+      "epoch": 8.828,
+      "grad_norm": 1.5640965700149536,
+      "learning_rate": 2e-05,
+      "loss": 0.06020439,
+      "step": 4414
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 1.4658604860305786,
+      "learning_rate": 2e-05,
+      "loss": 0.04439153,
+      "step": 4415
+    },
+    {
+      "epoch": 8.832,
+      "grad_norm": 1.7687060832977295,
+      "learning_rate": 2e-05,
+      "loss": 0.05248412,
+      "step": 4416
+    },
+    {
+      "epoch": 8.834,
+      "grad_norm": 1.102990746498108,
+      "learning_rate": 2e-05,
+      "loss": 0.03225242,
+      "step": 4417
+    },
+    {
+      "epoch": 8.836,
+      "grad_norm": 1.2133313417434692,
+      "learning_rate": 2e-05,
+      "loss": 0.03089228,
+      "step": 4418
+    },
+    {
+      "epoch": 8.838,
+      "grad_norm": 2.1788408756256104,
+      "learning_rate": 2e-05,
+      "loss": 0.05756471,
+      "step": 4419
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 1.5459779500961304,
+      "learning_rate": 2e-05,
+      "loss": 0.03825875,
+      "step": 4420
+    },
+    {
+      "epoch": 8.842,
+      "grad_norm": 2.9150943756103516,
+      "learning_rate": 2e-05,
+      "loss": 0.05558386,
+      "step": 4421
+    },
+    {
+      "epoch": 8.844,
+      "grad_norm": 1.6710635423660278,
+      "learning_rate": 2e-05,
+      "loss": 0.06062218,
+      "step": 4422
+    },
+    {
+      "epoch": 8.846,
+      "grad_norm": 1.0483447313308716,
+      "learning_rate": 2e-05,
+      "loss": 0.03988266,
+      "step": 4423
+    },
+    {
+      "epoch": 8.848,
+      "grad_norm": 1.760643720626831,
+      "learning_rate": 2e-05,
+      "loss": 0.04801723,
+      "step": 4424
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 1.5282479524612427,
+      "learning_rate": 2e-05,
+      "loss": 0.05213196,
+      "step": 4425
+    },
+    {
+      "epoch": 8.852,
+      "grad_norm": 1.7912678718566895,
+      "learning_rate": 2e-05,
+      "loss": 0.04636016,
+      "step": 4426
+    },
+    {
+      "epoch": 8.854,
+      "grad_norm": 0.8681804537773132,
+      "learning_rate": 2e-05,
+      "loss": 0.03162609,
+      "step": 4427
+    },
+    {
+      "epoch": 8.856,
+      "grad_norm": 1.3080195188522339,
+      "learning_rate": 2e-05,
+      "loss": 0.05422079,
+      "step": 4428
+    },
+    {
+      "epoch": 8.858,
+      "grad_norm": 1.337620735168457,
+      "learning_rate": 2e-05,
+      "loss": 0.05549601,
+      "step": 4429
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 1.0535590648651123,
+      "learning_rate": 2e-05,
+      "loss": 0.04026527,
+      "step": 4430
+    },
+    {
+      "epoch": 8.862,
+      "grad_norm": 1.1052578687667847,
+      "learning_rate": 2e-05,
+      "loss": 0.03884507,
+      "step": 4431
+    },
+    {
+      "epoch": 8.864,
+      "grad_norm": 1.7922765016555786,
+      "learning_rate": 2e-05,
+      "loss": 0.05645222,
+      "step": 4432
+    },
+    {
+      "epoch": 8.866,
+      "grad_norm": 1.553610920906067,
+      "learning_rate": 2e-05,
+      "loss": 0.055147,
+      "step": 4433
+    },
+    {
+      "epoch": 8.868,
+      "grad_norm": 1.6815420389175415,
+      "learning_rate": 2e-05,
+      "loss": 0.0513879,
+      "step": 4434
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 2.2995853424072266,
+      "learning_rate": 2e-05,
+      "loss": 0.0562899,
+      "step": 4435
+    },
+    {
+      "epoch": 8.872,
+      "grad_norm": 1.9725992679595947,
+      "learning_rate": 2e-05,
+      "loss": 0.057873,
+      "step": 4436
+    },
+    {
+      "epoch": 8.874,
+      "grad_norm": 2.180466890335083,
+      "learning_rate": 2e-05,
+      "loss": 0.04819129,
+      "step": 4437
+    },
+    {
+      "epoch": 8.876,
+      "grad_norm": 1.4554449319839478,
+      "learning_rate": 2e-05,
+      "loss": 0.04344202,
+      "step": 4438
+    },
+    {
+      "epoch": 8.878,
+      "grad_norm": 1.6162256002426147,
+      "learning_rate": 2e-05,
+      "loss": 0.04701218,
+      "step": 4439
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 0.8536230325698853,
+      "learning_rate": 2e-05,
+      "loss": 0.02937111,
+      "step": 4440
+    },
+    {
+      "epoch": 8.882,
+      "grad_norm": 1.6734143495559692,
+      "learning_rate": 2e-05,
+      "loss": 0.05779783,
+      "step": 4441
+    },
+    {
+      "epoch": 8.884,
+      "grad_norm": 1.528039574623108,
+      "learning_rate": 2e-05,
+      "loss": 0.0378891,
+      "step": 4442
+    },
+    {
+      "epoch": 8.886,
+      "grad_norm": 1.4532936811447144,
+      "learning_rate": 2e-05,
+      "loss": 0.04875976,
+      "step": 4443
+    },
+    {
+      "epoch": 8.888,
+      "grad_norm": 1.1509127616882324,
+      "learning_rate": 2e-05,
+      "loss": 0.04518646,
+      "step": 4444
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 1.0691490173339844,
+      "learning_rate": 2e-05,
+      "loss": 0.0388993,
+      "step": 4445
+    },
+    {
+      "epoch": 8.892,
+      "grad_norm": 1.3075459003448486,
+      "learning_rate": 2e-05,
+      "loss": 0.0404879,
+      "step": 4446
+    },
+    {
+      "epoch": 8.894,
+      "grad_norm": 1.5355843305587769,
+      "learning_rate": 2e-05,
+      "loss": 0.06093806,
+      "step": 4447
+    },
+    {
+      "epoch": 8.896,
+      "grad_norm": 1.2035893201828003,
+      "learning_rate": 2e-05,
+      "loss": 0.03944387,
+      "step": 4448
+    },
+    {
+      "epoch": 8.898,
+      "grad_norm": 1.2087056636810303,
+      "learning_rate": 2e-05,
+      "loss": 0.04656249,
+      "step": 4449
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 1.3271777629852295,
+      "learning_rate": 2e-05,
+      "loss": 0.04627533,
+      "step": 4450
+    },
+    {
+      "epoch": 8.902,
+      "grad_norm": 1.2811769247055054,
+      "learning_rate": 2e-05,
+      "loss": 0.04390157,
+      "step": 4451
+    },
+    {
+      "epoch": 8.904,
+      "grad_norm": 1.3496356010437012,
+      "learning_rate": 2e-05,
+      "loss": 0.05148339,
+      "step": 4452
+    },
+    {
+      "epoch": 8.906,
+      "grad_norm": 2.2644078731536865,
+      "learning_rate": 2e-05,
+      "loss": 0.06438026,
+      "step": 4453
+    },
+    {
+      "epoch": 8.908,
+      "grad_norm": 0.9715155363082886,
+      "learning_rate": 2e-05,
+      "loss": 0.02947888,
+      "step": 4454
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 1.787411093711853,
+      "learning_rate": 2e-05,
+      "loss": 0.04107827,
+      "step": 4455
+    },
+    {
+      "epoch": 8.912,
+      "grad_norm": 1.1115169525146484,
+      "learning_rate": 2e-05,
+      "loss": 0.0399262,
+      "step": 4456
+    },
+    {
+      "epoch": 8.914,
+      "grad_norm": 1.2536200284957886,
+      "learning_rate": 2e-05,
+      "loss": 0.03909591,
+      "step": 4457
+    },
+    {
+      "epoch": 8.916,
+      "grad_norm": 1.3224055767059326,
+      "learning_rate": 2e-05,
+      "loss": 0.04420022,
+      "step": 4458
+    },
+    {
+      "epoch": 8.918,
+      "grad_norm": 1.4246692657470703,
+      "learning_rate": 2e-05,
+      "loss": 0.03702656,
+      "step": 4459
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 1.9952268600463867,
+      "learning_rate": 2e-05,
+      "loss": 0.05900434,
+      "step": 4460
+    },
+    {
+      "epoch": 8.922,
+      "grad_norm": 2.0388376712799072,
+      "learning_rate": 2e-05,
+      "loss": 0.05166218,
+      "step": 4461
+    },
+    {
+      "epoch": 8.924,
+      "grad_norm": 1.1385499238967896,
+      "learning_rate": 2e-05,
+      "loss": 0.04556513,
+      "step": 4462
+    },
+    {
+      "epoch": 8.926,
+      "grad_norm": 1.5945521593093872,
+      "learning_rate": 2e-05,
+      "loss": 0.0496194,
+      "step": 4463
+    },
+    {
+      "epoch": 8.928,
+      "grad_norm": 1.6476668119430542,
+      "learning_rate": 2e-05,
+      "loss": 0.04806423,
+      "step": 4464
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 1.9072840213775635,
+      "learning_rate": 2e-05,
+      "loss": 0.05876537,
+      "step": 4465
+    },
+    {
+      "epoch": 8.932,
+      "grad_norm": 1.7775014638900757,
+      "learning_rate": 2e-05,
+      "loss": 0.05251986,
+      "step": 4466
+    },
+    {
+      "epoch": 8.934,
+      "grad_norm": 1.2985152006149292,
+      "learning_rate": 2e-05,
+      "loss": 0.0463018,
+      "step": 4467
+    },
+    {
+      "epoch": 8.936,
+      "grad_norm": 1.8547874689102173,
+      "learning_rate": 2e-05,
+      "loss": 0.06041159,
+      "step": 4468
+    },
+    {
+      "epoch": 8.938,
+      "grad_norm": 2.1976423263549805,
+      "learning_rate": 2e-05,
+      "loss": 0.03414884,
+      "step": 4469
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 0.9419854283332825,
+      "learning_rate": 2e-05,
+      "loss": 0.02789684,
+      "step": 4470
+    },
+    {
+      "epoch": 8.942,
+      "grad_norm": 1.1904888153076172,
+      "learning_rate": 2e-05,
+      "loss": 0.04499068,
+      "step": 4471
+    },
+    {
+      "epoch": 8.943999999999999,
+      "grad_norm": 1.1681838035583496,
+      "learning_rate": 2e-05,
+      "loss": 0.04271141,
+      "step": 4472
+    },
+    {
+      "epoch": 8.946,
+      "grad_norm": 1.388636827468872,
+      "learning_rate": 2e-05,
+      "loss": 0.04021178,
+      "step": 4473
+    },
+    {
+      "epoch": 8.948,
+      "grad_norm": 1.6245975494384766,
+      "learning_rate": 2e-05,
+      "loss": 0.05697055,
+      "step": 4474
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 1.1198829412460327,
+      "learning_rate": 2e-05,
+      "loss": 0.03857917,
+      "step": 4475
+    },
+    {
+      "epoch": 8.952,
+      "grad_norm": 1.9190189838409424,
+      "learning_rate": 2e-05,
+      "loss": 0.04297666,
+      "step": 4476
+    },
+    {
+      "epoch": 8.954,
+      "grad_norm": 1.584479808807373,
+      "learning_rate": 2e-05,
+      "loss": 0.06143427,
+      "step": 4477
+    },
+    {
+      "epoch": 8.956,
+      "grad_norm": 1.6050777435302734,
+      "learning_rate": 2e-05,
+      "loss": 0.04966013,
+      "step": 4478
+    },
+    {
+      "epoch": 8.958,
+      "grad_norm": 2.1203346252441406,
+      "learning_rate": 2e-05,
+      "loss": 0.06286226,
+      "step": 4479
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 1.5613123178482056,
+      "learning_rate": 2e-05,
+      "loss": 0.04335446,
+      "step": 4480
+    },
+    {
+      "epoch": 8.962,
+      "grad_norm": 1.7688831090927124,
+      "learning_rate": 2e-05,
+      "loss": 0.05140468,
+      "step": 4481
+    },
+    {
+      "epoch": 8.964,
+      "grad_norm": 1.1225483417510986,
+      "learning_rate": 2e-05,
+      "loss": 0.03839725,
+      "step": 4482
+    },
+    {
+      "epoch": 8.966,
+      "grad_norm": 1.546858549118042,
+      "learning_rate": 2e-05,
+      "loss": 0.04927346,
+      "step": 4483
+    },
+    {
+      "epoch": 8.968,
+      "grad_norm": 1.4144601821899414,
+      "learning_rate": 2e-05,
+      "loss": 0.0489133,
+      "step": 4484
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 1.2596808671951294,
+      "learning_rate": 2e-05,
+      "loss": 0.04895072,
+      "step": 4485
+    },
+    {
+      "epoch": 8.972,
+      "grad_norm": 1.3473176956176758,
+      "learning_rate": 2e-05,
+      "loss": 0.04276847,
+      "step": 4486
+    },
+    {
+      "epoch": 8.974,
+      "grad_norm": 1.0893629789352417,
+      "learning_rate": 2e-05,
+      "loss": 0.02666626,
+      "step": 4487
+    },
+    {
+      "epoch": 8.975999999999999,
+      "grad_norm": 1.3393633365631104,
+      "learning_rate": 2e-05,
+      "loss": 0.04725213,
+      "step": 4488
+    },
+    {
+      "epoch": 8.978,
+      "grad_norm": 1.2494399547576904,
+      "learning_rate": 2e-05,
+      "loss": 0.04335026,
+      "step": 4489
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 1.394800066947937,
+      "learning_rate": 2e-05,
+      "loss": 0.05108353,
+      "step": 4490
+    },
+    {
+      "epoch": 8.982,
+      "grad_norm": 2.5832700729370117,
+      "learning_rate": 2e-05,
+      "loss": 0.03944442,
+      "step": 4491
+    },
+    {
+      "epoch": 8.984,
+      "grad_norm": 2.0006890296936035,
+      "learning_rate": 2e-05,
+      "loss": 0.06387276,
+      "step": 4492
+    },
+    {
+      "epoch": 8.986,
+      "grad_norm": 1.664270043373108,
+      "learning_rate": 2e-05,
+      "loss": 0.0435832,
+      "step": 4493
+    },
+    {
+      "epoch": 8.988,
+      "grad_norm": 1.6541657447814941,
+      "learning_rate": 2e-05,
+      "loss": 0.05543558,
+      "step": 4494
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 1.3456977605819702,
+      "learning_rate": 2e-05,
+      "loss": 0.03495708,
+      "step": 4495
+    },
+    {
+      "epoch": 8.992,
+      "grad_norm": 1.4035969972610474,
+      "learning_rate": 2e-05,
+      "loss": 0.04944149,
+      "step": 4496
+    },
+    {
+      "epoch": 8.994,
+      "grad_norm": 2.1596810817718506,
+      "learning_rate": 2e-05,
+      "loss": 0.0587335,
+      "step": 4497
+    },
+    {
+      "epoch": 8.996,
+      "grad_norm": 1.4159624576568604,
+      "learning_rate": 2e-05,
+      "loss": 0.04697796,
+      "step": 4498
+    },
+    {
+      "epoch": 8.998,
+      "grad_norm": 2.2335352897644043,
+      "learning_rate": 2e-05,
+      "loss": 0.05628205,
+      "step": 4499
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 1.6927722692489624,
+      "learning_rate": 2e-05,
+      "loss": 0.04296227,
+      "step": 4500
+    },
+    {
+      "epoch": 9.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.992,
+        "AngleClassification_2": 0.99,
+        "AngleClassification_3": 0.8922155688622755,
+        "Equal_1": 0.99,
+        "Equal_2": 0.9600798403193613,
+        "Equal_3": 0.8662674650698603,
+        "LineComparison_1": 0.998,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9700598802395209,
+        "Parallel_1": 0.969939879759519,
+        "Parallel_2": 0.9919839679358717,
+        "Parallel_3": 0.988,
+        "Perpendicular_1": 0.992,
+        "Perpendicular_2": 0.828,
+        "Perpendicular_3": 0.533066132264529,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9896666666666667,
+        "PointLiesOnCircle_3": 0.9916,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9939879759519038,
+        "PointLiesOnLine_3": 0.9041916167664671
+      },
+      "eval_runtime": 320.19,
+      "eval_samples_per_second": 32.793,
+      "eval_steps_per_second": 0.656,
+      "step": 4500
+    },
+    {
+      "epoch": 9.002,
+      "grad_norm": 2.0088605880737305,
+      "learning_rate": 2e-05,
+      "loss": 0.05657805,
+      "step": 4501
+    },
+    {
+      "epoch": 9.004,
+      "grad_norm": 3.5889363288879395,
+      "learning_rate": 2e-05,
+      "loss": 0.08259207,
+      "step": 4502
+    },
+    {
+      "epoch": 9.006,
+      "grad_norm": 2.149068593978882,
+      "learning_rate": 2e-05,
+      "loss": 0.05930202,
+      "step": 4503
+    },
+    {
+      "epoch": 9.008,
+      "grad_norm": 4.037459373474121,
+      "learning_rate": 2e-05,
+      "loss": 0.0556262,
+      "step": 4504
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 1.781816840171814,
+      "learning_rate": 2e-05,
+      "loss": 0.05204342,
+      "step": 4505
+    },
+    {
+      "epoch": 9.012,
+      "grad_norm": 2.256213903427124,
+      "learning_rate": 2e-05,
+      "loss": 0.07549164,
+      "step": 4506
+    },
+    {
+      "epoch": 9.014,
+      "grad_norm": 2.2270400524139404,
+      "learning_rate": 2e-05,
+      "loss": 0.10077457,
+      "step": 4507
+    },
+    {
+      "epoch": 9.016,
+      "grad_norm": 1.9802483320236206,
+      "learning_rate": 2e-05,
+      "loss": 0.0692708,
+      "step": 4508
+    },
+    {
+      "epoch": 9.018,
+      "grad_norm": 1.4657119512557983,
+      "learning_rate": 2e-05,
+      "loss": 0.05361843,
+      "step": 4509
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 1.6600518226623535,
+      "learning_rate": 2e-05,
+      "loss": 0.05417612,
+      "step": 4510
+    },
+    {
+      "epoch": 9.022,
+      "grad_norm": 1.6829038858413696,
+      "learning_rate": 2e-05,
+      "loss": 0.05836887,
+      "step": 4511
+    },
+    {
+      "epoch": 9.024,
+      "grad_norm": 1.7397302389144897,
+      "learning_rate": 2e-05,
+      "loss": 0.06527451,
+      "step": 4512
+    },
+    {
+      "epoch": 9.026,
+      "grad_norm": 2.1521365642547607,
+      "learning_rate": 2e-05,
+      "loss": 0.0745489,
+      "step": 4513
+    },
+    {
+      "epoch": 9.028,
+      "grad_norm": 1.5954670906066895,
+      "learning_rate": 2e-05,
+      "loss": 0.04902204,
+      "step": 4514
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 1.6460261344909668,
+      "learning_rate": 2e-05,
+      "loss": 0.05216154,
+      "step": 4515
+    },
+    {
+      "epoch": 9.032,
+      "grad_norm": 1.4781848192214966,
+      "learning_rate": 2e-05,
+      "loss": 0.05158208,
+      "step": 4516
+    },
+    {
+      "epoch": 9.034,
+      "grad_norm": 1.3276011943817139,
+      "learning_rate": 2e-05,
+      "loss": 0.04840251,
+      "step": 4517
+    },
+    {
+      "epoch": 9.036,
+      "grad_norm": 1.4326094388961792,
+      "learning_rate": 2e-05,
+      "loss": 0.06191067,
+      "step": 4518
+    },
+    {
+      "epoch": 9.038,
+      "grad_norm": 1.330862045288086,
+      "learning_rate": 2e-05,
+      "loss": 0.05839656,
+      "step": 4519
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 2.0986196994781494,
+      "learning_rate": 2e-05,
+      "loss": 0.06784381,
+      "step": 4520
+    },
+    {
+      "epoch": 9.042,
+      "grad_norm": 2.1825203895568848,
+      "learning_rate": 2e-05,
+      "loss": 0.04452966,
+      "step": 4521
+    },
+    {
+      "epoch": 9.044,
+      "grad_norm": 1.90194571018219,
+      "learning_rate": 2e-05,
+      "loss": 0.06857677,
+      "step": 4522
+    },
+    {
+      "epoch": 9.046,
+      "grad_norm": 2.0310020446777344,
+      "learning_rate": 2e-05,
+      "loss": 0.04652988,
+      "step": 4523
+    },
+    {
+      "epoch": 9.048,
+      "grad_norm": 1.5352383852005005,
+      "learning_rate": 2e-05,
+      "loss": 0.04297383,
+      "step": 4524
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 2.4602177143096924,
+      "learning_rate": 2e-05,
+      "loss": 0.04376163,
+      "step": 4525
+    },
+    {
+      "epoch": 9.052,
+      "grad_norm": 1.6497063636779785,
+      "learning_rate": 2e-05,
+      "loss": 0.0518771,
+      "step": 4526
+    },
+    {
+      "epoch": 9.054,
+      "grad_norm": 1.3094106912612915,
+      "learning_rate": 2e-05,
+      "loss": 0.04263393,
+      "step": 4527
+    },
+    {
+      "epoch": 9.056,
+      "grad_norm": 2.1495842933654785,
+      "learning_rate": 2e-05,
+      "loss": 0.06488663,
+      "step": 4528
+    },
+    {
+      "epoch": 9.058,
+      "grad_norm": 1.4335927963256836,
+      "learning_rate": 2e-05,
+      "loss": 0.04027612,
+      "step": 4529
+    },
+    {
+      "epoch": 9.06,
+      "grad_norm": 1.6587024927139282,
+      "learning_rate": 2e-05,
+      "loss": 0.06693022,
+      "step": 4530
+    },
+    {
+      "epoch": 9.062,
+      "grad_norm": 3.3605124950408936,
+      "learning_rate": 2e-05,
+      "loss": 0.07780167,
+      "step": 4531
+    },
+    {
+      "epoch": 9.064,
+      "grad_norm": 1.3746273517608643,
+      "learning_rate": 2e-05,
+      "loss": 0.06263963,
+      "step": 4532
+    },
+    {
+      "epoch": 9.066,
+      "grad_norm": 1.8702830076217651,
+      "learning_rate": 2e-05,
+      "loss": 0.04445141,
+      "step": 4533
+    },
+    {
+      "epoch": 9.068,
+      "grad_norm": 1.272550344467163,
+      "learning_rate": 2e-05,
+      "loss": 0.04481889,
+      "step": 4534
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 1.7748891115188599,
+      "learning_rate": 2e-05,
+      "loss": 0.04557388,
+      "step": 4535
+    },
+    {
+      "epoch": 9.072,
+      "grad_norm": 6.206629276275635,
+      "learning_rate": 2e-05,
+      "loss": 0.07212655,
+      "step": 4536
+    },
+    {
+      "epoch": 9.074,
+      "grad_norm": 1.981056809425354,
+      "learning_rate": 2e-05,
+      "loss": 0.05645072,
+      "step": 4537
+    },
+    {
+      "epoch": 9.076,
+      "grad_norm": 2.8909831047058105,
+      "learning_rate": 2e-05,
+      "loss": 0.05474135,
+      "step": 4538
+    },
+    {
+      "epoch": 9.078,
+      "grad_norm": 1.981854796409607,
+      "learning_rate": 2e-05,
+      "loss": 0.05070494,
+      "step": 4539
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 1.315761923789978,
+      "learning_rate": 2e-05,
+      "loss": 0.04502026,
+      "step": 4540
+    },
+    {
+      "epoch": 9.082,
+      "grad_norm": 1.2324833869934082,
+      "learning_rate": 2e-05,
+      "loss": 0.04134127,
+      "step": 4541
+    },
+    {
+      "epoch": 9.084,
+      "grad_norm": 2.135284185409546,
+      "learning_rate": 2e-05,
+      "loss": 0.05175925,
+      "step": 4542
+    },
+    {
+      "epoch": 9.086,
+      "grad_norm": 4.033361434936523,
+      "learning_rate": 2e-05,
+      "loss": 0.07532454,
+      "step": 4543
+    },
+    {
+      "epoch": 9.088,
+      "grad_norm": 1.9662163257598877,
+      "learning_rate": 2e-05,
+      "loss": 0.05289062,
+      "step": 4544
+    },
+    {
+      "epoch": 9.09,
+      "grad_norm": 1.4949467182159424,
+      "learning_rate": 2e-05,
+      "loss": 0.05135939,
+      "step": 4545
+    },
+    {
+      "epoch": 9.092,
+      "grad_norm": 1.5621041059494019,
+      "learning_rate": 2e-05,
+      "loss": 0.05257202,
+      "step": 4546
+    },
+    {
+      "epoch": 9.094,
+      "grad_norm": 2.7926948070526123,
+      "learning_rate": 2e-05,
+      "loss": 0.05770051,
+      "step": 4547
+    },
+    {
+      "epoch": 9.096,
+      "grad_norm": 1.2026070356369019,
+      "learning_rate": 2e-05,
+      "loss": 0.03686951,
+      "step": 4548
+    },
+    {
+      "epoch": 9.098,
+      "grad_norm": 1.5503953695297241,
+      "learning_rate": 2e-05,
+      "loss": 0.06764029,
+      "step": 4549
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 1.4695518016815186,
+      "learning_rate": 2e-05,
+      "loss": 0.040582,
+      "step": 4550
+    },
+    {
+      "epoch": 9.102,
+      "grad_norm": 1.5409952402114868,
+      "learning_rate": 2e-05,
+      "loss": 0.04492978,
+      "step": 4551
+    },
+    {
+      "epoch": 9.104,
+      "grad_norm": 1.1860859394073486,
+      "learning_rate": 2e-05,
+      "loss": 0.03985904,
+      "step": 4552
+    },
+    {
+      "epoch": 9.106,
+      "grad_norm": 1.1103501319885254,
+      "learning_rate": 2e-05,
+      "loss": 0.03564502,
+      "step": 4553
+    },
+    {
+      "epoch": 9.108,
+      "grad_norm": 3.4660491943359375,
+      "learning_rate": 2e-05,
+      "loss": 0.07822317,
+      "step": 4554
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 2.640744686126709,
+      "learning_rate": 2e-05,
+      "loss": 0.05545685,
+      "step": 4555
+    },
+    {
+      "epoch": 9.112,
+      "grad_norm": 2.2998898029327393,
+      "learning_rate": 2e-05,
+      "loss": 0.06051971,
+      "step": 4556
+    },
+    {
+      "epoch": 9.114,
+      "grad_norm": 2.3359665870666504,
+      "learning_rate": 2e-05,
+      "loss": 0.06999737,
+      "step": 4557
+    },
+    {
+      "epoch": 9.116,
+      "grad_norm": 1.607606291770935,
+      "learning_rate": 2e-05,
+      "loss": 0.05824485,
+      "step": 4558
+    },
+    {
+      "epoch": 9.118,
+      "grad_norm": 1.0786031484603882,
+      "learning_rate": 2e-05,
+      "loss": 0.04007483,
+      "step": 4559
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 2.296783924102783,
+      "learning_rate": 2e-05,
+      "loss": 0.04448222,
+      "step": 4560
+    },
+    {
+      "epoch": 9.122,
+      "grad_norm": 1.5085766315460205,
+      "learning_rate": 2e-05,
+      "loss": 0.03726997,
+      "step": 4561
+    },
+    {
+      "epoch": 9.124,
+      "grad_norm": 2.368339776992798,
+      "learning_rate": 2e-05,
+      "loss": 0.05795577,
+      "step": 4562
+    },
+    {
+      "epoch": 9.126,
+      "grad_norm": 1.6994584798812866,
+      "learning_rate": 2e-05,
+      "loss": 0.04269854,
+      "step": 4563
+    },
+    {
+      "epoch": 9.128,
+      "grad_norm": 1.7187951803207397,
+      "learning_rate": 2e-05,
+      "loss": 0.05323853,
+      "step": 4564
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 2.053581953048706,
+      "learning_rate": 2e-05,
+      "loss": 0.0678404,
+      "step": 4565
+    },
+    {
+      "epoch": 9.132,
+      "grad_norm": 1.663622260093689,
+      "learning_rate": 2e-05,
+      "loss": 0.04938076,
+      "step": 4566
+    },
+    {
+      "epoch": 9.134,
+      "grad_norm": 2.431302785873413,
+      "learning_rate": 2e-05,
+      "loss": 0.08686221,
+      "step": 4567
+    },
+    {
+      "epoch": 9.136,
+      "grad_norm": 1.7747234106063843,
+      "learning_rate": 2e-05,
+      "loss": 0.06363907,
+      "step": 4568
+    },
+    {
+      "epoch": 9.138,
+      "grad_norm": 2.0806329250335693,
+      "learning_rate": 2e-05,
+      "loss": 0.06777623,
+      "step": 4569
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 1.9430458545684814,
+      "learning_rate": 2e-05,
+      "loss": 0.05345959,
+      "step": 4570
+    },
+    {
+      "epoch": 9.142,
+      "grad_norm": 1.6341060400009155,
+      "learning_rate": 2e-05,
+      "loss": 0.06348822,
+      "step": 4571
+    },
+    {
+      "epoch": 9.144,
+      "grad_norm": 2.909865140914917,
+      "learning_rate": 2e-05,
+      "loss": 0.06166807,
+      "step": 4572
+    },
+    {
+      "epoch": 9.146,
+      "grad_norm": 2.0202066898345947,
+      "learning_rate": 2e-05,
+      "loss": 0.06222503,
+      "step": 4573
+    },
+    {
+      "epoch": 9.148,
+      "grad_norm": 1.8997862339019775,
+      "learning_rate": 2e-05,
+      "loss": 0.04913793,
+      "step": 4574
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 1.336957573890686,
+      "learning_rate": 2e-05,
+      "loss": 0.07197899,
+      "step": 4575
+    },
+    {
+      "epoch": 9.152,
+      "grad_norm": 1.426227331161499,
+      "learning_rate": 2e-05,
+      "loss": 0.05871036,
+      "step": 4576
+    },
+    {
+      "epoch": 9.154,
+      "grad_norm": 1.507077693939209,
+      "learning_rate": 2e-05,
+      "loss": 0.04596441,
+      "step": 4577
+    },
+    {
+      "epoch": 9.156,
+      "grad_norm": 1.3802088499069214,
+      "learning_rate": 2e-05,
+      "loss": 0.04962367,
+      "step": 4578
+    },
+    {
+      "epoch": 9.158,
+      "grad_norm": 1.2211660146713257,
+      "learning_rate": 2e-05,
+      "loss": 0.04041466,
+      "step": 4579
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 1.5546852350234985,
+      "learning_rate": 2e-05,
+      "loss": 0.06333128,
+      "step": 4580
+    },
+    {
+      "epoch": 9.162,
+      "grad_norm": 1.5927609205245972,
+      "learning_rate": 2e-05,
+      "loss": 0.05327647,
+      "step": 4581
+    },
+    {
+      "epoch": 9.164,
+      "grad_norm": 1.427751898765564,
+      "learning_rate": 2e-05,
+      "loss": 0.04664875,
+      "step": 4582
+    },
+    {
+      "epoch": 9.166,
+      "grad_norm": 1.6107282638549805,
+      "learning_rate": 2e-05,
+      "loss": 0.06387123,
+      "step": 4583
+    },
+    {
+      "epoch": 9.168,
+      "grad_norm": 2.082274913787842,
+      "learning_rate": 2e-05,
+      "loss": 0.07043574,
+      "step": 4584
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 1.6113282442092896,
+      "learning_rate": 2e-05,
+      "loss": 0.05279049,
+      "step": 4585
+    },
+    {
+      "epoch": 9.172,
+      "grad_norm": 1.2764414548873901,
+      "learning_rate": 2e-05,
+      "loss": 0.05512537,
+      "step": 4586
+    },
+    {
+      "epoch": 9.174,
+      "grad_norm": 1.731431007385254,
+      "learning_rate": 2e-05,
+      "loss": 0.05489038,
+      "step": 4587
+    },
+    {
+      "epoch": 9.176,
+      "grad_norm": 2.2795941829681396,
+      "learning_rate": 2e-05,
+      "loss": 0.05099697,
+      "step": 4588
+    },
+    {
+      "epoch": 9.178,
+      "grad_norm": 1.9231106042861938,
+      "learning_rate": 2e-05,
+      "loss": 0.0609756,
+      "step": 4589
+    },
+    {
+      "epoch": 9.18,
+      "grad_norm": 1.3888475894927979,
+      "learning_rate": 2e-05,
+      "loss": 0.04252925,
+      "step": 4590
+    },
+    {
+      "epoch": 9.182,
+      "grad_norm": 1.7997218370437622,
+      "learning_rate": 2e-05,
+      "loss": 0.0603526,
+      "step": 4591
+    },
+    {
+      "epoch": 9.184,
+      "grad_norm": 2.605591297149658,
+      "learning_rate": 2e-05,
+      "loss": 0.05147631,
+      "step": 4592
+    },
+    {
+      "epoch": 9.186,
+      "grad_norm": 1.1323707103729248,
+      "learning_rate": 2e-05,
+      "loss": 0.03506859,
+      "step": 4593
+    },
+    {
+      "epoch": 9.188,
+      "grad_norm": 1.9055380821228027,
+      "learning_rate": 2e-05,
+      "loss": 0.05622992,
+      "step": 4594
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 1.66822350025177,
+      "learning_rate": 2e-05,
+      "loss": 0.047398,
+      "step": 4595
+    },
+    {
+      "epoch": 9.192,
+      "grad_norm": 1.536523461341858,
+      "learning_rate": 2e-05,
+      "loss": 0.05335706,
+      "step": 4596
+    },
+    {
+      "epoch": 9.194,
+      "grad_norm": 2.4968795776367188,
+      "learning_rate": 2e-05,
+      "loss": 0.04762807,
+      "step": 4597
+    },
+    {
+      "epoch": 9.196,
+      "grad_norm": 1.6647475957870483,
+      "learning_rate": 2e-05,
+      "loss": 0.06161051,
+      "step": 4598
+    },
+    {
+      "epoch": 9.198,
+      "grad_norm": 2.187264919281006,
+      "learning_rate": 2e-05,
+      "loss": 0.08325572,
+      "step": 4599
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 1.3294416666030884,
+      "learning_rate": 2e-05,
+      "loss": 0.05907375,
+      "step": 4600
+    },
+    {
+      "epoch": 9.202,
+      "grad_norm": 1.3682829141616821,
+      "learning_rate": 2e-05,
+      "loss": 0.0561332,
+      "step": 4601
+    },
+    {
+      "epoch": 9.204,
+      "grad_norm": 3.3193435668945312,
+      "learning_rate": 2e-05,
+      "loss": 0.06492063,
+      "step": 4602
+    },
+    {
+      "epoch": 9.206,
+      "grad_norm": 2.0114078521728516,
+      "learning_rate": 2e-05,
+      "loss": 0.06080581,
+      "step": 4603
+    },
+    {
+      "epoch": 9.208,
+      "grad_norm": 1.9517841339111328,
+      "learning_rate": 2e-05,
+      "loss": 0.06941114,
+      "step": 4604
+    },
+    {
+      "epoch": 9.21,
+      "grad_norm": 1.9020707607269287,
+      "learning_rate": 2e-05,
+      "loss": 0.06470014,
+      "step": 4605
+    },
+    {
+      "epoch": 9.212,
+      "grad_norm": 1.7281121015548706,
+      "learning_rate": 2e-05,
+      "loss": 0.05396245,
+      "step": 4606
+    },
+    {
+      "epoch": 9.214,
+      "grad_norm": 3.0623199939727783,
+      "learning_rate": 2e-05,
+      "loss": 0.08490207,
+      "step": 4607
+    },
+    {
+      "epoch": 9.216,
+      "grad_norm": 2.61993670463562,
+      "learning_rate": 2e-05,
+      "loss": 0.06834012,
+      "step": 4608
+    },
+    {
+      "epoch": 9.218,
+      "grad_norm": 1.3250088691711426,
+      "learning_rate": 2e-05,
+      "loss": 0.04134344,
+      "step": 4609
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 2.761884927749634,
+      "learning_rate": 2e-05,
+      "loss": 0.06493714,
+      "step": 4610
+    },
+    {
+      "epoch": 9.222,
+      "grad_norm": 1.54371178150177,
+      "learning_rate": 2e-05,
+      "loss": 0.05743696,
+      "step": 4611
+    },
+    {
+      "epoch": 9.224,
+      "grad_norm": 5.61605167388916,
+      "learning_rate": 2e-05,
+      "loss": 0.06084044,
+      "step": 4612
+    },
+    {
+      "epoch": 9.226,
+      "grad_norm": 1.3761494159698486,
+      "learning_rate": 2e-05,
+      "loss": 0.0450317,
+      "step": 4613
+    },
+    {
+      "epoch": 9.228,
+      "grad_norm": 1.8002984523773193,
+      "learning_rate": 2e-05,
+      "loss": 0.0667226,
+      "step": 4614
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 1.532179355621338,
+      "learning_rate": 2e-05,
+      "loss": 0.0425299,
+      "step": 4615
+    },
+    {
+      "epoch": 9.232,
+      "grad_norm": 1.6311419010162354,
+      "learning_rate": 2e-05,
+      "loss": 0.04938279,
+      "step": 4616
+    },
+    {
+      "epoch": 9.234,
+      "grad_norm": 1.4770641326904297,
+      "learning_rate": 2e-05,
+      "loss": 0.06095012,
+      "step": 4617
+    },
+    {
+      "epoch": 9.236,
+      "grad_norm": 1.8100465536117554,
+      "learning_rate": 2e-05,
+      "loss": 0.06146353,
+      "step": 4618
+    },
+    {
+      "epoch": 9.238,
+      "grad_norm": 2.3167026042938232,
+      "learning_rate": 2e-05,
+      "loss": 0.05474924,
+      "step": 4619
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 1.2826061248779297,
+      "learning_rate": 2e-05,
+      "loss": 0.04267112,
+      "step": 4620
+    },
+    {
+      "epoch": 9.242,
+      "grad_norm": 1.6880052089691162,
+      "learning_rate": 2e-05,
+      "loss": 0.05456819,
+      "step": 4621
+    },
+    {
+      "epoch": 9.244,
+      "grad_norm": 1.711029052734375,
+      "learning_rate": 2e-05,
+      "loss": 0.0621344,
+      "step": 4622
+    },
+    {
+      "epoch": 9.246,
+      "grad_norm": 2.999143600463867,
+      "learning_rate": 2e-05,
+      "loss": 0.08942266,
+      "step": 4623
+    },
+    {
+      "epoch": 9.248,
+      "grad_norm": 3.74527645111084,
+      "learning_rate": 2e-05,
+      "loss": 0.05644033,
+      "step": 4624
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 1.9168704748153687,
+      "learning_rate": 2e-05,
+      "loss": 0.03920052,
+      "step": 4625
+    },
+    {
+      "epoch": 9.252,
+      "grad_norm": 1.4764798879623413,
+      "learning_rate": 2e-05,
+      "loss": 0.06123003,
+      "step": 4626
+    },
+    {
+      "epoch": 9.254,
+      "grad_norm": 1.1448794603347778,
+      "learning_rate": 2e-05,
+      "loss": 0.04291178,
+      "step": 4627
+    },
+    {
+      "epoch": 9.256,
+      "grad_norm": 1.2735017538070679,
+      "learning_rate": 2e-05,
+      "loss": 0.0421702,
+      "step": 4628
+    },
+    {
+      "epoch": 9.258,
+      "grad_norm": 1.3639416694641113,
+      "learning_rate": 2e-05,
+      "loss": 0.04493828,
+      "step": 4629
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 1.3818625211715698,
+      "learning_rate": 2e-05,
+      "loss": 0.04133046,
+      "step": 4630
+    },
+    {
+      "epoch": 9.262,
+      "grad_norm": 1.436888575553894,
+      "learning_rate": 2e-05,
+      "loss": 0.05594487,
+      "step": 4631
+    },
+    {
+      "epoch": 9.264,
+      "grad_norm": 1.6045076847076416,
+      "learning_rate": 2e-05,
+      "loss": 0.07328442,
+      "step": 4632
+    },
+    {
+      "epoch": 9.266,
+      "grad_norm": 1.4170843362808228,
+      "learning_rate": 2e-05,
+      "loss": 0.04571041,
+      "step": 4633
+    },
+    {
+      "epoch": 9.268,
+      "grad_norm": 1.7131657600402832,
+      "learning_rate": 2e-05,
+      "loss": 0.06080949,
+      "step": 4634
+    },
+    {
+      "epoch": 9.27,
+      "grad_norm": 1.8091500997543335,
+      "learning_rate": 2e-05,
+      "loss": 0.0525728,
+      "step": 4635
+    },
+    {
+      "epoch": 9.272,
+      "grad_norm": 1.8526278734207153,
+      "learning_rate": 2e-05,
+      "loss": 0.06190364,
+      "step": 4636
+    },
+    {
+      "epoch": 9.274000000000001,
+      "grad_norm": 1.4877046346664429,
+      "learning_rate": 2e-05,
+      "loss": 0.06286225,
+      "step": 4637
+    },
+    {
+      "epoch": 9.276,
+      "grad_norm": 1.772066354751587,
+      "learning_rate": 2e-05,
+      "loss": 0.04873686,
+      "step": 4638
+    },
+    {
+      "epoch": 9.278,
+      "grad_norm": 2.045987606048584,
+      "learning_rate": 2e-05,
+      "loss": 0.06090255,
+      "step": 4639
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 1.1641405820846558,
+      "learning_rate": 2e-05,
+      "loss": 0.03027676,
+      "step": 4640
+    },
+    {
+      "epoch": 9.282,
+      "grad_norm": 1.6448951959609985,
+      "learning_rate": 2e-05,
+      "loss": 0.0530159,
+      "step": 4641
+    },
+    {
+      "epoch": 9.284,
+      "grad_norm": 1.8884637355804443,
+      "learning_rate": 2e-05,
+      "loss": 0.0639272,
+      "step": 4642
+    },
+    {
+      "epoch": 9.286,
+      "grad_norm": 1.355390191078186,
+      "learning_rate": 2e-05,
+      "loss": 0.04228202,
+      "step": 4643
+    },
+    {
+      "epoch": 9.288,
+      "grad_norm": 2.2434418201446533,
+      "learning_rate": 2e-05,
+      "loss": 0.07046621,
+      "step": 4644
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 2.844688892364502,
+      "learning_rate": 2e-05,
+      "loss": 0.08373692,
+      "step": 4645
+    },
+    {
+      "epoch": 9.292,
+      "grad_norm": 1.803123950958252,
+      "learning_rate": 2e-05,
+      "loss": 0.05633461,
+      "step": 4646
+    },
+    {
+      "epoch": 9.294,
+      "grad_norm": 2.8948237895965576,
+      "learning_rate": 2e-05,
+      "loss": 0.05847919,
+      "step": 4647
+    },
+    {
+      "epoch": 9.296,
+      "grad_norm": 2.4830639362335205,
+      "learning_rate": 2e-05,
+      "loss": 0.05488868,
+      "step": 4648
+    },
+    {
+      "epoch": 9.298,
+      "grad_norm": 1.4770337343215942,
+      "learning_rate": 2e-05,
+      "loss": 0.04579071,
+      "step": 4649
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 2.6344194412231445,
+      "learning_rate": 2e-05,
+      "loss": 0.05781864,
+      "step": 4650
+    },
+    {
+      "epoch": 9.302,
+      "grad_norm": 1.8816792964935303,
+      "learning_rate": 2e-05,
+      "loss": 0.05238046,
+      "step": 4651
+    },
+    {
+      "epoch": 9.304,
+      "grad_norm": 1.3099342584609985,
+      "learning_rate": 2e-05,
+      "loss": 0.04636721,
+      "step": 4652
+    },
+    {
+      "epoch": 9.306,
+      "grad_norm": 2.569141149520874,
+      "learning_rate": 2e-05,
+      "loss": 0.05769484,
+      "step": 4653
+    },
+    {
+      "epoch": 9.308,
+      "grad_norm": 1.5715465545654297,
+      "learning_rate": 2e-05,
+      "loss": 0.05822673,
+      "step": 4654
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 2.693352222442627,
+      "learning_rate": 2e-05,
+      "loss": 0.08047836,
+      "step": 4655
+    },
+    {
+      "epoch": 9.312,
+      "grad_norm": 2.2986159324645996,
+      "learning_rate": 2e-05,
+      "loss": 0.06082643,
+      "step": 4656
+    },
+    {
+      "epoch": 9.314,
+      "grad_norm": 1.2851147651672363,
+      "learning_rate": 2e-05,
+      "loss": 0.04421428,
+      "step": 4657
+    },
+    {
+      "epoch": 9.316,
+      "grad_norm": 2.6191282272338867,
+      "learning_rate": 2e-05,
+      "loss": 0.05954355,
+      "step": 4658
+    },
+    {
+      "epoch": 9.318,
+      "grad_norm": 1.7637144327163696,
+      "learning_rate": 2e-05,
+      "loss": 0.06339554,
+      "step": 4659
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 2.729525089263916,
+      "learning_rate": 2e-05,
+      "loss": 0.07117513,
+      "step": 4660
+    },
+    {
+      "epoch": 9.322,
+      "grad_norm": 1.4302338361740112,
+      "learning_rate": 2e-05,
+      "loss": 0.04973509,
+      "step": 4661
+    },
+    {
+      "epoch": 9.324,
+      "grad_norm": 1.4821285009384155,
+      "learning_rate": 2e-05,
+      "loss": 0.05108538,
+      "step": 4662
+    },
+    {
+      "epoch": 9.326,
+      "grad_norm": 1.5328965187072754,
+      "learning_rate": 2e-05,
+      "loss": 0.05789268,
+      "step": 4663
+    },
+    {
+      "epoch": 9.328,
+      "grad_norm": 1.9559437036514282,
+      "learning_rate": 2e-05,
+      "loss": 0.06492808,
+      "step": 4664
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 1.2809813022613525,
+      "learning_rate": 2e-05,
+      "loss": 0.04128128,
+      "step": 4665
+    },
+    {
+      "epoch": 9.332,
+      "grad_norm": 3.450357675552368,
+      "learning_rate": 2e-05,
+      "loss": 0.05096133,
+      "step": 4666
+    },
+    {
+      "epoch": 9.334,
+      "grad_norm": 2.293715238571167,
+      "learning_rate": 2e-05,
+      "loss": 0.03788307,
+      "step": 4667
+    },
+    {
+      "epoch": 9.336,
+      "grad_norm": 1.2826162576675415,
+      "learning_rate": 2e-05,
+      "loss": 0.05124827,
+      "step": 4668
+    },
+    {
+      "epoch": 9.338,
+      "grad_norm": 1.501015305519104,
+      "learning_rate": 2e-05,
+      "loss": 0.04789577,
+      "step": 4669
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 2.1823089122772217,
+      "learning_rate": 2e-05,
+      "loss": 0.06318521,
+      "step": 4670
+    },
+    {
+      "epoch": 9.342,
+      "grad_norm": 1.5042678117752075,
+      "learning_rate": 2e-05,
+      "loss": 0.0364994,
+      "step": 4671
+    },
+    {
+      "epoch": 9.344,
+      "grad_norm": 1.1942931413650513,
+      "learning_rate": 2e-05,
+      "loss": 0.04444024,
+      "step": 4672
+    },
+    {
+      "epoch": 9.346,
+      "grad_norm": 1.6051136255264282,
+      "learning_rate": 2e-05,
+      "loss": 0.06005847,
+      "step": 4673
+    },
+    {
+      "epoch": 9.348,
+      "grad_norm": 1.2597368955612183,
+      "learning_rate": 2e-05,
+      "loss": 0.04411948,
+      "step": 4674
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 1.3099509477615356,
+      "learning_rate": 2e-05,
+      "loss": 0.05771332,
+      "step": 4675
+    },
+    {
+      "epoch": 9.352,
+      "grad_norm": 2.081918954849243,
+      "learning_rate": 2e-05,
+      "loss": 0.04082836,
+      "step": 4676
+    },
+    {
+      "epoch": 9.354,
+      "grad_norm": 1.5217939615249634,
+      "learning_rate": 2e-05,
+      "loss": 0.05541312,
+      "step": 4677
+    },
+    {
+      "epoch": 9.356,
+      "grad_norm": 1.5885696411132812,
+      "learning_rate": 2e-05,
+      "loss": 0.06058114,
+      "step": 4678
+    },
+    {
+      "epoch": 9.358,
+      "grad_norm": 1.2582283020019531,
+      "learning_rate": 2e-05,
+      "loss": 0.02798801,
+      "step": 4679
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 1.9092825651168823,
+      "learning_rate": 2e-05,
+      "loss": 0.06019748,
+      "step": 4680
+    },
+    {
+      "epoch": 9.362,
+      "grad_norm": 3.1406259536743164,
+      "learning_rate": 2e-05,
+      "loss": 0.04014409,
+      "step": 4681
+    },
+    {
+      "epoch": 9.364,
+      "grad_norm": 1.404429316520691,
+      "learning_rate": 2e-05,
+      "loss": 0.04144133,
+      "step": 4682
+    },
+    {
+      "epoch": 9.366,
+      "grad_norm": 3.4135115146636963,
+      "learning_rate": 2e-05,
+      "loss": 0.06071433,
+      "step": 4683
+    },
+    {
+      "epoch": 9.368,
+      "grad_norm": 2.7143144607543945,
+      "learning_rate": 2e-05,
+      "loss": 0.06220663,
+      "step": 4684
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 1.7465615272521973,
+      "learning_rate": 2e-05,
+      "loss": 0.05392168,
+      "step": 4685
+    },
+    {
+      "epoch": 9.372,
+      "grad_norm": 2.8406643867492676,
+      "learning_rate": 2e-05,
+      "loss": 0.09703434,
+      "step": 4686
+    },
+    {
+      "epoch": 9.374,
+      "grad_norm": 1.889456868171692,
+      "learning_rate": 2e-05,
+      "loss": 0.04023182,
+      "step": 4687
+    },
+    {
+      "epoch": 9.376,
+      "grad_norm": 1.769338846206665,
+      "learning_rate": 2e-05,
+      "loss": 0.04535776,
+      "step": 4688
+    },
+    {
+      "epoch": 9.378,
+      "grad_norm": 1.3787952661514282,
+      "learning_rate": 2e-05,
+      "loss": 0.05142323,
+      "step": 4689
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 1.353203296661377,
+      "learning_rate": 2e-05,
+      "loss": 0.04485345,
+      "step": 4690
+    },
+    {
+      "epoch": 9.382,
+      "grad_norm": 1.3988724946975708,
+      "learning_rate": 2e-05,
+      "loss": 0.03410086,
+      "step": 4691
+    },
+    {
+      "epoch": 9.384,
+      "grad_norm": 1.2784860134124756,
+      "learning_rate": 2e-05,
+      "loss": 0.05209333,
+      "step": 4692
+    },
+    {
+      "epoch": 9.386,
+      "grad_norm": 1.9616296291351318,
+      "learning_rate": 2e-05,
+      "loss": 0.0617491,
+      "step": 4693
+    },
+    {
+      "epoch": 9.388,
+      "grad_norm": 2.1018028259277344,
+      "learning_rate": 2e-05,
+      "loss": 0.06020827,
+      "step": 4694
+    },
+    {
+      "epoch": 9.39,
+      "grad_norm": 1.3357975482940674,
+      "learning_rate": 2e-05,
+      "loss": 0.03462115,
+      "step": 4695
+    },
+    {
+      "epoch": 9.392,
+      "grad_norm": 2.2010691165924072,
+      "learning_rate": 2e-05,
+      "loss": 0.05162614,
+      "step": 4696
+    },
+    {
+      "epoch": 9.394,
+      "grad_norm": 1.4281353950500488,
+      "learning_rate": 2e-05,
+      "loss": 0.04901909,
+      "step": 4697
+    },
+    {
+      "epoch": 9.396,
+      "grad_norm": 1.978698492050171,
+      "learning_rate": 2e-05,
+      "loss": 0.06718802,
+      "step": 4698
+    },
+    {
+      "epoch": 9.398,
+      "grad_norm": 1.465959906578064,
+      "learning_rate": 2e-05,
+      "loss": 0.03534799,
+      "step": 4699
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 1.7106237411499023,
+      "learning_rate": 2e-05,
+      "loss": 0.06189927,
+      "step": 4700
+    },
+    {
+      "epoch": 9.402,
+      "grad_norm": 1.349387526512146,
+      "learning_rate": 2e-05,
+      "loss": 0.04897711,
+      "step": 4701
+    },
+    {
+      "epoch": 9.404,
+      "grad_norm": 1.9171662330627441,
+      "learning_rate": 2e-05,
+      "loss": 0.04699158,
+      "step": 4702
+    },
+    {
+      "epoch": 9.406,
+      "grad_norm": 2.517526388168335,
+      "learning_rate": 2e-05,
+      "loss": 0.07169571,
+      "step": 4703
+    },
+    {
+      "epoch": 9.408,
+      "grad_norm": 1.851466417312622,
+      "learning_rate": 2e-05,
+      "loss": 0.04784168,
+      "step": 4704
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 2.378621816635132,
+      "learning_rate": 2e-05,
+      "loss": 0.04365069,
+      "step": 4705
+    },
+    {
+      "epoch": 9.412,
+      "grad_norm": 1.615919828414917,
+      "learning_rate": 2e-05,
+      "loss": 0.05020372,
+      "step": 4706
+    },
+    {
+      "epoch": 9.414,
+      "grad_norm": 1.691772222518921,
+      "learning_rate": 2e-05,
+      "loss": 0.04145588,
+      "step": 4707
+    },
+    {
+      "epoch": 9.416,
+      "grad_norm": 3.0600521564483643,
+      "learning_rate": 2e-05,
+      "loss": 0.04704102,
+      "step": 4708
+    },
+    {
+      "epoch": 9.418,
+      "grad_norm": 1.2809059619903564,
+      "learning_rate": 2e-05,
+      "loss": 0.04573686,
+      "step": 4709
+    },
+    {
+      "epoch": 9.42,
+      "grad_norm": 1.2681528329849243,
+      "learning_rate": 2e-05,
+      "loss": 0.04140832,
+      "step": 4710
+    },
+    {
+      "epoch": 9.422,
+      "grad_norm": 1.5011893510818481,
+      "learning_rate": 2e-05,
+      "loss": 0.04901879,
+      "step": 4711
+    },
+    {
+      "epoch": 9.424,
+      "grad_norm": 1.7330307960510254,
+      "learning_rate": 2e-05,
+      "loss": 0.05393284,
+      "step": 4712
+    },
+    {
+      "epoch": 9.426,
+      "grad_norm": 1.1501388549804688,
+      "learning_rate": 2e-05,
+      "loss": 0.04429302,
+      "step": 4713
+    },
+    {
+      "epoch": 9.428,
+      "grad_norm": 1.5243293046951294,
+      "learning_rate": 2e-05,
+      "loss": 0.06208751,
+      "step": 4714
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 1.6041468381881714,
+      "learning_rate": 2e-05,
+      "loss": 0.05365137,
+      "step": 4715
+    },
+    {
+      "epoch": 9.432,
+      "grad_norm": 1.411880612373352,
+      "learning_rate": 2e-05,
+      "loss": 0.04755591,
+      "step": 4716
+    },
+    {
+      "epoch": 9.434,
+      "grad_norm": 1.7034417390823364,
+      "learning_rate": 2e-05,
+      "loss": 0.05169332,
+      "step": 4717
+    },
+    {
+      "epoch": 9.436,
+      "grad_norm": 1.398370623588562,
+      "learning_rate": 2e-05,
+      "loss": 0.04505457,
+      "step": 4718
+    },
+    {
+      "epoch": 9.438,
+      "grad_norm": 1.618485927581787,
+      "learning_rate": 2e-05,
+      "loss": 0.05048877,
+      "step": 4719
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 2.2976009845733643,
+      "learning_rate": 2e-05,
+      "loss": 0.05732608,
+      "step": 4720
+    },
+    {
+      "epoch": 9.442,
+      "grad_norm": 1.4977824687957764,
+      "learning_rate": 2e-05,
+      "loss": 0.05055287,
+      "step": 4721
+    },
+    {
+      "epoch": 9.444,
+      "grad_norm": 2.0112006664276123,
+      "learning_rate": 2e-05,
+      "loss": 0.05900105,
+      "step": 4722
+    },
+    {
+      "epoch": 9.446,
+      "grad_norm": 1.2168583869934082,
+      "learning_rate": 2e-05,
+      "loss": 0.04659557,
+      "step": 4723
+    },
+    {
+      "epoch": 9.448,
+      "grad_norm": 1.8147743940353394,
+      "learning_rate": 2e-05,
+      "loss": 0.04836474,
+      "step": 4724
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 2.1160666942596436,
+      "learning_rate": 2e-05,
+      "loss": 0.06309254,
+      "step": 4725
+    },
+    {
+      "epoch": 9.452,
+      "grad_norm": 1.0232397317886353,
+      "learning_rate": 2e-05,
+      "loss": 0.03620723,
+      "step": 4726
+    },
+    {
+      "epoch": 9.454,
+      "grad_norm": 1.4367437362670898,
+      "learning_rate": 2e-05,
+      "loss": 0.04998949,
+      "step": 4727
+    },
+    {
+      "epoch": 9.456,
+      "grad_norm": 1.9640501737594604,
+      "learning_rate": 2e-05,
+      "loss": 0.07069734,
+      "step": 4728
+    },
+    {
+      "epoch": 9.458,
+      "grad_norm": 4.114424228668213,
+      "learning_rate": 2e-05,
+      "loss": 0.09708965,
+      "step": 4729
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 1.8628021478652954,
+      "learning_rate": 2e-05,
+      "loss": 0.0480114,
+      "step": 4730
+    },
+    {
+      "epoch": 9.462,
+      "grad_norm": 1.896959662437439,
+      "learning_rate": 2e-05,
+      "loss": 0.057604,
+      "step": 4731
+    },
+    {
+      "epoch": 9.464,
+      "grad_norm": 1.6538161039352417,
+      "learning_rate": 2e-05,
+      "loss": 0.05080595,
+      "step": 4732
+    },
+    {
+      "epoch": 9.466,
+      "grad_norm": 1.931830883026123,
+      "learning_rate": 2e-05,
+      "loss": 0.053108,
+      "step": 4733
+    },
+    {
+      "epoch": 9.468,
+      "grad_norm": 1.452933430671692,
+      "learning_rate": 2e-05,
+      "loss": 0.041306,
+      "step": 4734
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 1.8941861391067505,
+      "learning_rate": 2e-05,
+      "loss": 0.05748565,
+      "step": 4735
+    },
+    {
+      "epoch": 9.472,
+      "grad_norm": 1.3065882921218872,
+      "learning_rate": 2e-05,
+      "loss": 0.05725362,
+      "step": 4736
+    },
+    {
+      "epoch": 9.474,
+      "grad_norm": 1.3110909461975098,
+      "learning_rate": 2e-05,
+      "loss": 0.04384652,
+      "step": 4737
+    },
+    {
+      "epoch": 9.475999999999999,
+      "grad_norm": 1.1904391050338745,
+      "learning_rate": 2e-05,
+      "loss": 0.0409494,
+      "step": 4738
+    },
+    {
+      "epoch": 9.478,
+      "grad_norm": 1.359976053237915,
+      "learning_rate": 2e-05,
+      "loss": 0.05123883,
+      "step": 4739
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 1.628815770149231,
+      "learning_rate": 2e-05,
+      "loss": 0.0685708,
+      "step": 4740
+    },
+    {
+      "epoch": 9.482,
+      "grad_norm": 1.5649138689041138,
+      "learning_rate": 2e-05,
+      "loss": 0.0603338,
+      "step": 4741
+    },
+    {
+      "epoch": 9.484,
+      "grad_norm": 1.9411360025405884,
+      "learning_rate": 2e-05,
+      "loss": 0.05299631,
+      "step": 4742
+    },
+    {
+      "epoch": 9.486,
+      "grad_norm": 2.5453522205352783,
+      "learning_rate": 2e-05,
+      "loss": 0.04924603,
+      "step": 4743
+    },
+    {
+      "epoch": 9.488,
+      "grad_norm": 1.5162389278411865,
+      "learning_rate": 2e-05,
+      "loss": 0.04518431,
+      "step": 4744
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 1.0650982856750488,
+      "learning_rate": 2e-05,
+      "loss": 0.03788474,
+      "step": 4745
+    },
+    {
+      "epoch": 9.492,
+      "grad_norm": 1.4776169061660767,
+      "learning_rate": 2e-05,
+      "loss": 0.04502506,
+      "step": 4746
+    },
+    {
+      "epoch": 9.494,
+      "grad_norm": 4.540150165557861,
+      "learning_rate": 2e-05,
+      "loss": 0.06305524,
+      "step": 4747
+    },
+    {
+      "epoch": 9.496,
+      "grad_norm": 1.8420084714889526,
+      "learning_rate": 2e-05,
+      "loss": 0.05973725,
+      "step": 4748
+    },
+    {
+      "epoch": 9.498,
+      "grad_norm": 1.177133321762085,
+      "learning_rate": 2e-05,
+      "loss": 0.04210076,
+      "step": 4749
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 1.4064905643463135,
+      "learning_rate": 2e-05,
+      "loss": 0.04571437,
+      "step": 4750
+    },
+    {
+      "epoch": 9.502,
+      "grad_norm": 1.7198388576507568,
+      "learning_rate": 2e-05,
+      "loss": 0.04831848,
+      "step": 4751
+    },
+    {
+      "epoch": 9.504,
+      "grad_norm": 1.373218297958374,
+      "learning_rate": 2e-05,
+      "loss": 0.04200636,
+      "step": 4752
+    },
+    {
+      "epoch": 9.506,
+      "grad_norm": 2.0869367122650146,
+      "learning_rate": 2e-05,
+      "loss": 0.05860278,
+      "step": 4753
+    },
+    {
+      "epoch": 9.508,
+      "grad_norm": 2.03802227973938,
+      "learning_rate": 2e-05,
+      "loss": 0.04599741,
+      "step": 4754
+    },
+    {
+      "epoch": 9.51,
+      "grad_norm": 2.718126058578491,
+      "learning_rate": 2e-05,
+      "loss": 0.05679199,
+      "step": 4755
+    },
+    {
+      "epoch": 9.512,
+      "grad_norm": 2.1345672607421875,
+      "learning_rate": 2e-05,
+      "loss": 0.07050245,
+      "step": 4756
+    },
+    {
+      "epoch": 9.514,
+      "grad_norm": 3.7088184356689453,
+      "learning_rate": 2e-05,
+      "loss": 0.07732576,
+      "step": 4757
+    },
+    {
+      "epoch": 9.516,
+      "grad_norm": 1.6699403524398804,
+      "learning_rate": 2e-05,
+      "loss": 0.04383042,
+      "step": 4758
+    },
+    {
+      "epoch": 9.518,
+      "grad_norm": 1.5457981824874878,
+      "learning_rate": 2e-05,
+      "loss": 0.03807002,
+      "step": 4759
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 1.9819368124008179,
+      "learning_rate": 2e-05,
+      "loss": 0.07731517,
+      "step": 4760
+    },
+    {
+      "epoch": 9.522,
+      "grad_norm": 2.0579514503479004,
+      "learning_rate": 2e-05,
+      "loss": 0.06062069,
+      "step": 4761
+    },
+    {
+      "epoch": 9.524000000000001,
+      "grad_norm": 1.8381377458572388,
+      "learning_rate": 2e-05,
+      "loss": 0.04691198,
+      "step": 4762
+    },
+    {
+      "epoch": 9.526,
+      "grad_norm": 2.709359645843506,
+      "learning_rate": 2e-05,
+      "loss": 0.05731576,
+      "step": 4763
+    },
+    {
+      "epoch": 9.528,
+      "grad_norm": 1.4428166151046753,
+      "learning_rate": 2e-05,
+      "loss": 0.04605105,
+      "step": 4764
+    },
+    {
+      "epoch": 9.53,
+      "grad_norm": 1.5184147357940674,
+      "learning_rate": 2e-05,
+      "loss": 0.05696104,
+      "step": 4765
+    },
+    {
+      "epoch": 9.532,
+      "grad_norm": 1.3034058809280396,
+      "learning_rate": 2e-05,
+      "loss": 0.03874236,
+      "step": 4766
+    },
+    {
+      "epoch": 9.534,
+      "grad_norm": 1.7730201482772827,
+      "learning_rate": 2e-05,
+      "loss": 0.05022582,
+      "step": 4767
+    },
+    {
+      "epoch": 9.536,
+      "grad_norm": 1.5613415241241455,
+      "learning_rate": 2e-05,
+      "loss": 0.06682578,
+      "step": 4768
+    },
+    {
+      "epoch": 9.538,
+      "grad_norm": 1.9399430751800537,
+      "learning_rate": 2e-05,
+      "loss": 0.05712492,
+      "step": 4769
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 2.3643603324890137,
+      "learning_rate": 2e-05,
+      "loss": 0.0664971,
+      "step": 4770
+    },
+    {
+      "epoch": 9.542,
+      "grad_norm": 1.1537812948226929,
+      "learning_rate": 2e-05,
+      "loss": 0.04907678,
+      "step": 4771
+    },
+    {
+      "epoch": 9.544,
+      "grad_norm": 1.476564884185791,
+      "learning_rate": 2e-05,
+      "loss": 0.05081899,
+      "step": 4772
+    },
+    {
+      "epoch": 9.546,
+      "grad_norm": 1.345429539680481,
+      "learning_rate": 2e-05,
+      "loss": 0.0485204,
+      "step": 4773
+    },
+    {
+      "epoch": 9.548,
+      "grad_norm": 1.3945995569229126,
+      "learning_rate": 2e-05,
+      "loss": 0.05414914,
+      "step": 4774
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 1.2747801542282104,
+      "learning_rate": 2e-05,
+      "loss": 0.04904943,
+      "step": 4775
+    },
+    {
+      "epoch": 9.552,
+      "grad_norm": 1.364349603652954,
+      "learning_rate": 2e-05,
+      "loss": 0.05541695,
+      "step": 4776
+    },
+    {
+      "epoch": 9.554,
+      "grad_norm": 1.40326726436615,
+      "learning_rate": 2e-05,
+      "loss": 0.0485206,
+      "step": 4777
+    },
+    {
+      "epoch": 9.556000000000001,
+      "grad_norm": 2.4218220710754395,
+      "learning_rate": 2e-05,
+      "loss": 0.05936872,
+      "step": 4778
+    },
+    {
+      "epoch": 9.558,
+      "grad_norm": 1.5374068021774292,
+      "learning_rate": 2e-05,
+      "loss": 0.03438132,
+      "step": 4779
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 2.0976953506469727,
+      "learning_rate": 2e-05,
+      "loss": 0.04670816,
+      "step": 4780
+    },
+    {
+      "epoch": 9.562,
+      "grad_norm": 1.6651930809020996,
+      "learning_rate": 2e-05,
+      "loss": 0.04262748,
+      "step": 4781
+    },
+    {
+      "epoch": 9.564,
+      "grad_norm": 1.433442234992981,
+      "learning_rate": 2e-05,
+      "loss": 0.05769576,
+      "step": 4782
+    },
+    {
+      "epoch": 9.566,
+      "grad_norm": 2.727137565612793,
+      "learning_rate": 2e-05,
+      "loss": 0.05369073,
+      "step": 4783
+    },
+    {
+      "epoch": 9.568,
+      "grad_norm": 2.403437852859497,
+      "learning_rate": 2e-05,
+      "loss": 0.05518568,
+      "step": 4784
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 1.5393919944763184,
+      "learning_rate": 2e-05,
+      "loss": 0.05311519,
+      "step": 4785
+    },
+    {
+      "epoch": 9.572,
+      "grad_norm": 1.650131344795227,
+      "learning_rate": 2e-05,
+      "loss": 0.05369199,
+      "step": 4786
+    },
+    {
+      "epoch": 9.574,
+      "grad_norm": 1.6887848377227783,
+      "learning_rate": 2e-05,
+      "loss": 0.04836053,
+      "step": 4787
+    },
+    {
+      "epoch": 9.576,
+      "grad_norm": 1.3404792547225952,
+      "learning_rate": 2e-05,
+      "loss": 0.05586416,
+      "step": 4788
+    },
+    {
+      "epoch": 9.578,
+      "grad_norm": 1.5387517213821411,
+      "learning_rate": 2e-05,
+      "loss": 0.05821295,
+      "step": 4789
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 1.3446617126464844,
+      "learning_rate": 2e-05,
+      "loss": 0.04860961,
+      "step": 4790
+    },
+    {
+      "epoch": 9.582,
+      "grad_norm": 1.9128057956695557,
+      "learning_rate": 2e-05,
+      "loss": 0.06139975,
+      "step": 4791
+    },
+    {
+      "epoch": 9.584,
+      "grad_norm": 1.8623543977737427,
+      "learning_rate": 2e-05,
+      "loss": 0.04838707,
+      "step": 4792
+    },
+    {
+      "epoch": 9.586,
+      "grad_norm": 2.0884079933166504,
+      "learning_rate": 2e-05,
+      "loss": 0.05497058,
+      "step": 4793
+    },
+    {
+      "epoch": 9.588,
+      "grad_norm": 1.26129949092865,
+      "learning_rate": 2e-05,
+      "loss": 0.05032674,
+      "step": 4794
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 1.5007047653198242,
+      "learning_rate": 2e-05,
+      "loss": 0.0500331,
+      "step": 4795
+    },
+    {
+      "epoch": 9.592,
+      "grad_norm": 2.1904873847961426,
+      "learning_rate": 2e-05,
+      "loss": 0.05238812,
+      "step": 4796
+    },
+    {
+      "epoch": 9.594,
+      "grad_norm": 2.0764124393463135,
+      "learning_rate": 2e-05,
+      "loss": 0.05428065,
+      "step": 4797
+    },
+    {
+      "epoch": 9.596,
+      "grad_norm": 1.8335858583450317,
+      "learning_rate": 2e-05,
+      "loss": 0.05815455,
+      "step": 4798
+    },
+    {
+      "epoch": 9.598,
+      "grad_norm": 1.186977744102478,
+      "learning_rate": 2e-05,
+      "loss": 0.04371549,
+      "step": 4799
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 6.732235908508301,
+      "learning_rate": 2e-05,
+      "loss": 0.06942256,
+      "step": 4800
+    },
+    {
+      "epoch": 9.602,
+      "grad_norm": 2.0716724395751953,
+      "learning_rate": 2e-05,
+      "loss": 0.0744697,
+      "step": 4801
+    },
+    {
+      "epoch": 9.604,
+      "grad_norm": 1.558214783668518,
+      "learning_rate": 2e-05,
+      "loss": 0.06264573,
+      "step": 4802
+    },
+    {
+      "epoch": 9.606,
+      "grad_norm": 4.206948280334473,
+      "learning_rate": 2e-05,
+      "loss": 0.055962,
+      "step": 4803
+    },
+    {
+      "epoch": 9.608,
+      "grad_norm": 1.3140082359313965,
+      "learning_rate": 2e-05,
+      "loss": 0.04739345,
+      "step": 4804
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 1.58623468875885,
+      "learning_rate": 2e-05,
+      "loss": 0.0550376,
+      "step": 4805
+    },
+    {
+      "epoch": 9.612,
+      "grad_norm": 1.9726959466934204,
+      "learning_rate": 2e-05,
+      "loss": 0.07566231,
+      "step": 4806
+    },
+    {
+      "epoch": 9.614,
+      "grad_norm": 1.6119412183761597,
+      "learning_rate": 2e-05,
+      "loss": 0.05679318,
+      "step": 4807
+    },
+    {
+      "epoch": 9.616,
+      "grad_norm": 1.3407459259033203,
+      "learning_rate": 2e-05,
+      "loss": 0.04322591,
+      "step": 4808
+    },
+    {
+      "epoch": 9.618,
+      "grad_norm": 1.3803037405014038,
+      "learning_rate": 2e-05,
+      "loss": 0.04514508,
+      "step": 4809
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 1.5106648206710815,
+      "learning_rate": 2e-05,
+      "loss": 0.06446012,
+      "step": 4810
+    },
+    {
+      "epoch": 9.622,
+      "grad_norm": 1.2000592947006226,
+      "learning_rate": 2e-05,
+      "loss": 0.04472655,
+      "step": 4811
+    },
+    {
+      "epoch": 9.624,
+      "grad_norm": 3.254822015762329,
+      "learning_rate": 2e-05,
+      "loss": 0.07974194,
+      "step": 4812
+    },
+    {
+      "epoch": 9.626,
+      "grad_norm": 1.7189629077911377,
+      "learning_rate": 2e-05,
+      "loss": 0.05568657,
+      "step": 4813
+    },
+    {
+      "epoch": 9.628,
+      "grad_norm": 1.4042153358459473,
+      "learning_rate": 2e-05,
+      "loss": 0.04132761,
+      "step": 4814
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 2.279066562652588,
+      "learning_rate": 2e-05,
+      "loss": 0.04985248,
+      "step": 4815
+    },
+    {
+      "epoch": 9.632,
+      "grad_norm": 1.4138849973678589,
+      "learning_rate": 2e-05,
+      "loss": 0.04461198,
+      "step": 4816
+    },
+    {
+      "epoch": 9.634,
+      "grad_norm": 2.0648868083953857,
+      "learning_rate": 2e-05,
+      "loss": 0.04070833,
+      "step": 4817
+    },
+    {
+      "epoch": 9.636,
+      "grad_norm": 3.6020028591156006,
+      "learning_rate": 2e-05,
+      "loss": 0.05208477,
+      "step": 4818
+    },
+    {
+      "epoch": 9.638,
+      "grad_norm": 1.8781304359436035,
+      "learning_rate": 2e-05,
+      "loss": 0.05021086,
+      "step": 4819
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 1.0614352226257324,
+      "learning_rate": 2e-05,
+      "loss": 0.03605043,
+      "step": 4820
+    },
+    {
+      "epoch": 9.642,
+      "grad_norm": 1.910072684288025,
+      "learning_rate": 2e-05,
+      "loss": 0.06666766,
+      "step": 4821
+    },
+    {
+      "epoch": 9.644,
+      "grad_norm": 1.2410168647766113,
+      "learning_rate": 2e-05,
+      "loss": 0.04226237,
+      "step": 4822
+    },
+    {
+      "epoch": 9.646,
+      "grad_norm": 1.6553242206573486,
+      "learning_rate": 2e-05,
+      "loss": 0.05368885,
+      "step": 4823
+    },
+    {
+      "epoch": 9.648,
+      "grad_norm": 1.3068299293518066,
+      "learning_rate": 2e-05,
+      "loss": 0.03962949,
+      "step": 4824
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 2.591618537902832,
+      "learning_rate": 2e-05,
+      "loss": 0.06662943,
+      "step": 4825
+    },
+    {
+      "epoch": 9.652,
+      "grad_norm": 2.577542781829834,
+      "learning_rate": 2e-05,
+      "loss": 0.06383407,
+      "step": 4826
+    },
+    {
+      "epoch": 9.654,
+      "grad_norm": 1.6264952421188354,
+      "learning_rate": 2e-05,
+      "loss": 0.0523547,
+      "step": 4827
+    },
+    {
+      "epoch": 9.656,
+      "grad_norm": 2.4551913738250732,
+      "learning_rate": 2e-05,
+      "loss": 0.04400642,
+      "step": 4828
+    },
+    {
+      "epoch": 9.658,
+      "grad_norm": 3.1886560916900635,
+      "learning_rate": 2e-05,
+      "loss": 0.07130433,
+      "step": 4829
+    },
+    {
+      "epoch": 9.66,
+      "grad_norm": 1.5255032777786255,
+      "learning_rate": 2e-05,
+      "loss": 0.04827879,
+      "step": 4830
+    },
+    {
+      "epoch": 9.662,
+      "grad_norm": 2.4944541454315186,
+      "learning_rate": 2e-05,
+      "loss": 0.07604385,
+      "step": 4831
+    },
+    {
+      "epoch": 9.664,
+      "grad_norm": 1.9444361925125122,
+      "learning_rate": 2e-05,
+      "loss": 0.04895295,
+      "step": 4832
+    },
+    {
+      "epoch": 9.666,
+      "grad_norm": 1.206829309463501,
+      "learning_rate": 2e-05,
+      "loss": 0.04027706,
+      "step": 4833
+    },
+    {
+      "epoch": 9.668,
+      "grad_norm": 1.0655403137207031,
+      "learning_rate": 2e-05,
+      "loss": 0.03361625,
+      "step": 4834
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 1.6680309772491455,
+      "learning_rate": 2e-05,
+      "loss": 0.05498701,
+      "step": 4835
+    },
+    {
+      "epoch": 9.672,
+      "grad_norm": 1.6686928272247314,
+      "learning_rate": 2e-05,
+      "loss": 0.05405208,
+      "step": 4836
+    },
+    {
+      "epoch": 9.674,
+      "grad_norm": 1.5973371267318726,
+      "learning_rate": 2e-05,
+      "loss": 0.03787687,
+      "step": 4837
+    },
+    {
+      "epoch": 9.676,
+      "grad_norm": 1.5057041645050049,
+      "learning_rate": 2e-05,
+      "loss": 0.05292775,
+      "step": 4838
+    },
+    {
+      "epoch": 9.678,
+      "grad_norm": 2.048788547515869,
+      "learning_rate": 2e-05,
+      "loss": 0.05188146,
+      "step": 4839
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 1.5165146589279175,
+      "learning_rate": 2e-05,
+      "loss": 0.05076149,
+      "step": 4840
+    },
+    {
+      "epoch": 9.682,
+      "grad_norm": 1.3067723512649536,
+      "learning_rate": 2e-05,
+      "loss": 0.04745375,
+      "step": 4841
+    },
+    {
+      "epoch": 9.684,
+      "grad_norm": 2.40364670753479,
+      "learning_rate": 2e-05,
+      "loss": 0.06933152,
+      "step": 4842
+    },
+    {
+      "epoch": 9.686,
+      "grad_norm": 2.187824010848999,
+      "learning_rate": 2e-05,
+      "loss": 0.06234965,
+      "step": 4843
+    },
+    {
+      "epoch": 9.688,
+      "grad_norm": 2.3751494884490967,
+      "learning_rate": 2e-05,
+      "loss": 0.05942409,
+      "step": 4844
+    },
+    {
+      "epoch": 9.69,
+      "grad_norm": 1.3556272983551025,
+      "learning_rate": 2e-05,
+      "loss": 0.0271932,
+      "step": 4845
+    },
+    {
+      "epoch": 9.692,
+      "grad_norm": 2.0702574253082275,
+      "learning_rate": 2e-05,
+      "loss": 0.04416346,
+      "step": 4846
+    },
+    {
+      "epoch": 9.693999999999999,
+      "grad_norm": 2.776580572128296,
+      "learning_rate": 2e-05,
+      "loss": 0.06459181,
+      "step": 4847
+    },
+    {
+      "epoch": 9.696,
+      "grad_norm": 1.423103928565979,
+      "learning_rate": 2e-05,
+      "loss": 0.04712578,
+      "step": 4848
+    },
+    {
+      "epoch": 9.698,
+      "grad_norm": 1.3185054063796997,
+      "learning_rate": 2e-05,
+      "loss": 0.03745867,
+      "step": 4849
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 1.4054712057113647,
+      "learning_rate": 2e-05,
+      "loss": 0.05372046,
+      "step": 4850
+    },
+    {
+      "epoch": 9.702,
+      "grad_norm": 1.5396513938903809,
+      "learning_rate": 2e-05,
+      "loss": 0.04799499,
+      "step": 4851
+    },
+    {
+      "epoch": 9.704,
+      "grad_norm": 1.3242496252059937,
+      "learning_rate": 2e-05,
+      "loss": 0.04319259,
+      "step": 4852
+    },
+    {
+      "epoch": 9.706,
+      "grad_norm": 1.6607216596603394,
+      "learning_rate": 2e-05,
+      "loss": 0.04303396,
+      "step": 4853
+    },
+    {
+      "epoch": 9.708,
+      "grad_norm": 1.570757508277893,
+      "learning_rate": 2e-05,
+      "loss": 0.06974973,
+      "step": 4854
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 1.5978914499282837,
+      "learning_rate": 2e-05,
+      "loss": 0.05531528,
+      "step": 4855
+    },
+    {
+      "epoch": 9.712,
+      "grad_norm": 2.2332658767700195,
+      "learning_rate": 2e-05,
+      "loss": 0.05066529,
+      "step": 4856
+    },
+    {
+      "epoch": 9.714,
+      "grad_norm": 1.695459246635437,
+      "learning_rate": 2e-05,
+      "loss": 0.04276437,
+      "step": 4857
+    },
+    {
+      "epoch": 9.716,
+      "grad_norm": 1.4969645738601685,
+      "learning_rate": 2e-05,
+      "loss": 0.0565182,
+      "step": 4858
+    },
+    {
+      "epoch": 9.718,
+      "grad_norm": 2.46620512008667,
+      "learning_rate": 2e-05,
+      "loss": 0.04657774,
+      "step": 4859
+    },
+    {
+      "epoch": 9.72,
+      "grad_norm": 2.0684173107147217,
+      "learning_rate": 2e-05,
+      "loss": 0.05510051,
+      "step": 4860
+    },
+    {
+      "epoch": 9.722,
+      "grad_norm": 1.5146620273590088,
+      "learning_rate": 2e-05,
+      "loss": 0.04686711,
+      "step": 4861
+    },
+    {
+      "epoch": 9.724,
+      "grad_norm": 2.6633617877960205,
+      "learning_rate": 2e-05,
+      "loss": 0.05086003,
+      "step": 4862
+    },
+    {
+      "epoch": 9.725999999999999,
+      "grad_norm": 1.2176555395126343,
+      "learning_rate": 2e-05,
+      "loss": 0.0497752,
+      "step": 4863
+    },
+    {
+      "epoch": 9.728,
+      "grad_norm": 1.2809525728225708,
+      "learning_rate": 2e-05,
+      "loss": 0.04838467,
+      "step": 4864
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 1.3932509422302246,
+      "learning_rate": 2e-05,
+      "loss": 0.0427762,
+      "step": 4865
+    },
+    {
+      "epoch": 9.732,
+      "grad_norm": 3.854522466659546,
+      "learning_rate": 2e-05,
+      "loss": 0.0666661,
+      "step": 4866
+    },
+    {
+      "epoch": 9.734,
+      "grad_norm": 1.537145972251892,
+      "learning_rate": 2e-05,
+      "loss": 0.03788938,
+      "step": 4867
+    },
+    {
+      "epoch": 9.736,
+      "grad_norm": 2.0760679244995117,
+      "learning_rate": 2e-05,
+      "loss": 0.05522866,
+      "step": 4868
+    },
+    {
+      "epoch": 9.738,
+      "grad_norm": 1.693951964378357,
+      "learning_rate": 2e-05,
+      "loss": 0.06659015,
+      "step": 4869
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 1.2150261402130127,
+      "learning_rate": 2e-05,
+      "loss": 0.04565568,
+      "step": 4870
+    },
+    {
+      "epoch": 9.742,
+      "grad_norm": 1.8635746240615845,
+      "learning_rate": 2e-05,
+      "loss": 0.0635218,
+      "step": 4871
+    },
+    {
+      "epoch": 9.744,
+      "grad_norm": 1.8649309873580933,
+      "learning_rate": 2e-05,
+      "loss": 0.06209478,
+      "step": 4872
+    },
+    {
+      "epoch": 9.746,
+      "grad_norm": 1.2454828023910522,
+      "learning_rate": 2e-05,
+      "loss": 0.04050411,
+      "step": 4873
+    },
+    {
+      "epoch": 9.748,
+      "grad_norm": 1.7118374109268188,
+      "learning_rate": 2e-05,
+      "loss": 0.0501698,
+      "step": 4874
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 1.9825464487075806,
+      "learning_rate": 2e-05,
+      "loss": 0.0551829,
+      "step": 4875
+    },
+    {
+      "epoch": 9.752,
+      "grad_norm": 1.6307551860809326,
+      "learning_rate": 2e-05,
+      "loss": 0.06142005,
+      "step": 4876
+    },
+    {
+      "epoch": 9.754,
+      "grad_norm": 2.2640645503997803,
+      "learning_rate": 2e-05,
+      "loss": 0.06031884,
+      "step": 4877
+    },
+    {
+      "epoch": 9.756,
+      "grad_norm": 1.0484297275543213,
+      "learning_rate": 2e-05,
+      "loss": 0.05024759,
+      "step": 4878
+    },
+    {
+      "epoch": 9.758,
+      "grad_norm": 1.469162940979004,
+      "learning_rate": 2e-05,
+      "loss": 0.04584752,
+      "step": 4879
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 1.5308761596679688,
+      "learning_rate": 2e-05,
+      "loss": 0.03732128,
+      "step": 4880
+    },
+    {
+      "epoch": 9.762,
+      "grad_norm": 1.904474139213562,
+      "learning_rate": 2e-05,
+      "loss": 0.05384236,
+      "step": 4881
+    },
+    {
+      "epoch": 9.764,
+      "grad_norm": 6.474231243133545,
+      "learning_rate": 2e-05,
+      "loss": 0.07180381,
+      "step": 4882
+    },
+    {
+      "epoch": 9.766,
+      "grad_norm": 1.5289386510849,
+      "learning_rate": 2e-05,
+      "loss": 0.05518559,
+      "step": 4883
+    },
+    {
+      "epoch": 9.768,
+      "grad_norm": 1.3522764444351196,
+      "learning_rate": 2e-05,
+      "loss": 0.05258585,
+      "step": 4884
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 1.4351115226745605,
+      "learning_rate": 2e-05,
+      "loss": 0.03858275,
+      "step": 4885
+    },
+    {
+      "epoch": 9.772,
+      "grad_norm": 2.1494507789611816,
+      "learning_rate": 2e-05,
+      "loss": 0.05100446,
+      "step": 4886
+    },
+    {
+      "epoch": 9.774000000000001,
+      "grad_norm": 2.7354249954223633,
+      "learning_rate": 2e-05,
+      "loss": 0.05425896,
+      "step": 4887
+    },
+    {
+      "epoch": 9.776,
+      "grad_norm": 1.6631423234939575,
+      "learning_rate": 2e-05,
+      "loss": 0.04931759,
+      "step": 4888
+    },
+    {
+      "epoch": 9.778,
+      "grad_norm": 2.6517434120178223,
+      "learning_rate": 2e-05,
+      "loss": 0.06315567,
+      "step": 4889
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 2.4273080825805664,
+      "learning_rate": 2e-05,
+      "loss": 0.07165055,
+      "step": 4890
+    },
+    {
+      "epoch": 9.782,
+      "grad_norm": 1.5773106813430786,
+      "learning_rate": 2e-05,
+      "loss": 0.06337696,
+      "step": 4891
+    },
+    {
+      "epoch": 9.784,
+      "grad_norm": 1.3168998956680298,
+      "learning_rate": 2e-05,
+      "loss": 0.04947937,
+      "step": 4892
+    },
+    {
+      "epoch": 9.786,
+      "grad_norm": 1.380509614944458,
+      "learning_rate": 2e-05,
+      "loss": 0.04712594,
+      "step": 4893
+    },
+    {
+      "epoch": 9.788,
+      "grad_norm": 1.479285717010498,
+      "learning_rate": 2e-05,
+      "loss": 0.04698952,
+      "step": 4894
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 1.8223832845687866,
+      "learning_rate": 2e-05,
+      "loss": 0.04973854,
+      "step": 4895
+    },
+    {
+      "epoch": 9.792,
+      "grad_norm": 1.8875789642333984,
+      "learning_rate": 2e-05,
+      "loss": 0.05310801,
+      "step": 4896
+    },
+    {
+      "epoch": 9.794,
+      "grad_norm": 2.261950969696045,
+      "learning_rate": 2e-05,
+      "loss": 0.05974211,
+      "step": 4897
+    },
+    {
+      "epoch": 9.796,
+      "grad_norm": 1.628933072090149,
+      "learning_rate": 2e-05,
+      "loss": 0.04541996,
+      "step": 4898
+    },
+    {
+      "epoch": 9.798,
+      "grad_norm": 2.749540328979492,
+      "learning_rate": 2e-05,
+      "loss": 0.06158432,
+      "step": 4899
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 2.566983938217163,
+      "learning_rate": 2e-05,
+      "loss": 0.05563569,
+      "step": 4900
+    },
+    {
+      "epoch": 9.802,
+      "grad_norm": 1.902557134628296,
+      "learning_rate": 2e-05,
+      "loss": 0.04794583,
+      "step": 4901
+    },
+    {
+      "epoch": 9.804,
+      "grad_norm": 3.0628104209899902,
+      "learning_rate": 2e-05,
+      "loss": 0.06039095,
+      "step": 4902
+    },
+    {
+      "epoch": 9.806000000000001,
+      "grad_norm": 1.1863480806350708,
+      "learning_rate": 2e-05,
+      "loss": 0.03568749,
+      "step": 4903
+    },
+    {
+      "epoch": 9.808,
+      "grad_norm": 2.010585308074951,
+      "learning_rate": 2e-05,
+      "loss": 0.04390045,
+      "step": 4904
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 1.8428614139556885,
+      "learning_rate": 2e-05,
+      "loss": 0.06692166,
+      "step": 4905
+    },
+    {
+      "epoch": 9.812,
+      "grad_norm": 1.32927668094635,
+      "learning_rate": 2e-05,
+      "loss": 0.04278745,
+      "step": 4906
+    },
+    {
+      "epoch": 9.814,
+      "grad_norm": 1.290330410003662,
+      "learning_rate": 2e-05,
+      "loss": 0.0594513,
+      "step": 4907
+    },
+    {
+      "epoch": 9.816,
+      "grad_norm": 0.9869194030761719,
+      "learning_rate": 2e-05,
+      "loss": 0.0311925,
+      "step": 4908
+    },
+    {
+      "epoch": 9.818,
+      "grad_norm": 1.442022681236267,
+      "learning_rate": 2e-05,
+      "loss": 0.04770211,
+      "step": 4909
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 1.677892804145813,
+      "learning_rate": 2e-05,
+      "loss": 0.04223461,
+      "step": 4910
+    },
+    {
+      "epoch": 9.822,
+      "grad_norm": 1.2613219022750854,
+      "learning_rate": 2e-05,
+      "loss": 0.04329883,
+      "step": 4911
+    },
+    {
+      "epoch": 9.824,
+      "grad_norm": 2.5619208812713623,
+      "learning_rate": 2e-05,
+      "loss": 0.04533758,
+      "step": 4912
+    },
+    {
+      "epoch": 9.826,
+      "grad_norm": 1.5010818243026733,
+      "learning_rate": 2e-05,
+      "loss": 0.0549286,
+      "step": 4913
+    },
+    {
+      "epoch": 9.828,
+      "grad_norm": 1.5867867469787598,
+      "learning_rate": 2e-05,
+      "loss": 0.05643577,
+      "step": 4914
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 2.7611711025238037,
+      "learning_rate": 2e-05,
+      "loss": 0.04502559,
+      "step": 4915
+    },
+    {
+      "epoch": 9.832,
+      "grad_norm": 1.104813575744629,
+      "learning_rate": 2e-05,
+      "loss": 0.03817623,
+      "step": 4916
+    },
+    {
+      "epoch": 9.834,
+      "grad_norm": 1.927043080329895,
+      "learning_rate": 2e-05,
+      "loss": 0.05190308,
+      "step": 4917
+    },
+    {
+      "epoch": 9.836,
+      "grad_norm": 1.5192525386810303,
+      "learning_rate": 2e-05,
+      "loss": 0.03937745,
+      "step": 4918
+    },
+    {
+      "epoch": 9.838,
+      "grad_norm": 1.665751576423645,
+      "learning_rate": 2e-05,
+      "loss": 0.05758657,
+      "step": 4919
+    },
+    {
+      "epoch": 9.84,
+      "grad_norm": 2.303105354309082,
+      "learning_rate": 2e-05,
+      "loss": 0.05672823,
+      "step": 4920
+    },
+    {
+      "epoch": 9.842,
+      "grad_norm": 1.7795699834823608,
+      "learning_rate": 2e-05,
+      "loss": 0.05460572,
+      "step": 4921
+    },
+    {
+      "epoch": 9.844,
+      "grad_norm": 1.8190152645111084,
+      "learning_rate": 2e-05,
+      "loss": 0.04848269,
+      "step": 4922
+    },
+    {
+      "epoch": 9.846,
+      "grad_norm": 1.7064250707626343,
+      "learning_rate": 2e-05,
+      "loss": 0.05619481,
+      "step": 4923
+    },
+    {
+      "epoch": 9.848,
+      "grad_norm": 1.413002610206604,
+      "learning_rate": 2e-05,
+      "loss": 0.04609952,
+      "step": 4924
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 1.523747444152832,
+      "learning_rate": 2e-05,
+      "loss": 0.06341836,
+      "step": 4925
+    },
+    {
+      "epoch": 9.852,
+      "grad_norm": 1.346929669380188,
+      "learning_rate": 2e-05,
+      "loss": 0.03450187,
+      "step": 4926
+    },
+    {
+      "epoch": 9.854,
+      "grad_norm": 1.8626230955123901,
+      "learning_rate": 2e-05,
+      "loss": 0.04854961,
+      "step": 4927
+    },
+    {
+      "epoch": 9.856,
+      "grad_norm": 1.1077322959899902,
+      "learning_rate": 2e-05,
+      "loss": 0.03540801,
+      "step": 4928
+    },
+    {
+      "epoch": 9.858,
+      "grad_norm": 1.6448397636413574,
+      "learning_rate": 2e-05,
+      "loss": 0.04669225,
+      "step": 4929
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 2.1358819007873535,
+      "learning_rate": 2e-05,
+      "loss": 0.05960096,
+      "step": 4930
+    },
+    {
+      "epoch": 9.862,
+      "grad_norm": 4.25274658203125,
+      "learning_rate": 2e-05,
+      "loss": 0.08006147,
+      "step": 4931
+    },
+    {
+      "epoch": 9.864,
+      "grad_norm": 1.9246102571487427,
+      "learning_rate": 2e-05,
+      "loss": 0.07823011,
+      "step": 4932
+    },
+    {
+      "epoch": 9.866,
+      "grad_norm": 1.5427088737487793,
+      "learning_rate": 2e-05,
+      "loss": 0.04289813,
+      "step": 4933
+    },
+    {
+      "epoch": 9.868,
+      "grad_norm": 1.953423261642456,
+      "learning_rate": 2e-05,
+      "loss": 0.05337017,
+      "step": 4934
+    },
+    {
+      "epoch": 9.87,
+      "grad_norm": 1.5664440393447876,
+      "learning_rate": 2e-05,
+      "loss": 0.05721167,
+      "step": 4935
+    },
+    {
+      "epoch": 9.872,
+      "grad_norm": 2.196546792984009,
+      "learning_rate": 2e-05,
+      "loss": 0.05689994,
+      "step": 4936
+    },
+    {
+      "epoch": 9.874,
+      "grad_norm": 1.7827051877975464,
+      "learning_rate": 2e-05,
+      "loss": 0.05442892,
+      "step": 4937
+    },
+    {
+      "epoch": 9.876,
+      "grad_norm": 1.2307844161987305,
+      "learning_rate": 2e-05,
+      "loss": 0.04137726,
+      "step": 4938
+    },
+    {
+      "epoch": 9.878,
+      "grad_norm": 1.3965389728546143,
+      "learning_rate": 2e-05,
+      "loss": 0.04177804,
+      "step": 4939
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 1.301183819770813,
+      "learning_rate": 2e-05,
+      "loss": 0.04382346,
+      "step": 4940
+    },
+    {
+      "epoch": 9.882,
+      "grad_norm": 2.091052293777466,
+      "learning_rate": 2e-05,
+      "loss": 0.05116628,
+      "step": 4941
+    },
+    {
+      "epoch": 9.884,
+      "grad_norm": 1.581408143043518,
+      "learning_rate": 2e-05,
+      "loss": 0.05110285,
+      "step": 4942
+    },
+    {
+      "epoch": 9.886,
+      "grad_norm": 2.7291102409362793,
+      "learning_rate": 2e-05,
+      "loss": 0.06126875,
+      "step": 4943
+    },
+    {
+      "epoch": 9.888,
+      "grad_norm": 2.2793638706207275,
+      "learning_rate": 2e-05,
+      "loss": 0.05009011,
+      "step": 4944
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 3.306272506713867,
+      "learning_rate": 2e-05,
+      "loss": 0.06589854,
+      "step": 4945
+    },
+    {
+      "epoch": 9.892,
+      "grad_norm": 1.4898868799209595,
+      "learning_rate": 2e-05,
+      "loss": 0.05594926,
+      "step": 4946
+    },
+    {
+      "epoch": 9.894,
+      "grad_norm": 2.2353460788726807,
+      "learning_rate": 2e-05,
+      "loss": 0.06084714,
+      "step": 4947
+    },
+    {
+      "epoch": 9.896,
+      "grad_norm": 2.330730676651001,
+      "learning_rate": 2e-05,
+      "loss": 0.05106577,
+      "step": 4948
+    },
+    {
+      "epoch": 9.898,
+      "grad_norm": 1.3714518547058105,
+      "learning_rate": 2e-05,
+      "loss": 0.04674903,
+      "step": 4949
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 1.4012672901153564,
+      "learning_rate": 2e-05,
+      "loss": 0.05167598,
+      "step": 4950
+    },
+    {
+      "epoch": 9.902,
+      "grad_norm": 1.0671659708023071,
+      "learning_rate": 2e-05,
+      "loss": 0.02830281,
+      "step": 4951
+    },
+    {
+      "epoch": 9.904,
+      "grad_norm": 1.8455958366394043,
+      "learning_rate": 2e-05,
+      "loss": 0.05752491,
+      "step": 4952
+    },
+    {
+      "epoch": 9.906,
+      "grad_norm": 1.4932456016540527,
+      "learning_rate": 2e-05,
+      "loss": 0.04646926,
+      "step": 4953
+    },
+    {
+      "epoch": 9.908,
+      "grad_norm": 1.6366294622421265,
+      "learning_rate": 2e-05,
+      "loss": 0.0468614,
+      "step": 4954
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 2.687469720840454,
+      "learning_rate": 2e-05,
+      "loss": 0.0775534,
+      "step": 4955
+    },
+    {
+      "epoch": 9.912,
+      "grad_norm": 3.0272841453552246,
+      "learning_rate": 2e-05,
+      "loss": 0.06905685,
+      "step": 4956
+    },
+    {
+      "epoch": 9.914,
+      "grad_norm": 2.0984997749328613,
+      "learning_rate": 2e-05,
+      "loss": 0.07175948,
+      "step": 4957
+    },
+    {
+      "epoch": 9.916,
+      "grad_norm": 2.3406221866607666,
+      "learning_rate": 2e-05,
+      "loss": 0.06852732,
+      "step": 4958
+    },
+    {
+      "epoch": 9.918,
+      "grad_norm": 2.1772301197052,
+      "learning_rate": 2e-05,
+      "loss": 0.0689801,
+      "step": 4959
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 1.6257960796356201,
+      "learning_rate": 2e-05,
+      "loss": 0.05941542,
+      "step": 4960
+    },
+    {
+      "epoch": 9.922,
+      "grad_norm": 2.138094663619995,
+      "learning_rate": 2e-05,
+      "loss": 0.04775422,
+      "step": 4961
+    },
+    {
+      "epoch": 9.924,
+      "grad_norm": 2.711124897003174,
+      "learning_rate": 2e-05,
+      "loss": 0.06104578,
+      "step": 4962
+    },
+    {
+      "epoch": 9.926,
+      "grad_norm": 2.045482873916626,
+      "learning_rate": 2e-05,
+      "loss": 0.06203432,
+      "step": 4963
+    },
+    {
+      "epoch": 9.928,
+      "grad_norm": 1.3155605792999268,
+      "learning_rate": 2e-05,
+      "loss": 0.05150554,
+      "step": 4964
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 1.79209566116333,
+      "learning_rate": 2e-05,
+      "loss": 0.04305602,
+      "step": 4965
+    },
+    {
+      "epoch": 9.932,
+      "grad_norm": 2.037360191345215,
+      "learning_rate": 2e-05,
+      "loss": 0.05929771,
+      "step": 4966
+    },
+    {
+      "epoch": 9.934,
+      "grad_norm": 1.8728766441345215,
+      "learning_rate": 2e-05,
+      "loss": 0.05034747,
+      "step": 4967
+    },
+    {
+      "epoch": 9.936,
+      "grad_norm": 1.58304762840271,
+      "learning_rate": 2e-05,
+      "loss": 0.04624342,
+      "step": 4968
+    },
+    {
+      "epoch": 9.938,
+      "grad_norm": 1.070703387260437,
+      "learning_rate": 2e-05,
+      "loss": 0.04769868,
+      "step": 4969
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 1.5392509698867798,
+      "learning_rate": 2e-05,
+      "loss": 0.04278378,
+      "step": 4970
+    },
+    {
+      "epoch": 9.942,
+      "grad_norm": 1.853967308998108,
+      "learning_rate": 2e-05,
+      "loss": 0.06206827,
+      "step": 4971
+    },
+    {
+      "epoch": 9.943999999999999,
+      "grad_norm": 1.6598646640777588,
+      "learning_rate": 2e-05,
+      "loss": 0.05348482,
+      "step": 4972
+    },
+    {
+      "epoch": 9.946,
+      "grad_norm": 1.3386852741241455,
+      "learning_rate": 2e-05,
+      "loss": 0.05379867,
+      "step": 4973
+    },
+    {
+      "epoch": 9.948,
+      "grad_norm": 1.0884183645248413,
+      "learning_rate": 2e-05,
+      "loss": 0.04097927,
+      "step": 4974
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 2.1080918312072754,
+      "learning_rate": 2e-05,
+      "loss": 0.05956004,
+      "step": 4975
+    },
+    {
+      "epoch": 9.952,
+      "grad_norm": 3.454415798187256,
+      "learning_rate": 2e-05,
+      "loss": 0.05866547,
+      "step": 4976
+    },
+    {
+      "epoch": 9.954,
+      "grad_norm": 1.8443541526794434,
+      "learning_rate": 2e-05,
+      "loss": 0.05498111,
+      "step": 4977
+    },
+    {
+      "epoch": 9.956,
+      "grad_norm": 1.2437260150909424,
+      "learning_rate": 2e-05,
+      "loss": 0.04896142,
+      "step": 4978
+    },
+    {
+      "epoch": 9.958,
+      "grad_norm": 2.130713701248169,
+      "learning_rate": 2e-05,
+      "loss": 0.05966695,
+      "step": 4979
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 2.0753307342529297,
+      "learning_rate": 2e-05,
+      "loss": 0.0618369,
+      "step": 4980
+    },
+    {
+      "epoch": 9.962,
+      "grad_norm": 2.517958879470825,
+      "learning_rate": 2e-05,
+      "loss": 0.06341064,
+      "step": 4981
+    },
+    {
+      "epoch": 9.964,
+      "grad_norm": 1.4736145734786987,
+      "learning_rate": 2e-05,
+      "loss": 0.05388065,
+      "step": 4982
+    },
+    {
+      "epoch": 9.966,
+      "grad_norm": 2.1390750408172607,
+      "learning_rate": 2e-05,
+      "loss": 0.0556998,
+      "step": 4983
+    },
+    {
+      "epoch": 9.968,
+      "grad_norm": 1.417447566986084,
+      "learning_rate": 2e-05,
+      "loss": 0.04931127,
+      "step": 4984
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 1.8981767892837524,
+      "learning_rate": 2e-05,
+      "loss": 0.04368909,
+      "step": 4985
+    },
+    {
+      "epoch": 9.972,
+      "grad_norm": 2.0923216342926025,
+      "learning_rate": 2e-05,
+      "loss": 0.0481521,
+      "step": 4986
+    },
+    {
+      "epoch": 9.974,
+      "grad_norm": 2.3560774326324463,
+      "learning_rate": 2e-05,
+      "loss": 0.04617333,
+      "step": 4987
+    },
+    {
+      "epoch": 9.975999999999999,
+      "grad_norm": 2.1225790977478027,
+      "learning_rate": 2e-05,
+      "loss": 0.06341647,
+      "step": 4988
+    },
+    {
+      "epoch": 9.978,
+      "grad_norm": 1.4702603816986084,
+      "learning_rate": 2e-05,
+      "loss": 0.06608592,
+      "step": 4989
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 1.437772512435913,
+      "learning_rate": 2e-05,
+      "loss": 0.0440442,
+      "step": 4990
+    },
+    {
+      "epoch": 9.982,
+      "grad_norm": 1.7781963348388672,
+      "learning_rate": 2e-05,
+      "loss": 0.04864323,
+      "step": 4991
+    },
+    {
+      "epoch": 9.984,
+      "grad_norm": 2.0641047954559326,
+      "learning_rate": 2e-05,
+      "loss": 0.05038781,
+      "step": 4992
+    },
+    {
+      "epoch": 9.986,
+      "grad_norm": 1.471038818359375,
+      "learning_rate": 2e-05,
+      "loss": 0.05805187,
+      "step": 4993
+    },
+    {
+      "epoch": 9.988,
+      "grad_norm": 1.7074260711669922,
+      "learning_rate": 2e-05,
+      "loss": 0.04913766,
+      "step": 4994
+    },
+    {
+      "epoch": 9.99,
+      "grad_norm": 1.5524414777755737,
+      "learning_rate": 2e-05,
+      "loss": 0.04417773,
+      "step": 4995
+    },
+    {
+      "epoch": 9.992,
+      "grad_norm": 2.4501919746398926,
+      "learning_rate": 2e-05,
+      "loss": 0.05671235,
+      "step": 4996
+    },
+    {
+      "epoch": 9.994,
+      "grad_norm": 1.801255226135254,
+      "learning_rate": 2e-05,
+      "loss": 0.06374493,
+      "step": 4997
+    },
+    {
+      "epoch": 9.996,
+      "grad_norm": 1.989261507987976,
+      "learning_rate": 2e-05,
+      "loss": 0.04868555,
+      "step": 4998
+    },
+    {
+      "epoch": 9.998,
+      "grad_norm": 1.389630675315857,
+      "learning_rate": 2e-05,
+      "loss": 0.0455512,
+      "step": 4999
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 1.467817783355713,
+      "learning_rate": 2e-05,
+      "loss": 0.06919758,
+      "step": 5000
+    },
+    {
+      "epoch": 10.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.988,
+        "AngleClassification_2": 0.994,
+        "AngleClassification_3": 0.9560878243512974,
+        "Equal_1": 0.998,
+        "Equal_2": 0.9600798403193613,
+        "Equal_3": 0.8542914171656687,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9900199600798403,
+        "Parallel_1": 0.9899799599198397,
+        "Parallel_2": 0.9959919839679359,
+        "Parallel_3": 0.988,
+        "Perpendicular_1": 0.988,
+        "Perpendicular_2": 0.916,
+        "Perpendicular_3": 0.5741482965931863,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.994,
+        "PointLiesOnCircle_3": 0.996,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9899799599198397,
+        "PointLiesOnLine_3": 0.9580838323353293
+      },
+      "eval_runtime": 320.6792,
+      "eval_samples_per_second": 32.743,
+      "eval_steps_per_second": 0.655,
+      "step": 5000
+    },
+    {
+      "epoch": 10.002,
+      "grad_norm": 1.6074891090393066,
+      "learning_rate": 2e-05,
+      "loss": 0.05639043,
+      "step": 5001
+    },
+    {
+      "epoch": 10.004,
+      "grad_norm": 1.6433279514312744,
+      "learning_rate": 2e-05,
+      "loss": 0.06391891,
+      "step": 5002
+    },
+    {
+      "epoch": 10.006,
+      "grad_norm": 1.0623618364334106,
+      "learning_rate": 2e-05,
+      "loss": 0.0363953,
+      "step": 5003
+    },
+    {
+      "epoch": 10.008,
+      "grad_norm": 2.0282094478607178,
+      "learning_rate": 2e-05,
+      "loss": 0.05439749,
+      "step": 5004
+    },
+    {
+      "epoch": 10.01,
+      "grad_norm": 2.1043431758880615,
+      "learning_rate": 2e-05,
+      "loss": 0.05599514,
+      "step": 5005
+    },
+    {
+      "epoch": 10.012,
+      "grad_norm": 1.4989700317382812,
+      "learning_rate": 2e-05,
+      "loss": 0.04970724,
+      "step": 5006
+    },
+    {
+      "epoch": 10.014,
+      "grad_norm": 1.6386247873306274,
+      "learning_rate": 2e-05,
+      "loss": 0.04516533,
+      "step": 5007
+    },
+    {
+      "epoch": 10.016,
+      "grad_norm": 1.3612985610961914,
+      "learning_rate": 2e-05,
+      "loss": 0.05006377,
+      "step": 5008
+    },
+    {
+      "epoch": 10.018,
+      "grad_norm": 1.6199952363967896,
+      "learning_rate": 2e-05,
+      "loss": 0.0529905,
+      "step": 5009
+    },
+    {
+      "epoch": 10.02,
+      "grad_norm": 1.690873384475708,
+      "learning_rate": 2e-05,
+      "loss": 0.06474019,
+      "step": 5010
+    },
+    {
+      "epoch": 10.022,
+      "grad_norm": 1.0925859212875366,
+      "learning_rate": 2e-05,
+      "loss": 0.04544307,
+      "step": 5011
+    },
+    {
+      "epoch": 10.024,
+      "grad_norm": 2.489445447921753,
+      "learning_rate": 2e-05,
+      "loss": 0.04316767,
+      "step": 5012
+    },
+    {
+      "epoch": 10.026,
+      "grad_norm": 1.0583429336547852,
+      "learning_rate": 2e-05,
+      "loss": 0.03954223,
+      "step": 5013
+    },
+    {
+      "epoch": 10.028,
+      "grad_norm": 2.259428024291992,
+      "learning_rate": 2e-05,
+      "loss": 0.05867039,
+      "step": 5014
+    },
+    {
+      "epoch": 10.03,
+      "grad_norm": 1.1833878755569458,
+      "learning_rate": 2e-05,
+      "loss": 0.04461697,
+      "step": 5015
+    },
+    {
+      "epoch": 10.032,
+      "grad_norm": 1.7030174732208252,
+      "learning_rate": 2e-05,
+      "loss": 0.06960414,
+      "step": 5016
+    },
+    {
+      "epoch": 10.034,
+      "grad_norm": 2.5388693809509277,
+      "learning_rate": 2e-05,
+      "loss": 0.04828655,
+      "step": 5017
+    },
+    {
+      "epoch": 10.036,
+      "grad_norm": 1.4048974514007568,
+      "learning_rate": 2e-05,
+      "loss": 0.03989777,
+      "step": 5018
+    },
+    {
+      "epoch": 10.038,
+      "grad_norm": 1.6983534097671509,
+      "learning_rate": 2e-05,
+      "loss": 0.04251143,
+      "step": 5019
+    },
+    {
+      "epoch": 10.04,
+      "grad_norm": 2.109666347503662,
+      "learning_rate": 2e-05,
+      "loss": 0.04322487,
+      "step": 5020
+    },
+    {
+      "epoch": 10.042,
+      "grad_norm": 1.6029553413391113,
+      "learning_rate": 2e-05,
+      "loss": 0.06342378,
+      "step": 5021
+    },
+    {
+      "epoch": 10.044,
+      "grad_norm": 1.3456430435180664,
+      "learning_rate": 2e-05,
+      "loss": 0.04814564,
+      "step": 5022
+    },
+    {
+      "epoch": 10.046,
+      "grad_norm": 1.7737743854522705,
+      "learning_rate": 2e-05,
+      "loss": 0.0458786,
+      "step": 5023
+    },
+    {
+      "epoch": 10.048,
+      "grad_norm": 1.2273495197296143,
+      "learning_rate": 2e-05,
+      "loss": 0.03703401,
+      "step": 5024
+    },
+    {
+      "epoch": 10.05,
+      "grad_norm": 2.4264330863952637,
+      "learning_rate": 2e-05,
+      "loss": 0.07618839,
+      "step": 5025
+    },
+    {
+      "epoch": 10.052,
+      "grad_norm": 1.851207971572876,
+      "learning_rate": 2e-05,
+      "loss": 0.07478073,
+      "step": 5026
+    },
+    {
+      "epoch": 10.054,
+      "grad_norm": 1.3361573219299316,
+      "learning_rate": 2e-05,
+      "loss": 0.05228508,
+      "step": 5027
+    },
+    {
+      "epoch": 10.056,
+      "grad_norm": 2.115432024002075,
+      "learning_rate": 2e-05,
+      "loss": 0.06497653,
+      "step": 5028
+    },
+    {
+      "epoch": 10.058,
+      "grad_norm": 1.5752816200256348,
+      "learning_rate": 2e-05,
+      "loss": 0.04064302,
+      "step": 5029
+    },
+    {
+      "epoch": 10.06,
+      "grad_norm": 1.224853754043579,
+      "learning_rate": 2e-05,
+      "loss": 0.04076491,
+      "step": 5030
+    },
+    {
+      "epoch": 10.062,
+      "grad_norm": 1.0174431800842285,
+      "learning_rate": 2e-05,
+      "loss": 0.02776841,
+      "step": 5031
+    },
+    {
+      "epoch": 10.064,
+      "grad_norm": 1.4870493412017822,
+      "learning_rate": 2e-05,
+      "loss": 0.05205495,
+      "step": 5032
+    },
+    {
+      "epoch": 10.066,
+      "grad_norm": 1.233358383178711,
+      "learning_rate": 2e-05,
+      "loss": 0.04476997,
+      "step": 5033
+    },
+    {
+      "epoch": 10.068,
+      "grad_norm": 1.4965168237686157,
+      "learning_rate": 2e-05,
+      "loss": 0.06173176,
+      "step": 5034
+    },
+    {
+      "epoch": 10.07,
+      "grad_norm": 1.2926126718521118,
+      "learning_rate": 2e-05,
+      "loss": 0.05273168,
+      "step": 5035
+    },
+    {
+      "epoch": 10.072,
+      "grad_norm": 1.9668138027191162,
+      "learning_rate": 2e-05,
+      "loss": 0.05729368,
+      "step": 5036
+    },
+    {
+      "epoch": 10.074,
+      "grad_norm": 1.9118866920471191,
+      "learning_rate": 2e-05,
+      "loss": 0.07089907,
+      "step": 5037
+    },
+    {
+      "epoch": 10.076,
+      "grad_norm": 1.567081093788147,
+      "learning_rate": 2e-05,
+      "loss": 0.04848254,
+      "step": 5038
+    },
+    {
+      "epoch": 10.078,
+      "grad_norm": 1.3504266738891602,
+      "learning_rate": 2e-05,
+      "loss": 0.04428206,
+      "step": 5039
+    },
+    {
+      "epoch": 10.08,
+      "grad_norm": 1.5226638317108154,
+      "learning_rate": 2e-05,
+      "loss": 0.05267453,
+      "step": 5040
+    },
+    {
+      "epoch": 10.082,
+      "grad_norm": 2.2543962001800537,
+      "learning_rate": 2e-05,
+      "loss": 0.05228227,
+      "step": 5041
+    },
+    {
+      "epoch": 10.084,
+      "grad_norm": 1.4806350469589233,
+      "learning_rate": 2e-05,
+      "loss": 0.04181156,
+      "step": 5042
+    },
+    {
+      "epoch": 10.086,
+      "grad_norm": 1.2908692359924316,
+      "learning_rate": 2e-05,
+      "loss": 0.04454116,
+      "step": 5043
+    },
+    {
+      "epoch": 10.088,
+      "grad_norm": 2.006575584411621,
+      "learning_rate": 2e-05,
+      "loss": 0.05524797,
+      "step": 5044
+    },
+    {
+      "epoch": 10.09,
+      "grad_norm": 1.949083685874939,
+      "learning_rate": 2e-05,
+      "loss": 0.04127013,
+      "step": 5045
+    },
+    {
+      "epoch": 10.092,
+      "grad_norm": 1.6486314535140991,
+      "learning_rate": 2e-05,
+      "loss": 0.06116631,
+      "step": 5046
+    },
+    {
+      "epoch": 10.094,
+      "grad_norm": 1.8648558855056763,
+      "learning_rate": 2e-05,
+      "loss": 0.04682179,
+      "step": 5047
+    },
+    {
+      "epoch": 10.096,
+      "grad_norm": 2.304664134979248,
+      "learning_rate": 2e-05,
+      "loss": 0.05680639,
+      "step": 5048
+    },
+    {
+      "epoch": 10.098,
+      "grad_norm": 1.4848788976669312,
+      "learning_rate": 2e-05,
+      "loss": 0.03981915,
+      "step": 5049
+    },
+    {
+      "epoch": 10.1,
+      "grad_norm": 1.9800957441329956,
+      "learning_rate": 2e-05,
+      "loss": 0.03908926,
+      "step": 5050
+    },
+    {
+      "epoch": 10.102,
+      "grad_norm": 1.5394706726074219,
+      "learning_rate": 2e-05,
+      "loss": 0.05664773,
+      "step": 5051
+    },
+    {
+      "epoch": 10.104,
+      "grad_norm": 2.171658992767334,
+      "learning_rate": 2e-05,
+      "loss": 0.05971353,
+      "step": 5052
+    },
+    {
+      "epoch": 10.106,
+      "grad_norm": 2.1404213905334473,
+      "learning_rate": 2e-05,
+      "loss": 0.05590626,
+      "step": 5053
+    },
+    {
+      "epoch": 10.108,
+      "grad_norm": 3.454113483428955,
+      "learning_rate": 2e-05,
+      "loss": 0.0404954,
+      "step": 5054
+    },
+    {
+      "epoch": 10.11,
+      "grad_norm": 4.064882278442383,
+      "learning_rate": 2e-05,
+      "loss": 0.0535365,
+      "step": 5055
+    },
+    {
+      "epoch": 10.112,
+      "grad_norm": 1.671109914779663,
+      "learning_rate": 2e-05,
+      "loss": 0.05773462,
+      "step": 5056
+    },
+    {
+      "epoch": 10.114,
+      "grad_norm": 2.3881421089172363,
+      "learning_rate": 2e-05,
+      "loss": 0.05546438,
+      "step": 5057
+    },
+    {
+      "epoch": 10.116,
+      "grad_norm": 2.157351493835449,
+      "learning_rate": 2e-05,
+      "loss": 0.05277005,
+      "step": 5058
+    },
+    {
+      "epoch": 10.118,
+      "grad_norm": 2.0532100200653076,
+      "learning_rate": 2e-05,
+      "loss": 0.05053755,
+      "step": 5059
+    },
+    {
+      "epoch": 10.12,
+      "grad_norm": 1.3669548034667969,
+      "learning_rate": 2e-05,
+      "loss": 0.05167579,
+      "step": 5060
+    },
+    {
+      "epoch": 10.122,
+      "grad_norm": 1.5119102001190186,
+      "learning_rate": 2e-05,
+      "loss": 0.05561122,
+      "step": 5061
+    },
+    {
+      "epoch": 10.124,
+      "grad_norm": 1.244271159172058,
+      "learning_rate": 2e-05,
+      "loss": 0.03419596,
+      "step": 5062
+    },
+    {
+      "epoch": 10.126,
+      "grad_norm": 1.6181620359420776,
+      "learning_rate": 2e-05,
+      "loss": 0.06645474,
+      "step": 5063
+    },
+    {
+      "epoch": 10.128,
+      "grad_norm": 2.0490105152130127,
+      "learning_rate": 2e-05,
+      "loss": 0.05366251,
+      "step": 5064
+    },
+    {
+      "epoch": 10.13,
+      "grad_norm": 1.2846488952636719,
+      "learning_rate": 2e-05,
+      "loss": 0.04805609,
+      "step": 5065
+    },
+    {
+      "epoch": 10.132,
+      "grad_norm": 1.5411098003387451,
+      "learning_rate": 2e-05,
+      "loss": 0.04068913,
+      "step": 5066
+    },
+    {
+      "epoch": 10.134,
+      "grad_norm": 1.4525972604751587,
+      "learning_rate": 2e-05,
+      "loss": 0.04148395,
+      "step": 5067
+    },
+    {
+      "epoch": 10.136,
+      "grad_norm": 1.9913413524627686,
+      "learning_rate": 2e-05,
+      "loss": 0.05251299,
+      "step": 5068
+    },
+    {
+      "epoch": 10.138,
+      "grad_norm": 1.329590082168579,
+      "learning_rate": 2e-05,
+      "loss": 0.03482792,
+      "step": 5069
+    },
+    {
+      "epoch": 10.14,
+      "grad_norm": 1.917649507522583,
+      "learning_rate": 2e-05,
+      "loss": 0.0389032,
+      "step": 5070
+    },
+    {
+      "epoch": 10.142,
+      "grad_norm": 1.8143230676651,
+      "learning_rate": 2e-05,
+      "loss": 0.06296816,
+      "step": 5071
+    },
+    {
+      "epoch": 10.144,
+      "grad_norm": 1.9044404029846191,
+      "learning_rate": 2e-05,
+      "loss": 0.07797446,
+      "step": 5072
+    },
+    {
+      "epoch": 10.146,
+      "grad_norm": 1.3988782167434692,
+      "learning_rate": 2e-05,
+      "loss": 0.03789392,
+      "step": 5073
+    },
+    {
+      "epoch": 10.148,
+      "grad_norm": 2.06538987159729,
+      "learning_rate": 2e-05,
+      "loss": 0.05376375,
+      "step": 5074
+    },
+    {
+      "epoch": 10.15,
+      "grad_norm": 1.908994436264038,
+      "learning_rate": 2e-05,
+      "loss": 0.05527605,
+      "step": 5075
+    },
+    {
+      "epoch": 10.152,
+      "grad_norm": 1.5503228902816772,
+      "learning_rate": 2e-05,
+      "loss": 0.04854936,
+      "step": 5076
+    },
+    {
+      "epoch": 10.154,
+      "grad_norm": 1.1571024656295776,
+      "learning_rate": 2e-05,
+      "loss": 0.03305773,
+      "step": 5077
+    },
+    {
+      "epoch": 10.156,
+      "grad_norm": 2.0785417556762695,
+      "learning_rate": 2e-05,
+      "loss": 0.06657989,
+      "step": 5078
+    },
+    {
+      "epoch": 10.158,
+      "grad_norm": 2.85941743850708,
+      "learning_rate": 2e-05,
+      "loss": 0.05014766,
+      "step": 5079
+    },
+    {
+      "epoch": 10.16,
+      "grad_norm": 2.1629788875579834,
+      "learning_rate": 2e-05,
+      "loss": 0.05455408,
+      "step": 5080
+    },
+    {
+      "epoch": 10.162,
+      "grad_norm": 3.536186933517456,
+      "learning_rate": 2e-05,
+      "loss": 0.04600221,
+      "step": 5081
+    },
+    {
+      "epoch": 10.164,
+      "grad_norm": 1.7894084453582764,
+      "learning_rate": 2e-05,
+      "loss": 0.03903076,
+      "step": 5082
+    },
+    {
+      "epoch": 10.166,
+      "grad_norm": 1.6507161855697632,
+      "learning_rate": 2e-05,
+      "loss": 0.03541517,
+      "step": 5083
+    },
+    {
+      "epoch": 10.168,
+      "grad_norm": 1.4610087871551514,
+      "learning_rate": 2e-05,
+      "loss": 0.05326772,
+      "step": 5084
+    },
+    {
+      "epoch": 10.17,
+      "grad_norm": 1.1885812282562256,
+      "learning_rate": 2e-05,
+      "loss": 0.04884129,
+      "step": 5085
+    },
+    {
+      "epoch": 10.172,
+      "grad_norm": 1.5193572044372559,
+      "learning_rate": 2e-05,
+      "loss": 0.05214652,
+      "step": 5086
+    },
+    {
+      "epoch": 10.174,
+      "grad_norm": 2.79036808013916,
+      "learning_rate": 2e-05,
+      "loss": 0.07177282,
+      "step": 5087
+    },
+    {
+      "epoch": 10.176,
+      "grad_norm": 2.4795050621032715,
+      "learning_rate": 2e-05,
+      "loss": 0.05598135,
+      "step": 5088
+    },
+    {
+      "epoch": 10.178,
+      "grad_norm": 1.8323618173599243,
+      "learning_rate": 2e-05,
+      "loss": 0.04974123,
+      "step": 5089
+    },
+    {
+      "epoch": 10.18,
+      "grad_norm": 1.8436342477798462,
+      "learning_rate": 2e-05,
+      "loss": 0.05738843,
+      "step": 5090
+    },
+    {
+      "epoch": 10.182,
+      "grad_norm": 1.8771424293518066,
+      "learning_rate": 2e-05,
+      "loss": 0.06057268,
+      "step": 5091
+    },
+    {
+      "epoch": 10.184,
+      "grad_norm": 1.8508983850479126,
+      "learning_rate": 2e-05,
+      "loss": 0.05036641,
+      "step": 5092
+    },
+    {
+      "epoch": 10.186,
+      "grad_norm": 1.3412444591522217,
+      "learning_rate": 2e-05,
+      "loss": 0.03264994,
+      "step": 5093
+    },
+    {
+      "epoch": 10.188,
+      "grad_norm": 1.8948533535003662,
+      "learning_rate": 2e-05,
+      "loss": 0.04661043,
+      "step": 5094
+    },
+    {
+      "epoch": 10.19,
+      "grad_norm": 1.1658676862716675,
+      "learning_rate": 2e-05,
+      "loss": 0.0359153,
+      "step": 5095
+    },
+    {
+      "epoch": 10.192,
+      "grad_norm": 2.3413681983947754,
+      "learning_rate": 2e-05,
+      "loss": 0.06752145,
+      "step": 5096
+    },
+    {
+      "epoch": 10.194,
+      "grad_norm": 2.0791563987731934,
+      "learning_rate": 2e-05,
+      "loss": 0.03742487,
+      "step": 5097
+    },
+    {
+      "epoch": 10.196,
+      "grad_norm": 2.3035924434661865,
+      "learning_rate": 2e-05,
+      "loss": 0.05462205,
+      "step": 5098
+    },
+    {
+      "epoch": 10.198,
+      "grad_norm": 1.5185017585754395,
+      "learning_rate": 2e-05,
+      "loss": 0.0364113,
+      "step": 5099
+    },
+    {
+      "epoch": 10.2,
+      "grad_norm": 2.8155453205108643,
+      "learning_rate": 2e-05,
+      "loss": 0.06460752,
+      "step": 5100
+    },
+    {
+      "epoch": 10.202,
+      "grad_norm": 2.565056800842285,
+      "learning_rate": 2e-05,
+      "loss": 0.04938511,
+      "step": 5101
+    },
+    {
+      "epoch": 10.204,
+      "grad_norm": 2.4186198711395264,
+      "learning_rate": 2e-05,
+      "loss": 0.05762579,
+      "step": 5102
+    },
+    {
+      "epoch": 10.206,
+      "grad_norm": 1.7323216199874878,
+      "learning_rate": 2e-05,
+      "loss": 0.04478491,
+      "step": 5103
+    },
+    {
+      "epoch": 10.208,
+      "grad_norm": 2.4618570804595947,
+      "learning_rate": 2e-05,
+      "loss": 0.0601164,
+      "step": 5104
+    },
+    {
+      "epoch": 10.21,
+      "grad_norm": 1.4251370429992676,
+      "learning_rate": 2e-05,
+      "loss": 0.05556741,
+      "step": 5105
+    },
+    {
+      "epoch": 10.212,
+      "grad_norm": 2.33821177482605,
+      "learning_rate": 2e-05,
+      "loss": 0.07309254,
+      "step": 5106
+    },
+    {
+      "epoch": 10.214,
+      "grad_norm": 1.3154981136322021,
+      "learning_rate": 2e-05,
+      "loss": 0.04135425,
+      "step": 5107
+    },
+    {
+      "epoch": 10.216,
+      "grad_norm": 2.5417816638946533,
+      "learning_rate": 2e-05,
+      "loss": 0.06982885,
+      "step": 5108
+    },
+    {
+      "epoch": 10.218,
+      "grad_norm": 2.1002843379974365,
+      "learning_rate": 2e-05,
+      "loss": 0.05075446,
+      "step": 5109
+    },
+    {
+      "epoch": 10.22,
+      "grad_norm": 1.574785828590393,
+      "learning_rate": 2e-05,
+      "loss": 0.04077484,
+      "step": 5110
+    },
+    {
+      "epoch": 10.222,
+      "grad_norm": 1.7429002523422241,
+      "learning_rate": 2e-05,
+      "loss": 0.04333501,
+      "step": 5111
+    },
+    {
+      "epoch": 10.224,
+      "grad_norm": 1.4146769046783447,
+      "learning_rate": 2e-05,
+      "loss": 0.04594901,
+      "step": 5112
+    },
+    {
+      "epoch": 10.226,
+      "grad_norm": 2.0971226692199707,
+      "learning_rate": 2e-05,
+      "loss": 0.06643675,
+      "step": 5113
+    },
+    {
+      "epoch": 10.228,
+      "grad_norm": 1.6116442680358887,
+      "learning_rate": 2e-05,
+      "loss": 0.04431507,
+      "step": 5114
+    },
+    {
+      "epoch": 10.23,
+      "grad_norm": 3.321749210357666,
+      "learning_rate": 2e-05,
+      "loss": 0.06666962,
+      "step": 5115
+    },
+    {
+      "epoch": 10.232,
+      "grad_norm": 1.248750925064087,
+      "learning_rate": 2e-05,
+      "loss": 0.05188263,
+      "step": 5116
+    },
+    {
+      "epoch": 10.234,
+      "grad_norm": 1.4745240211486816,
+      "learning_rate": 2e-05,
+      "loss": 0.0565898,
+      "step": 5117
+    },
+    {
+      "epoch": 10.236,
+      "grad_norm": 1.7704201936721802,
+      "learning_rate": 2e-05,
+      "loss": 0.04397082,
+      "step": 5118
+    },
+    {
+      "epoch": 10.238,
+      "grad_norm": 1.7633501291275024,
+      "learning_rate": 2e-05,
+      "loss": 0.0520197,
+      "step": 5119
+    },
+    {
+      "epoch": 10.24,
+      "grad_norm": 1.8023216724395752,
+      "learning_rate": 2e-05,
+      "loss": 0.06178421,
+      "step": 5120
+    },
+    {
+      "epoch": 10.242,
+      "grad_norm": 1.9784626960754395,
+      "learning_rate": 2e-05,
+      "loss": 0.05770921,
+      "step": 5121
+    },
+    {
+      "epoch": 10.244,
+      "grad_norm": 1.6166801452636719,
+      "learning_rate": 2e-05,
+      "loss": 0.05028439,
+      "step": 5122
+    },
+    {
+      "epoch": 10.246,
+      "grad_norm": 1.2488207817077637,
+      "learning_rate": 2e-05,
+      "loss": 0.05257055,
+      "step": 5123
+    },
+    {
+      "epoch": 10.248,
+      "grad_norm": 1.3418329954147339,
+      "learning_rate": 2e-05,
+      "loss": 0.03734386,
+      "step": 5124
+    },
+    {
+      "epoch": 10.25,
+      "grad_norm": 1.5167311429977417,
+      "learning_rate": 2e-05,
+      "loss": 0.0557439,
+      "step": 5125
+    },
+    {
+      "epoch": 10.252,
+      "grad_norm": 1.087599515914917,
+      "learning_rate": 2e-05,
+      "loss": 0.02959022,
+      "step": 5126
+    },
+    {
+      "epoch": 10.254,
+      "grad_norm": 1.6808693408966064,
+      "learning_rate": 2e-05,
+      "loss": 0.04390074,
+      "step": 5127
+    },
+    {
+      "epoch": 10.256,
+      "grad_norm": 1.909277319908142,
+      "learning_rate": 2e-05,
+      "loss": 0.05320806,
+      "step": 5128
+    },
+    {
+      "epoch": 10.258,
+      "grad_norm": 2.03977370262146,
+      "learning_rate": 2e-05,
+      "loss": 0.05649532,
+      "step": 5129
+    },
+    {
+      "epoch": 10.26,
+      "grad_norm": 1.8067188262939453,
+      "learning_rate": 2e-05,
+      "loss": 0.06265157,
+      "step": 5130
+    },
+    {
+      "epoch": 10.262,
+      "grad_norm": 2.781466007232666,
+      "learning_rate": 2e-05,
+      "loss": 0.06059079,
+      "step": 5131
+    },
+    {
+      "epoch": 10.264,
+      "grad_norm": 2.2374839782714844,
+      "learning_rate": 2e-05,
+      "loss": 0.04488031,
+      "step": 5132
+    },
+    {
+      "epoch": 10.266,
+      "grad_norm": 1.5937252044677734,
+      "learning_rate": 2e-05,
+      "loss": 0.04937337,
+      "step": 5133
+    },
+    {
+      "epoch": 10.268,
+      "grad_norm": 1.5034781694412231,
+      "learning_rate": 2e-05,
+      "loss": 0.04895452,
+      "step": 5134
+    },
+    {
+      "epoch": 10.27,
+      "grad_norm": 2.400557279586792,
+      "learning_rate": 2e-05,
+      "loss": 0.06906497,
+      "step": 5135
+    },
+    {
+      "epoch": 10.272,
+      "grad_norm": 1.635102391242981,
+      "learning_rate": 2e-05,
+      "loss": 0.03693788,
+      "step": 5136
+    },
+    {
+      "epoch": 10.274000000000001,
+      "grad_norm": 1.3570059537887573,
+      "learning_rate": 2e-05,
+      "loss": 0.05358264,
+      "step": 5137
+    },
+    {
+      "epoch": 10.276,
+      "grad_norm": 1.6576945781707764,
+      "learning_rate": 2e-05,
+      "loss": 0.05347702,
+      "step": 5138
+    },
+    {
+      "epoch": 10.278,
+      "grad_norm": 1.7237348556518555,
+      "learning_rate": 2e-05,
+      "loss": 0.04945024,
+      "step": 5139
+    },
+    {
+      "epoch": 10.28,
+      "grad_norm": 1.9865562915802002,
+      "learning_rate": 2e-05,
+      "loss": 0.05404085,
+      "step": 5140
+    },
+    {
+      "epoch": 10.282,
+      "grad_norm": 2.0699520111083984,
+      "learning_rate": 2e-05,
+      "loss": 0.04398727,
+      "step": 5141
+    },
+    {
+      "epoch": 10.284,
+      "grad_norm": 1.870590329170227,
+      "learning_rate": 2e-05,
+      "loss": 0.05769954,
+      "step": 5142
+    },
+    {
+      "epoch": 10.286,
+      "grad_norm": 1.6009098291397095,
+      "learning_rate": 2e-05,
+      "loss": 0.05270899,
+      "step": 5143
+    },
+    {
+      "epoch": 10.288,
+      "grad_norm": 1.0542221069335938,
+      "learning_rate": 2e-05,
+      "loss": 0.03725312,
+      "step": 5144
+    },
+    {
+      "epoch": 10.29,
+      "grad_norm": 1.4165904521942139,
+      "learning_rate": 2e-05,
+      "loss": 0.04188281,
+      "step": 5145
+    },
+    {
+      "epoch": 10.292,
+      "grad_norm": 3.5521252155303955,
+      "learning_rate": 2e-05,
+      "loss": 0.06759865,
+      "step": 5146
+    },
+    {
+      "epoch": 10.294,
+      "grad_norm": 1.433131217956543,
+      "learning_rate": 2e-05,
+      "loss": 0.05106829,
+      "step": 5147
+    },
+    {
+      "epoch": 10.296,
+      "grad_norm": 1.4577850103378296,
+      "learning_rate": 2e-05,
+      "loss": 0.05083516,
+      "step": 5148
+    },
+    {
+      "epoch": 10.298,
+      "grad_norm": 1.3248356580734253,
+      "learning_rate": 2e-05,
+      "loss": 0.04911773,
+      "step": 5149
+    },
+    {
+      "epoch": 10.3,
+      "grad_norm": 1.37599778175354,
+      "learning_rate": 2e-05,
+      "loss": 0.04473313,
+      "step": 5150
+    },
+    {
+      "epoch": 10.302,
+      "grad_norm": 1.5513116121292114,
+      "learning_rate": 2e-05,
+      "loss": 0.04096273,
+      "step": 5151
+    },
+    {
+      "epoch": 10.304,
+      "grad_norm": 1.411054015159607,
+      "learning_rate": 2e-05,
+      "loss": 0.04515607,
+      "step": 5152
+    },
+    {
+      "epoch": 10.306,
+      "grad_norm": 2.094132661819458,
+      "learning_rate": 2e-05,
+      "loss": 0.04543452,
+      "step": 5153
+    },
+    {
+      "epoch": 10.308,
+      "grad_norm": 1.480191707611084,
+      "learning_rate": 2e-05,
+      "loss": 0.04611487,
+      "step": 5154
+    },
+    {
+      "epoch": 10.31,
+      "grad_norm": 1.5969020128250122,
+      "learning_rate": 2e-05,
+      "loss": 0.05584262,
+      "step": 5155
+    },
+    {
+      "epoch": 10.312,
+      "grad_norm": 1.505147099494934,
+      "learning_rate": 2e-05,
+      "loss": 0.05068387,
+      "step": 5156
+    },
+    {
+      "epoch": 10.314,
+      "grad_norm": 1.7500067949295044,
+      "learning_rate": 2e-05,
+      "loss": 0.04701269,
+      "step": 5157
+    },
+    {
+      "epoch": 10.316,
+      "grad_norm": 1.2172950506210327,
+      "learning_rate": 2e-05,
+      "loss": 0.05006229,
+      "step": 5158
+    },
+    {
+      "epoch": 10.318,
+      "grad_norm": 1.3415688276290894,
+      "learning_rate": 2e-05,
+      "loss": 0.04002773,
+      "step": 5159
+    },
+    {
+      "epoch": 10.32,
+      "grad_norm": 1.4847948551177979,
+      "learning_rate": 2e-05,
+      "loss": 0.05143999,
+      "step": 5160
+    },
+    {
+      "epoch": 10.322,
+      "grad_norm": 1.2036330699920654,
+      "learning_rate": 2e-05,
+      "loss": 0.05749707,
+      "step": 5161
+    },
+    {
+      "epoch": 10.324,
+      "grad_norm": 1.8553268909454346,
+      "learning_rate": 2e-05,
+      "loss": 0.06461729,
+      "step": 5162
+    },
+    {
+      "epoch": 10.326,
+      "grad_norm": 1.2511320114135742,
+      "learning_rate": 2e-05,
+      "loss": 0.04641117,
+      "step": 5163
+    },
+    {
+      "epoch": 10.328,
+      "grad_norm": 3.30887770652771,
+      "learning_rate": 2e-05,
+      "loss": 0.0428102,
+      "step": 5164
+    },
+    {
+      "epoch": 10.33,
+      "grad_norm": 3.343966007232666,
+      "learning_rate": 2e-05,
+      "loss": 0.06479934,
+      "step": 5165
+    },
+    {
+      "epoch": 10.332,
+      "grad_norm": 1.185491919517517,
+      "learning_rate": 2e-05,
+      "loss": 0.0378601,
+      "step": 5166
+    },
+    {
+      "epoch": 10.334,
+      "grad_norm": 1.7531931400299072,
+      "learning_rate": 2e-05,
+      "loss": 0.04356651,
+      "step": 5167
+    },
+    {
+      "epoch": 10.336,
+      "grad_norm": 1.5299067497253418,
+      "learning_rate": 2e-05,
+      "loss": 0.0489331,
+      "step": 5168
+    },
+    {
+      "epoch": 10.338,
+      "grad_norm": 1.7191680669784546,
+      "learning_rate": 2e-05,
+      "loss": 0.05703234,
+      "step": 5169
+    },
+    {
+      "epoch": 10.34,
+      "grad_norm": 2.2998485565185547,
+      "learning_rate": 2e-05,
+      "loss": 0.05554244,
+      "step": 5170
+    },
+    {
+      "epoch": 10.342,
+      "grad_norm": 1.4917652606964111,
+      "learning_rate": 2e-05,
+      "loss": 0.0459786,
+      "step": 5171
+    },
+    {
+      "epoch": 10.344,
+      "grad_norm": 2.3985610008239746,
+      "learning_rate": 2e-05,
+      "loss": 0.07849705,
+      "step": 5172
+    },
+    {
+      "epoch": 10.346,
+      "grad_norm": 1.3253332376480103,
+      "learning_rate": 2e-05,
+      "loss": 0.03654364,
+      "step": 5173
+    },
+    {
+      "epoch": 10.348,
+      "grad_norm": 1.9052636623382568,
+      "learning_rate": 2e-05,
+      "loss": 0.06166274,
+      "step": 5174
+    },
+    {
+      "epoch": 10.35,
+      "grad_norm": 1.6939311027526855,
+      "learning_rate": 2e-05,
+      "loss": 0.04354223,
+      "step": 5175
+    },
+    {
+      "epoch": 10.352,
+      "grad_norm": 2.325695753097534,
+      "learning_rate": 2e-05,
+      "loss": 0.06419559,
+      "step": 5176
+    },
+    {
+      "epoch": 10.354,
+      "grad_norm": 2.277292013168335,
+      "learning_rate": 2e-05,
+      "loss": 0.05193496,
+      "step": 5177
+    },
+    {
+      "epoch": 10.356,
+      "grad_norm": 1.4175512790679932,
+      "learning_rate": 2e-05,
+      "loss": 0.04061021,
+      "step": 5178
+    },
+    {
+      "epoch": 10.358,
+      "grad_norm": 2.847827434539795,
+      "learning_rate": 2e-05,
+      "loss": 0.05605824,
+      "step": 5179
+    },
+    {
+      "epoch": 10.36,
+      "grad_norm": 1.4422760009765625,
+      "learning_rate": 2e-05,
+      "loss": 0.05320713,
+      "step": 5180
+    },
+    {
+      "epoch": 10.362,
+      "grad_norm": 1.720396637916565,
+      "learning_rate": 2e-05,
+      "loss": 0.05881133,
+      "step": 5181
+    },
+    {
+      "epoch": 10.364,
+      "grad_norm": 1.4462811946868896,
+      "learning_rate": 2e-05,
+      "loss": 0.05044965,
+      "step": 5182
+    },
+    {
+      "epoch": 10.366,
+      "grad_norm": 1.1778342723846436,
+      "learning_rate": 2e-05,
+      "loss": 0.03916883,
+      "step": 5183
+    },
+    {
+      "epoch": 10.368,
+      "grad_norm": 0.9606398344039917,
+      "learning_rate": 2e-05,
+      "loss": 0.0260303,
+      "step": 5184
+    },
+    {
+      "epoch": 10.37,
+      "grad_norm": 3.025618553161621,
+      "learning_rate": 2e-05,
+      "loss": 0.07479943,
+      "step": 5185
+    },
+    {
+      "epoch": 10.372,
+      "grad_norm": 1.7541829347610474,
+      "learning_rate": 2e-05,
+      "loss": 0.05436843,
+      "step": 5186
+    },
+    {
+      "epoch": 10.374,
+      "grad_norm": 1.5163887739181519,
+      "learning_rate": 2e-05,
+      "loss": 0.06724942,
+      "step": 5187
+    },
+    {
+      "epoch": 10.376,
+      "grad_norm": 1.2254157066345215,
+      "learning_rate": 2e-05,
+      "loss": 0.03878136,
+      "step": 5188
+    },
+    {
+      "epoch": 10.378,
+      "grad_norm": 1.7440659999847412,
+      "learning_rate": 2e-05,
+      "loss": 0.04667617,
+      "step": 5189
+    },
+    {
+      "epoch": 10.38,
+      "grad_norm": 1.921148657798767,
+      "learning_rate": 2e-05,
+      "loss": 0.0511181,
+      "step": 5190
+    },
+    {
+      "epoch": 10.382,
+      "grad_norm": 2.249518871307373,
+      "learning_rate": 2e-05,
+      "loss": 0.05295843,
+      "step": 5191
+    },
+    {
+      "epoch": 10.384,
+      "grad_norm": 1.9301743507385254,
+      "learning_rate": 2e-05,
+      "loss": 0.05415469,
+      "step": 5192
+    },
+    {
+      "epoch": 10.386,
+      "grad_norm": 2.933361291885376,
+      "learning_rate": 2e-05,
+      "loss": 0.05823458,
+      "step": 5193
+    },
+    {
+      "epoch": 10.388,
+      "grad_norm": 1.8775449991226196,
+      "learning_rate": 2e-05,
+      "loss": 0.03668153,
+      "step": 5194
+    },
+    {
+      "epoch": 10.39,
+      "grad_norm": 1.2913427352905273,
+      "learning_rate": 2e-05,
+      "loss": 0.03676874,
+      "step": 5195
+    },
+    {
+      "epoch": 10.392,
+      "grad_norm": 3.2641937732696533,
+      "learning_rate": 2e-05,
+      "loss": 0.03394835,
+      "step": 5196
+    },
+    {
+      "epoch": 10.394,
+      "grad_norm": 1.5311733484268188,
+      "learning_rate": 2e-05,
+      "loss": 0.05097717,
+      "step": 5197
+    },
+    {
+      "epoch": 10.396,
+      "grad_norm": 1.488006830215454,
+      "learning_rate": 2e-05,
+      "loss": 0.06110777,
+      "step": 5198
+    },
+    {
+      "epoch": 10.398,
+      "grad_norm": 1.3041068315505981,
+      "learning_rate": 2e-05,
+      "loss": 0.04042555,
+      "step": 5199
+    },
+    {
+      "epoch": 10.4,
+      "grad_norm": 1.8183988332748413,
+      "learning_rate": 2e-05,
+      "loss": 0.04754098,
+      "step": 5200
+    },
+    {
+      "epoch": 10.402,
+      "grad_norm": 2.3569672107696533,
+      "learning_rate": 2e-05,
+      "loss": 0.04439374,
+      "step": 5201
+    },
+    {
+      "epoch": 10.404,
+      "grad_norm": 1.4490054845809937,
+      "learning_rate": 2e-05,
+      "loss": 0.0474748,
+      "step": 5202
+    },
+    {
+      "epoch": 10.406,
+      "grad_norm": 1.513960838317871,
+      "learning_rate": 2e-05,
+      "loss": 0.05048071,
+      "step": 5203
+    },
+    {
+      "epoch": 10.408,
+      "grad_norm": 1.6078133583068848,
+      "learning_rate": 2e-05,
+      "loss": 0.06171016,
+      "step": 5204
+    },
+    {
+      "epoch": 10.41,
+      "grad_norm": 2.1041479110717773,
+      "learning_rate": 2e-05,
+      "loss": 0.07458056,
+      "step": 5205
+    },
+    {
+      "epoch": 10.412,
+      "grad_norm": 1.2768281698226929,
+      "learning_rate": 2e-05,
+      "loss": 0.05074968,
+      "step": 5206
+    },
+    {
+      "epoch": 10.414,
+      "grad_norm": 1.2827049493789673,
+      "learning_rate": 2e-05,
+      "loss": 0.03884158,
+      "step": 5207
+    },
+    {
+      "epoch": 10.416,
+      "grad_norm": 1.3199437856674194,
+      "learning_rate": 2e-05,
+      "loss": 0.04091244,
+      "step": 5208
+    },
+    {
+      "epoch": 10.418,
+      "grad_norm": 1.4176160097122192,
+      "learning_rate": 2e-05,
+      "loss": 0.05431768,
+      "step": 5209
+    },
+    {
+      "epoch": 10.42,
+      "grad_norm": 1.8141944408416748,
+      "learning_rate": 2e-05,
+      "loss": 0.04813956,
+      "step": 5210
+    },
+    {
+      "epoch": 10.422,
+      "grad_norm": 1.3609751462936401,
+      "learning_rate": 2e-05,
+      "loss": 0.05530354,
+      "step": 5211
+    },
+    {
+      "epoch": 10.424,
+      "grad_norm": 1.780599594116211,
+      "learning_rate": 2e-05,
+      "loss": 0.0568007,
+      "step": 5212
+    },
+    {
+      "epoch": 10.426,
+      "grad_norm": 1.6819827556610107,
+      "learning_rate": 2e-05,
+      "loss": 0.03991935,
+      "step": 5213
+    },
+    {
+      "epoch": 10.428,
+      "grad_norm": 2.557753801345825,
+      "learning_rate": 2e-05,
+      "loss": 0.0696613,
+      "step": 5214
+    },
+    {
+      "epoch": 10.43,
+      "grad_norm": 1.8446578979492188,
+      "learning_rate": 2e-05,
+      "loss": 0.04753678,
+      "step": 5215
+    },
+    {
+      "epoch": 10.432,
+      "grad_norm": 0.939564049243927,
+      "learning_rate": 2e-05,
+      "loss": 0.02770604,
+      "step": 5216
+    },
+    {
+      "epoch": 10.434,
+      "grad_norm": 1.930539846420288,
+      "learning_rate": 2e-05,
+      "loss": 0.05070189,
+      "step": 5217
+    },
+    {
+      "epoch": 10.436,
+      "grad_norm": 1.480115532875061,
+      "learning_rate": 2e-05,
+      "loss": 0.03808409,
+      "step": 5218
+    },
+    {
+      "epoch": 10.438,
+      "grad_norm": 1.2075250148773193,
+      "learning_rate": 2e-05,
+      "loss": 0.03466361,
+      "step": 5219
+    },
+    {
+      "epoch": 10.44,
+      "grad_norm": 2.400930166244507,
+      "learning_rate": 2e-05,
+      "loss": 0.05821132,
+      "step": 5220
+    },
+    {
+      "epoch": 10.442,
+      "grad_norm": 1.5590664148330688,
+      "learning_rate": 2e-05,
+      "loss": 0.04925883,
+      "step": 5221
+    },
+    {
+      "epoch": 10.444,
+      "grad_norm": 1.9377199411392212,
+      "learning_rate": 2e-05,
+      "loss": 0.05342076,
+      "step": 5222
+    },
+    {
+      "epoch": 10.446,
+      "grad_norm": 1.9182746410369873,
+      "learning_rate": 2e-05,
+      "loss": 0.06361227,
+      "step": 5223
+    },
+    {
+      "epoch": 10.448,
+      "grad_norm": 1.3133718967437744,
+      "learning_rate": 2e-05,
+      "loss": 0.04293478,
+      "step": 5224
+    },
+    {
+      "epoch": 10.45,
+      "grad_norm": 1.6044734716415405,
+      "learning_rate": 2e-05,
+      "loss": 0.06643161,
+      "step": 5225
+    },
+    {
+      "epoch": 10.452,
+      "grad_norm": 1.3510301113128662,
+      "learning_rate": 2e-05,
+      "loss": 0.0424097,
+      "step": 5226
+    },
+    {
+      "epoch": 10.454,
+      "grad_norm": 2.926586389541626,
+      "learning_rate": 2e-05,
+      "loss": 0.05616639,
+      "step": 5227
+    },
+    {
+      "epoch": 10.456,
+      "grad_norm": 1.7717212438583374,
+      "learning_rate": 2e-05,
+      "loss": 0.04580856,
+      "step": 5228
+    },
+    {
+      "epoch": 10.458,
+      "grad_norm": 1.2072011232376099,
+      "learning_rate": 2e-05,
+      "loss": 0.05067045,
+      "step": 5229
+    },
+    {
+      "epoch": 10.46,
+      "grad_norm": 2.7207653522491455,
+      "learning_rate": 2e-05,
+      "loss": 0.05808522,
+      "step": 5230
+    },
+    {
+      "epoch": 10.462,
+      "grad_norm": 1.8373502492904663,
+      "learning_rate": 2e-05,
+      "loss": 0.0528776,
+      "step": 5231
+    },
+    {
+      "epoch": 10.464,
+      "grad_norm": 1.2759414911270142,
+      "learning_rate": 2e-05,
+      "loss": 0.04170848,
+      "step": 5232
+    },
+    {
+      "epoch": 10.466,
+      "grad_norm": 1.9643059968948364,
+      "learning_rate": 2e-05,
+      "loss": 0.04919844,
+      "step": 5233
+    },
+    {
+      "epoch": 10.468,
+      "grad_norm": 3.1205170154571533,
+      "learning_rate": 2e-05,
+      "loss": 0.06371839,
+      "step": 5234
+    },
+    {
+      "epoch": 10.47,
+      "grad_norm": 13.818527221679688,
+      "learning_rate": 2e-05,
+      "loss": 0.0568444,
+      "step": 5235
+    },
+    {
+      "epoch": 10.472,
+      "grad_norm": 5.399960517883301,
+      "learning_rate": 2e-05,
+      "loss": 0.04257676,
+      "step": 5236
+    },
+    {
+      "epoch": 10.474,
+      "grad_norm": 1.6575089693069458,
+      "learning_rate": 2e-05,
+      "loss": 0.05032737,
+      "step": 5237
+    },
+    {
+      "epoch": 10.475999999999999,
+      "grad_norm": 1.2673888206481934,
+      "learning_rate": 2e-05,
+      "loss": 0.04592997,
+      "step": 5238
+    },
+    {
+      "epoch": 10.478,
+      "grad_norm": 1.8720520734786987,
+      "learning_rate": 2e-05,
+      "loss": 0.05344952,
+      "step": 5239
+    },
+    {
+      "epoch": 10.48,
+      "grad_norm": 5.716723918914795,
+      "learning_rate": 2e-05,
+      "loss": 0.05336998,
+      "step": 5240
+    },
+    {
+      "epoch": 10.482,
+      "grad_norm": 1.8104397058486938,
+      "learning_rate": 2e-05,
+      "loss": 0.05812651,
+      "step": 5241
+    },
+    {
+      "epoch": 10.484,
+      "grad_norm": 1.6072144508361816,
+      "learning_rate": 2e-05,
+      "loss": 0.05022549,
+      "step": 5242
+    },
+    {
+      "epoch": 10.486,
+      "grad_norm": 1.7834798097610474,
+      "learning_rate": 2e-05,
+      "loss": 0.0548949,
+      "step": 5243
+    },
+    {
+      "epoch": 10.488,
+      "grad_norm": 1.4747883081436157,
+      "learning_rate": 2e-05,
+      "loss": 0.07916712,
+      "step": 5244
+    },
+    {
+      "epoch": 10.49,
+      "grad_norm": 1.4801729917526245,
+      "learning_rate": 2e-05,
+      "loss": 0.0434589,
+      "step": 5245
+    },
+    {
+      "epoch": 10.492,
+      "grad_norm": 1.3738034963607788,
+      "learning_rate": 2e-05,
+      "loss": 0.04868829,
+      "step": 5246
+    },
+    {
+      "epoch": 10.494,
+      "grad_norm": 1.1676543951034546,
+      "learning_rate": 2e-05,
+      "loss": 0.03874259,
+      "step": 5247
+    },
+    {
+      "epoch": 10.496,
+      "grad_norm": 1.4973816871643066,
+      "learning_rate": 2e-05,
+      "loss": 0.04599489,
+      "step": 5248
+    },
+    {
+      "epoch": 10.498,
+      "grad_norm": 1.5680755376815796,
+      "learning_rate": 2e-05,
+      "loss": 0.04904544,
+      "step": 5249
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 1.9003678560256958,
+      "learning_rate": 2e-05,
+      "loss": 0.07093397,
+      "step": 5250
+    },
+    {
+      "epoch": 10.502,
+      "grad_norm": 1.6360976696014404,
+      "learning_rate": 2e-05,
+      "loss": 0.04844421,
+      "step": 5251
+    },
+    {
+      "epoch": 10.504,
+      "grad_norm": 1.8371424674987793,
+      "learning_rate": 2e-05,
+      "loss": 0.0682577,
+      "step": 5252
+    },
+    {
+      "epoch": 10.506,
+      "grad_norm": 2.010197401046753,
+      "learning_rate": 2e-05,
+      "loss": 0.04919277,
+      "step": 5253
+    },
+    {
+      "epoch": 10.508,
+      "grad_norm": 1.8417555093765259,
+      "learning_rate": 2e-05,
+      "loss": 0.04783285,
+      "step": 5254
+    },
+    {
+      "epoch": 10.51,
+      "grad_norm": 1.6858887672424316,
+      "learning_rate": 2e-05,
+      "loss": 0.05456799,
+      "step": 5255
+    },
+    {
+      "epoch": 10.512,
+      "grad_norm": 3.253695011138916,
+      "learning_rate": 2e-05,
+      "loss": 0.03823632,
+      "step": 5256
+    },
+    {
+      "epoch": 10.514,
+      "grad_norm": 1.8431847095489502,
+      "learning_rate": 2e-05,
+      "loss": 0.06339058,
+      "step": 5257
+    },
+    {
+      "epoch": 10.516,
+      "grad_norm": 1.3688884973526,
+      "learning_rate": 2e-05,
+      "loss": 0.06101316,
+      "step": 5258
+    },
+    {
+      "epoch": 10.518,
+      "grad_norm": 0.9589589834213257,
+      "learning_rate": 2e-05,
+      "loss": 0.03332622,
+      "step": 5259
+    },
+    {
+      "epoch": 10.52,
+      "grad_norm": 1.5747522115707397,
+      "learning_rate": 2e-05,
+      "loss": 0.05914874,
+      "step": 5260
+    },
+    {
+      "epoch": 10.522,
+      "grad_norm": 1.884902834892273,
+      "learning_rate": 2e-05,
+      "loss": 0.07614301,
+      "step": 5261
+    },
+    {
+      "epoch": 10.524000000000001,
+      "grad_norm": 1.4136667251586914,
+      "learning_rate": 2e-05,
+      "loss": 0.05749642,
+      "step": 5262
+    },
+    {
+      "epoch": 10.526,
+      "grad_norm": 1.4973257780075073,
+      "learning_rate": 2e-05,
+      "loss": 0.05710443,
+      "step": 5263
+    },
+    {
+      "epoch": 10.528,
+      "grad_norm": 2.1103341579437256,
+      "learning_rate": 2e-05,
+      "loss": 0.06917971,
+      "step": 5264
+    },
+    {
+      "epoch": 10.53,
+      "grad_norm": 1.6362884044647217,
+      "learning_rate": 2e-05,
+      "loss": 0.04992104,
+      "step": 5265
+    },
+    {
+      "epoch": 10.532,
+      "grad_norm": 2.8185596466064453,
+      "learning_rate": 2e-05,
+      "loss": 0.05004818,
+      "step": 5266
+    },
+    {
+      "epoch": 10.534,
+      "grad_norm": 1.6403809785842896,
+      "learning_rate": 2e-05,
+      "loss": 0.05555485,
+      "step": 5267
+    },
+    {
+      "epoch": 10.536,
+      "grad_norm": 2.137666702270508,
+      "learning_rate": 2e-05,
+      "loss": 0.05246295,
+      "step": 5268
+    },
+    {
+      "epoch": 10.538,
+      "grad_norm": 1.6086887121200562,
+      "learning_rate": 2e-05,
+      "loss": 0.04962714,
+      "step": 5269
+    },
+    {
+      "epoch": 10.54,
+      "grad_norm": 1.4475809335708618,
+      "learning_rate": 2e-05,
+      "loss": 0.0453802,
+      "step": 5270
+    },
+    {
+      "epoch": 10.542,
+      "grad_norm": 1.6052064895629883,
+      "learning_rate": 2e-05,
+      "loss": 0.04843373,
+      "step": 5271
+    },
+    {
+      "epoch": 10.544,
+      "grad_norm": 1.8348846435546875,
+      "learning_rate": 2e-05,
+      "loss": 0.05354032,
+      "step": 5272
+    },
+    {
+      "epoch": 10.546,
+      "grad_norm": 1.5598926544189453,
+      "learning_rate": 2e-05,
+      "loss": 0.04522219,
+      "step": 5273
+    },
+    {
+      "epoch": 10.548,
+      "grad_norm": 1.7176995277404785,
+      "learning_rate": 2e-05,
+      "loss": 0.06097797,
+      "step": 5274
+    },
+    {
+      "epoch": 10.55,
+      "grad_norm": 1.6066391468048096,
+      "learning_rate": 2e-05,
+      "loss": 0.04459933,
+      "step": 5275
+    },
+    {
+      "epoch": 10.552,
+      "grad_norm": 1.2835681438446045,
+      "learning_rate": 2e-05,
+      "loss": 0.04766644,
+      "step": 5276
+    },
+    {
+      "epoch": 10.554,
+      "grad_norm": 1.5211418867111206,
+      "learning_rate": 2e-05,
+      "loss": 0.05268958,
+      "step": 5277
+    },
+    {
+      "epoch": 10.556000000000001,
+      "grad_norm": 1.3323153257369995,
+      "learning_rate": 2e-05,
+      "loss": 0.04630678,
+      "step": 5278
+    },
+    {
+      "epoch": 10.558,
+      "grad_norm": 2.196124315261841,
+      "learning_rate": 2e-05,
+      "loss": 0.04796411,
+      "step": 5279
+    },
+    {
+      "epoch": 10.56,
+      "grad_norm": 6.703586101531982,
+      "learning_rate": 2e-05,
+      "loss": 0.08236083,
+      "step": 5280
+    },
+    {
+      "epoch": 10.562,
+      "grad_norm": 2.39101505279541,
+      "learning_rate": 2e-05,
+      "loss": 0.0583214,
+      "step": 5281
+    },
+    {
+      "epoch": 10.564,
+      "grad_norm": 1.674933671951294,
+      "learning_rate": 2e-05,
+      "loss": 0.05167703,
+      "step": 5282
+    },
+    {
+      "epoch": 10.566,
+      "grad_norm": 1.468446135520935,
+      "learning_rate": 2e-05,
+      "loss": 0.04566929,
+      "step": 5283
+    },
+    {
+      "epoch": 10.568,
+      "grad_norm": 1.4748491048812866,
+      "learning_rate": 2e-05,
+      "loss": 0.04062827,
+      "step": 5284
+    },
+    {
+      "epoch": 10.57,
+      "grad_norm": 2.5267021656036377,
+      "learning_rate": 2e-05,
+      "loss": 0.05230495,
+      "step": 5285
+    },
+    {
+      "epoch": 10.572,
+      "grad_norm": 2.1810038089752197,
+      "learning_rate": 2e-05,
+      "loss": 0.04218719,
+      "step": 5286
+    },
+    {
+      "epoch": 10.574,
+      "grad_norm": 1.3321385383605957,
+      "learning_rate": 2e-05,
+      "loss": 0.04871097,
+      "step": 5287
+    },
+    {
+      "epoch": 10.576,
+      "grad_norm": 1.7109707593917847,
+      "learning_rate": 2e-05,
+      "loss": 0.04826786,
+      "step": 5288
+    },
+    {
+      "epoch": 10.578,
+      "grad_norm": 1.9152201414108276,
+      "learning_rate": 2e-05,
+      "loss": 0.04686288,
+      "step": 5289
+    },
+    {
+      "epoch": 10.58,
+      "grad_norm": 5.031193256378174,
+      "learning_rate": 2e-05,
+      "loss": 0.04337011,
+      "step": 5290
+    },
+    {
+      "epoch": 10.582,
+      "grad_norm": 3.426795482635498,
+      "learning_rate": 2e-05,
+      "loss": 0.03990654,
+      "step": 5291
+    },
+    {
+      "epoch": 10.584,
+      "grad_norm": 2.406569719314575,
+      "learning_rate": 2e-05,
+      "loss": 0.03668533,
+      "step": 5292
+    },
+    {
+      "epoch": 10.586,
+      "grad_norm": 3.1804933547973633,
+      "learning_rate": 2e-05,
+      "loss": 0.06070712,
+      "step": 5293
+    },
+    {
+      "epoch": 10.588,
+      "grad_norm": 2.061739206314087,
+      "learning_rate": 2e-05,
+      "loss": 0.02787476,
+      "step": 5294
+    },
+    {
+      "epoch": 10.59,
+      "grad_norm": 1.6928627490997314,
+      "learning_rate": 2e-05,
+      "loss": 0.04576159,
+      "step": 5295
+    },
+    {
+      "epoch": 10.592,
+      "grad_norm": 2.2229623794555664,
+      "learning_rate": 2e-05,
+      "loss": 0.05602458,
+      "step": 5296
+    },
+    {
+      "epoch": 10.594,
+      "grad_norm": 1.289467692375183,
+      "learning_rate": 2e-05,
+      "loss": 0.05179123,
+      "step": 5297
+    },
+    {
+      "epoch": 10.596,
+      "grad_norm": 1.5103563070297241,
+      "learning_rate": 2e-05,
+      "loss": 0.05836169,
+      "step": 5298
+    },
+    {
+      "epoch": 10.598,
+      "grad_norm": 1.6638998985290527,
+      "learning_rate": 2e-05,
+      "loss": 0.04362776,
+      "step": 5299
+    },
+    {
+      "epoch": 10.6,
+      "grad_norm": 1.3501155376434326,
+      "learning_rate": 2e-05,
+      "loss": 0.04407869,
+      "step": 5300
+    },
+    {
+      "epoch": 10.602,
+      "grad_norm": 1.708865761756897,
+      "learning_rate": 2e-05,
+      "loss": 0.05075856,
+      "step": 5301
+    },
+    {
+      "epoch": 10.604,
+      "grad_norm": 10.003942489624023,
+      "learning_rate": 2e-05,
+      "loss": 0.07709937,
+      "step": 5302
+    },
+    {
+      "epoch": 10.606,
+      "grad_norm": 1.8898708820343018,
+      "learning_rate": 2e-05,
+      "loss": 0.03913818,
+      "step": 5303
+    },
+    {
+      "epoch": 10.608,
+      "grad_norm": 1.95207679271698,
+      "learning_rate": 2e-05,
+      "loss": 0.05876492,
+      "step": 5304
+    },
+    {
+      "epoch": 10.61,
+      "grad_norm": 1.6811554431915283,
+      "learning_rate": 2e-05,
+      "loss": 0.03337777,
+      "step": 5305
+    },
+    {
+      "epoch": 10.612,
+      "grad_norm": 2.222196102142334,
+      "learning_rate": 2e-05,
+      "loss": 0.0575141,
+      "step": 5306
+    },
+    {
+      "epoch": 10.614,
+      "grad_norm": 2.2834959030151367,
+      "learning_rate": 2e-05,
+      "loss": 0.05859252,
+      "step": 5307
+    },
+    {
+      "epoch": 10.616,
+      "grad_norm": 2.21087646484375,
+      "learning_rate": 2e-05,
+      "loss": 0.05571397,
+      "step": 5308
+    },
+    {
+      "epoch": 10.618,
+      "grad_norm": 3.0877723693847656,
+      "learning_rate": 2e-05,
+      "loss": 0.10596378,
+      "step": 5309
+    },
+    {
+      "epoch": 10.62,
+      "grad_norm": 3.603444814682007,
+      "learning_rate": 2e-05,
+      "loss": 0.06696412,
+      "step": 5310
+    },
+    {
+      "epoch": 10.622,
+      "grad_norm": 1.8475040197372437,
+      "learning_rate": 2e-05,
+      "loss": 0.04399508,
+      "step": 5311
+    },
+    {
+      "epoch": 10.624,
+      "grad_norm": 1.4669568538665771,
+      "learning_rate": 2e-05,
+      "loss": 0.0431044,
+      "step": 5312
+    },
+    {
+      "epoch": 10.626,
+      "grad_norm": 1.3643040657043457,
+      "learning_rate": 2e-05,
+      "loss": 0.0470515,
+      "step": 5313
+    },
+    {
+      "epoch": 10.628,
+      "grad_norm": 1.2287578582763672,
+      "learning_rate": 2e-05,
+      "loss": 0.05811661,
+      "step": 5314
+    },
+    {
+      "epoch": 10.63,
+      "grad_norm": 1.0902774333953857,
+      "learning_rate": 2e-05,
+      "loss": 0.03484159,
+      "step": 5315
+    },
+    {
+      "epoch": 10.632,
+      "grad_norm": 1.6369050741195679,
+      "learning_rate": 2e-05,
+      "loss": 0.04110289,
+      "step": 5316
+    },
+    {
+      "epoch": 10.634,
+      "grad_norm": 1.275800108909607,
+      "learning_rate": 2e-05,
+      "loss": 0.05027898,
+      "step": 5317
+    },
+    {
+      "epoch": 10.636,
+      "grad_norm": 1.4583206176757812,
+      "learning_rate": 2e-05,
+      "loss": 0.03882696,
+      "step": 5318
+    },
+    {
+      "epoch": 10.638,
+      "grad_norm": 1.3193122148513794,
+      "learning_rate": 2e-05,
+      "loss": 0.03158198,
+      "step": 5319
+    },
+    {
+      "epoch": 10.64,
+      "grad_norm": 1.4229142665863037,
+      "learning_rate": 2e-05,
+      "loss": 0.04588968,
+      "step": 5320
+    },
+    {
+      "epoch": 10.642,
+      "grad_norm": 1.5470317602157593,
+      "learning_rate": 2e-05,
+      "loss": 0.04177276,
+      "step": 5321
+    },
+    {
+      "epoch": 10.644,
+      "grad_norm": 1.322806477546692,
+      "learning_rate": 2e-05,
+      "loss": 0.04808362,
+      "step": 5322
+    },
+    {
+      "epoch": 10.646,
+      "grad_norm": 1.7866078615188599,
+      "learning_rate": 2e-05,
+      "loss": 0.04570613,
+      "step": 5323
+    },
+    {
+      "epoch": 10.648,
+      "grad_norm": 1.8436026573181152,
+      "learning_rate": 2e-05,
+      "loss": 0.06046174,
+      "step": 5324
+    },
+    {
+      "epoch": 10.65,
+      "grad_norm": 1.6099350452423096,
+      "learning_rate": 2e-05,
+      "loss": 0.05201571,
+      "step": 5325
+    },
+    {
+      "epoch": 10.652,
+      "grad_norm": 1.4118268489837646,
+      "learning_rate": 2e-05,
+      "loss": 0.04571307,
+      "step": 5326
+    },
+    {
+      "epoch": 10.654,
+      "grad_norm": 1.7007311582565308,
+      "learning_rate": 2e-05,
+      "loss": 0.05410545,
+      "step": 5327
+    },
+    {
+      "epoch": 10.656,
+      "grad_norm": 1.3747484683990479,
+      "learning_rate": 2e-05,
+      "loss": 0.04396842,
+      "step": 5328
+    },
+    {
+      "epoch": 10.658,
+      "grad_norm": 1.5627105236053467,
+      "learning_rate": 2e-05,
+      "loss": 0.05439685,
+      "step": 5329
+    },
+    {
+      "epoch": 10.66,
+      "grad_norm": 1.6469447612762451,
+      "learning_rate": 2e-05,
+      "loss": 0.05709826,
+      "step": 5330
+    },
+    {
+      "epoch": 10.662,
+      "grad_norm": 3.3534231185913086,
+      "learning_rate": 2e-05,
+      "loss": 0.05348397,
+      "step": 5331
+    },
+    {
+      "epoch": 10.664,
+      "grad_norm": 1.3663334846496582,
+      "learning_rate": 2e-05,
+      "loss": 0.04078261,
+      "step": 5332
+    },
+    {
+      "epoch": 10.666,
+      "grad_norm": 1.6048120260238647,
+      "learning_rate": 2e-05,
+      "loss": 0.05477417,
+      "step": 5333
+    },
+    {
+      "epoch": 10.668,
+      "grad_norm": 1.6299818754196167,
+      "learning_rate": 2e-05,
+      "loss": 0.05402359,
+      "step": 5334
+    },
+    {
+      "epoch": 10.67,
+      "grad_norm": 1.9940632581710815,
+      "learning_rate": 2e-05,
+      "loss": 0.05533578,
+      "step": 5335
+    },
+    {
+      "epoch": 10.672,
+      "grad_norm": 1.1183878183364868,
+      "learning_rate": 2e-05,
+      "loss": 0.03753264,
+      "step": 5336
+    },
+    {
+      "epoch": 10.674,
+      "grad_norm": 1.5522422790527344,
+      "learning_rate": 2e-05,
+      "loss": 0.05828452,
+      "step": 5337
+    },
+    {
+      "epoch": 10.676,
+      "grad_norm": 1.3103322982788086,
+      "learning_rate": 2e-05,
+      "loss": 0.04702647,
+      "step": 5338
+    },
+    {
+      "epoch": 10.678,
+      "grad_norm": 1.2099779844284058,
+      "learning_rate": 2e-05,
+      "loss": 0.04113594,
+      "step": 5339
+    },
+    {
+      "epoch": 10.68,
+      "grad_norm": 3.985682964324951,
+      "learning_rate": 2e-05,
+      "loss": 0.07098438,
+      "step": 5340
+    },
+    {
+      "epoch": 10.682,
+      "grad_norm": 1.8771806955337524,
+      "learning_rate": 2e-05,
+      "loss": 0.04530764,
+      "step": 5341
+    },
+    {
+      "epoch": 10.684,
+      "grad_norm": 2.1567442417144775,
+      "learning_rate": 2e-05,
+      "loss": 0.04229917,
+      "step": 5342
+    },
+    {
+      "epoch": 10.686,
+      "grad_norm": 1.6238269805908203,
+      "learning_rate": 2e-05,
+      "loss": 0.04225535,
+      "step": 5343
+    },
+    {
+      "epoch": 10.688,
+      "grad_norm": 1.869265079498291,
+      "learning_rate": 2e-05,
+      "loss": 0.05382503,
+      "step": 5344
+    },
+    {
+      "epoch": 10.69,
+      "grad_norm": 1.8355045318603516,
+      "learning_rate": 2e-05,
+      "loss": 0.0423015,
+      "step": 5345
+    },
+    {
+      "epoch": 10.692,
+      "grad_norm": 1.3736673593521118,
+      "learning_rate": 2e-05,
+      "loss": 0.04386007,
+      "step": 5346
+    },
+    {
+      "epoch": 10.693999999999999,
+      "grad_norm": 2.0322253704071045,
+      "learning_rate": 2e-05,
+      "loss": 0.04041972,
+      "step": 5347
+    },
+    {
+      "epoch": 10.696,
+      "grad_norm": 2.336951732635498,
+      "learning_rate": 2e-05,
+      "loss": 0.05086974,
+      "step": 5348
+    },
+    {
+      "epoch": 10.698,
+      "grad_norm": 2.0457851886749268,
+      "learning_rate": 2e-05,
+      "loss": 0.05006742,
+      "step": 5349
+    },
+    {
+      "epoch": 10.7,
+      "grad_norm": 1.992970585823059,
+      "learning_rate": 2e-05,
+      "loss": 0.06358136,
+      "step": 5350
+    },
+    {
+      "epoch": 10.702,
+      "grad_norm": 1.2050204277038574,
+      "learning_rate": 2e-05,
+      "loss": 0.04032322,
+      "step": 5351
+    },
+    {
+      "epoch": 10.704,
+      "grad_norm": 1.4412949085235596,
+      "learning_rate": 2e-05,
+      "loss": 0.04633742,
+      "step": 5352
+    },
+    {
+      "epoch": 10.706,
+      "grad_norm": 1.5254440307617188,
+      "learning_rate": 2e-05,
+      "loss": 0.05168087,
+      "step": 5353
+    },
+    {
+      "epoch": 10.708,
+      "grad_norm": 1.4087555408477783,
+      "learning_rate": 2e-05,
+      "loss": 0.05477347,
+      "step": 5354
+    },
+    {
+      "epoch": 10.71,
+      "grad_norm": 1.431321382522583,
+      "learning_rate": 2e-05,
+      "loss": 0.03768836,
+      "step": 5355
+    },
+    {
+      "epoch": 10.712,
+      "grad_norm": 1.1506340503692627,
+      "learning_rate": 2e-05,
+      "loss": 0.04004525,
+      "step": 5356
+    },
+    {
+      "epoch": 10.714,
+      "grad_norm": 2.427537441253662,
+      "learning_rate": 2e-05,
+      "loss": 0.06314078,
+      "step": 5357
+    },
+    {
+      "epoch": 10.716,
+      "grad_norm": 1.5225038528442383,
+      "learning_rate": 2e-05,
+      "loss": 0.05141336,
+      "step": 5358
+    },
+    {
+      "epoch": 10.718,
+      "grad_norm": 3.697816848754883,
+      "learning_rate": 2e-05,
+      "loss": 0.05468082,
+      "step": 5359
+    },
+    {
+      "epoch": 10.72,
+      "grad_norm": 1.5800566673278809,
+      "learning_rate": 2e-05,
+      "loss": 0.058851,
+      "step": 5360
+    },
+    {
+      "epoch": 10.722,
+      "grad_norm": 0.8254619240760803,
+      "learning_rate": 2e-05,
+      "loss": 0.02738049,
+      "step": 5361
+    },
+    {
+      "epoch": 10.724,
+      "grad_norm": 1.369780421257019,
+      "learning_rate": 2e-05,
+      "loss": 0.04381582,
+      "step": 5362
+    },
+    {
+      "epoch": 10.725999999999999,
+      "grad_norm": 1.2137796878814697,
+      "learning_rate": 2e-05,
+      "loss": 0.0480298,
+      "step": 5363
+    },
+    {
+      "epoch": 10.728,
+      "grad_norm": 1.4857935905456543,
+      "learning_rate": 2e-05,
+      "loss": 0.03931804,
+      "step": 5364
+    },
+    {
+      "epoch": 10.73,
+      "grad_norm": 1.4152106046676636,
+      "learning_rate": 2e-05,
+      "loss": 0.03949886,
+      "step": 5365
+    },
+    {
+      "epoch": 10.732,
+      "grad_norm": 1.3397774696350098,
+      "learning_rate": 2e-05,
+      "loss": 0.05219603,
+      "step": 5366
+    },
+    {
+      "epoch": 10.734,
+      "grad_norm": 1.2872146368026733,
+      "learning_rate": 2e-05,
+      "loss": 0.042691,
+      "step": 5367
+    },
+    {
+      "epoch": 10.736,
+      "grad_norm": 2.1003661155700684,
+      "learning_rate": 2e-05,
+      "loss": 0.05165257,
+      "step": 5368
+    },
+    {
+      "epoch": 10.738,
+      "grad_norm": 2.007760524749756,
+      "learning_rate": 2e-05,
+      "loss": 0.06440516,
+      "step": 5369
+    },
+    {
+      "epoch": 10.74,
+      "grad_norm": 1.6404235363006592,
+      "learning_rate": 2e-05,
+      "loss": 0.05111123,
+      "step": 5370
+    },
+    {
+      "epoch": 10.742,
+      "grad_norm": 2.8288562297821045,
+      "learning_rate": 2e-05,
+      "loss": 0.05282591,
+      "step": 5371
+    },
+    {
+      "epoch": 10.744,
+      "grad_norm": 2.6334850788116455,
+      "learning_rate": 2e-05,
+      "loss": 0.06181185,
+      "step": 5372
+    },
+    {
+      "epoch": 10.746,
+      "grad_norm": 1.4054573774337769,
+      "learning_rate": 2e-05,
+      "loss": 0.05052005,
+      "step": 5373
+    },
+    {
+      "epoch": 10.748,
+      "grad_norm": 1.4174977540969849,
+      "learning_rate": 2e-05,
+      "loss": 0.04603645,
+      "step": 5374
+    },
+    {
+      "epoch": 10.75,
+      "grad_norm": 1.3553341627120972,
+      "learning_rate": 2e-05,
+      "loss": 0.04983129,
+      "step": 5375
+    },
+    {
+      "epoch": 10.752,
+      "grad_norm": 2.03845477104187,
+      "learning_rate": 2e-05,
+      "loss": 0.06038283,
+      "step": 5376
+    },
+    {
+      "epoch": 10.754,
+      "grad_norm": 2.294628858566284,
+      "learning_rate": 2e-05,
+      "loss": 0.04681225,
+      "step": 5377
+    },
+    {
+      "epoch": 10.756,
+      "grad_norm": 1.303605079650879,
+      "learning_rate": 2e-05,
+      "loss": 0.04528525,
+      "step": 5378
+    },
+    {
+      "epoch": 10.758,
+      "grad_norm": 1.6137866973876953,
+      "learning_rate": 2e-05,
+      "loss": 0.0557767,
+      "step": 5379
+    },
+    {
+      "epoch": 10.76,
+      "grad_norm": 1.5829466581344604,
+      "learning_rate": 2e-05,
+      "loss": 0.04809469,
+      "step": 5380
+    },
+    {
+      "epoch": 10.762,
+      "grad_norm": 1.2748074531555176,
+      "learning_rate": 2e-05,
+      "loss": 0.04331703,
+      "step": 5381
+    },
+    {
+      "epoch": 10.764,
+      "grad_norm": 1.5433510541915894,
+      "learning_rate": 2e-05,
+      "loss": 0.06725293,
+      "step": 5382
+    },
+    {
+      "epoch": 10.766,
+      "grad_norm": 1.6846797466278076,
+      "learning_rate": 2e-05,
+      "loss": 0.04823402,
+      "step": 5383
+    },
+    {
+      "epoch": 10.768,
+      "grad_norm": 2.141554355621338,
+      "learning_rate": 2e-05,
+      "loss": 0.06744844,
+      "step": 5384
+    },
+    {
+      "epoch": 10.77,
+      "grad_norm": 1.4289065599441528,
+      "learning_rate": 2e-05,
+      "loss": 0.05341191,
+      "step": 5385
+    },
+    {
+      "epoch": 10.772,
+      "grad_norm": 1.3509269952774048,
+      "learning_rate": 2e-05,
+      "loss": 0.05020216,
+      "step": 5386
+    },
+    {
+      "epoch": 10.774000000000001,
+      "grad_norm": 2.071425676345825,
+      "learning_rate": 2e-05,
+      "loss": 0.06101374,
+      "step": 5387
+    },
+    {
+      "epoch": 10.776,
+      "grad_norm": 1.1163301467895508,
+      "learning_rate": 2e-05,
+      "loss": 0.04933468,
+      "step": 5388
+    },
+    {
+      "epoch": 10.778,
+      "grad_norm": 1.2288962602615356,
+      "learning_rate": 2e-05,
+      "loss": 0.04980216,
+      "step": 5389
+    },
+    {
+      "epoch": 10.78,
+      "grad_norm": 1.9011704921722412,
+      "learning_rate": 2e-05,
+      "loss": 0.05300172,
+      "step": 5390
+    },
+    {
+      "epoch": 10.782,
+      "grad_norm": 1.5152117013931274,
+      "learning_rate": 2e-05,
+      "loss": 0.04199405,
+      "step": 5391
+    },
+    {
+      "epoch": 10.784,
+      "grad_norm": 1.2738306522369385,
+      "learning_rate": 2e-05,
+      "loss": 0.04918165,
+      "step": 5392
+    },
+    {
+      "epoch": 10.786,
+      "grad_norm": 1.598197102546692,
+      "learning_rate": 2e-05,
+      "loss": 0.05998203,
+      "step": 5393
+    },
+    {
+      "epoch": 10.788,
+      "grad_norm": 1.4320733547210693,
+      "learning_rate": 2e-05,
+      "loss": 0.04948084,
+      "step": 5394
+    },
+    {
+      "epoch": 10.79,
+      "grad_norm": 1.6892844438552856,
+      "learning_rate": 2e-05,
+      "loss": 0.06663223,
+      "step": 5395
+    },
+    {
+      "epoch": 10.792,
+      "grad_norm": 1.3951140642166138,
+      "learning_rate": 2e-05,
+      "loss": 0.05481715,
+      "step": 5396
+    },
+    {
+      "epoch": 10.794,
+      "grad_norm": 1.895914077758789,
+      "learning_rate": 2e-05,
+      "loss": 0.050419,
+      "step": 5397
+    },
+    {
+      "epoch": 10.796,
+      "grad_norm": 1.6895439624786377,
+      "learning_rate": 2e-05,
+      "loss": 0.05903822,
+      "step": 5398
+    },
+    {
+      "epoch": 10.798,
+      "grad_norm": 1.3120181560516357,
+      "learning_rate": 2e-05,
+      "loss": 0.03543907,
+      "step": 5399
+    },
+    {
+      "epoch": 10.8,
+      "grad_norm": 1.598132610321045,
+      "learning_rate": 2e-05,
+      "loss": 0.03839812,
+      "step": 5400
+    },
+    {
+      "epoch": 10.802,
+      "grad_norm": 1.5897235870361328,
+      "learning_rate": 2e-05,
+      "loss": 0.04522756,
+      "step": 5401
+    },
+    {
+      "epoch": 10.804,
+      "grad_norm": 1.8060530424118042,
+      "learning_rate": 2e-05,
+      "loss": 0.06058746,
+      "step": 5402
+    },
+    {
+      "epoch": 10.806000000000001,
+      "grad_norm": 1.4628931283950806,
+      "learning_rate": 2e-05,
+      "loss": 0.03918256,
+      "step": 5403
+    },
+    {
+      "epoch": 10.808,
+      "grad_norm": 1.6146814823150635,
+      "learning_rate": 2e-05,
+      "loss": 0.05647987,
+      "step": 5404
+    },
+    {
+      "epoch": 10.81,
+      "grad_norm": 2.082378387451172,
+      "learning_rate": 2e-05,
+      "loss": 0.04440005,
+      "step": 5405
+    },
+    {
+      "epoch": 10.812,
+      "grad_norm": 2.013496160507202,
+      "learning_rate": 2e-05,
+      "loss": 0.06099815,
+      "step": 5406
+    },
+    {
+      "epoch": 10.814,
+      "grad_norm": 2.267378807067871,
+      "learning_rate": 2e-05,
+      "loss": 0.06538484,
+      "step": 5407
+    },
+    {
+      "epoch": 10.816,
+      "grad_norm": 1.9403990507125854,
+      "learning_rate": 2e-05,
+      "loss": 0.05472123,
+      "step": 5408
+    },
+    {
+      "epoch": 10.818,
+      "grad_norm": 1.1750283241271973,
+      "learning_rate": 2e-05,
+      "loss": 0.03978173,
+      "step": 5409
+    },
+    {
+      "epoch": 10.82,
+      "grad_norm": 1.5307939052581787,
+      "learning_rate": 2e-05,
+      "loss": 0.06439765,
+      "step": 5410
+    },
+    {
+      "epoch": 10.822,
+      "grad_norm": 1.0061407089233398,
+      "learning_rate": 2e-05,
+      "loss": 0.03479624,
+      "step": 5411
+    },
+    {
+      "epoch": 10.824,
+      "grad_norm": 1.3248248100280762,
+      "learning_rate": 2e-05,
+      "loss": 0.04627767,
+      "step": 5412
+    },
+    {
+      "epoch": 10.826,
+      "grad_norm": 2.050506114959717,
+      "learning_rate": 2e-05,
+      "loss": 0.03434724,
+      "step": 5413
+    },
+    {
+      "epoch": 10.828,
+      "grad_norm": 1.4149460792541504,
+      "learning_rate": 2e-05,
+      "loss": 0.05991215,
+      "step": 5414
+    },
+    {
+      "epoch": 10.83,
+      "grad_norm": 1.5017156600952148,
+      "learning_rate": 2e-05,
+      "loss": 0.05083961,
+      "step": 5415
+    },
+    {
+      "epoch": 10.832,
+      "grad_norm": 1.0457175970077515,
+      "learning_rate": 2e-05,
+      "loss": 0.03952884,
+      "step": 5416
+    },
+    {
+      "epoch": 10.834,
+      "grad_norm": 1.5834999084472656,
+      "learning_rate": 2e-05,
+      "loss": 0.05933148,
+      "step": 5417
+    },
+    {
+      "epoch": 10.836,
+      "grad_norm": 1.1834666728973389,
+      "learning_rate": 2e-05,
+      "loss": 0.03205921,
+      "step": 5418
+    },
+    {
+      "epoch": 10.838,
+      "grad_norm": 1.2450237274169922,
+      "learning_rate": 2e-05,
+      "loss": 0.04426446,
+      "step": 5419
+    },
+    {
+      "epoch": 10.84,
+      "grad_norm": 1.2155392169952393,
+      "learning_rate": 2e-05,
+      "loss": 0.03819549,
+      "step": 5420
+    },
+    {
+      "epoch": 10.842,
+      "grad_norm": 1.4611936807632446,
+      "learning_rate": 2e-05,
+      "loss": 0.04256415,
+      "step": 5421
+    },
+    {
+      "epoch": 10.844,
+      "grad_norm": 0.9903550744056702,
+      "learning_rate": 2e-05,
+      "loss": 0.03737554,
+      "step": 5422
+    },
+    {
+      "epoch": 10.846,
+      "grad_norm": 1.4878544807434082,
+      "learning_rate": 2e-05,
+      "loss": 0.04571939,
+      "step": 5423
+    },
+    {
+      "epoch": 10.848,
+      "grad_norm": 3.3160300254821777,
+      "learning_rate": 2e-05,
+      "loss": 0.06830984,
+      "step": 5424
+    },
+    {
+      "epoch": 10.85,
+      "grad_norm": 1.3133734464645386,
+      "learning_rate": 2e-05,
+      "loss": 0.05649231,
+      "step": 5425
+    },
+    {
+      "epoch": 10.852,
+      "grad_norm": 1.1060919761657715,
+      "learning_rate": 2e-05,
+      "loss": 0.03377821,
+      "step": 5426
+    },
+    {
+      "epoch": 10.854,
+      "grad_norm": 1.686246395111084,
+      "learning_rate": 2e-05,
+      "loss": 0.04799533,
+      "step": 5427
+    },
+    {
+      "epoch": 10.856,
+      "grad_norm": 2.3602640628814697,
+      "learning_rate": 2e-05,
+      "loss": 0.03780733,
+      "step": 5428
+    },
+    {
+      "epoch": 10.858,
+      "grad_norm": 1.5096981525421143,
+      "learning_rate": 2e-05,
+      "loss": 0.04364054,
+      "step": 5429
+    },
+    {
+      "epoch": 10.86,
+      "grad_norm": 3.4624457359313965,
+      "learning_rate": 2e-05,
+      "loss": 0.07765651,
+      "step": 5430
+    },
+    {
+      "epoch": 10.862,
+      "grad_norm": 1.9215983152389526,
+      "learning_rate": 2e-05,
+      "loss": 0.0564341,
+      "step": 5431
+    },
+    {
+      "epoch": 10.864,
+      "grad_norm": 2.2769806385040283,
+      "learning_rate": 2e-05,
+      "loss": 0.0673317,
+      "step": 5432
+    },
+    {
+      "epoch": 10.866,
+      "grad_norm": 1.4568067789077759,
+      "learning_rate": 2e-05,
+      "loss": 0.04599699,
+      "step": 5433
+    },
+    {
+      "epoch": 10.868,
+      "grad_norm": 1.3727821111679077,
+      "learning_rate": 2e-05,
+      "loss": 0.05360325,
+      "step": 5434
+    },
+    {
+      "epoch": 10.87,
+      "grad_norm": 2.1646981239318848,
+      "learning_rate": 2e-05,
+      "loss": 0.05432039,
+      "step": 5435
+    },
+    {
+      "epoch": 10.872,
+      "grad_norm": 1.466217041015625,
+      "learning_rate": 2e-05,
+      "loss": 0.05962711,
+      "step": 5436
+    },
+    {
+      "epoch": 10.874,
+      "grad_norm": 1.4684398174285889,
+      "learning_rate": 2e-05,
+      "loss": 0.03964592,
+      "step": 5437
+    },
+    {
+      "epoch": 10.876,
+      "grad_norm": 1.3081820011138916,
+      "learning_rate": 2e-05,
+      "loss": 0.03898221,
+      "step": 5438
+    },
+    {
+      "epoch": 10.878,
+      "grad_norm": 1.5860309600830078,
+      "learning_rate": 2e-05,
+      "loss": 0.03976344,
+      "step": 5439
+    },
+    {
+      "epoch": 10.88,
+      "grad_norm": 1.7701555490493774,
+      "learning_rate": 2e-05,
+      "loss": 0.04799116,
+      "step": 5440
+    },
+    {
+      "epoch": 10.882,
+      "grad_norm": 2.2064380645751953,
+      "learning_rate": 2e-05,
+      "loss": 0.0660604,
+      "step": 5441
+    },
+    {
+      "epoch": 10.884,
+      "grad_norm": 1.732148289680481,
+      "learning_rate": 2e-05,
+      "loss": 0.05492194,
+      "step": 5442
+    },
+    {
+      "epoch": 10.886,
+      "grad_norm": 1.9355525970458984,
+      "learning_rate": 2e-05,
+      "loss": 0.04734231,
+      "step": 5443
+    },
+    {
+      "epoch": 10.888,
+      "grad_norm": 1.2587823867797852,
+      "learning_rate": 2e-05,
+      "loss": 0.03634328,
+      "step": 5444
+    },
+    {
+      "epoch": 10.89,
+      "grad_norm": 1.5590167045593262,
+      "learning_rate": 2e-05,
+      "loss": 0.06726646,
+      "step": 5445
+    },
+    {
+      "epoch": 10.892,
+      "grad_norm": 1.3101967573165894,
+      "learning_rate": 2e-05,
+      "loss": 0.04764952,
+      "step": 5446
+    },
+    {
+      "epoch": 10.894,
+      "grad_norm": 2.4456593990325928,
+      "learning_rate": 2e-05,
+      "loss": 0.05624555,
+      "step": 5447
+    },
+    {
+      "epoch": 10.896,
+      "grad_norm": 2.824155330657959,
+      "learning_rate": 2e-05,
+      "loss": 0.05465455,
+      "step": 5448
+    },
+    {
+      "epoch": 10.898,
+      "grad_norm": 1.4769179821014404,
+      "learning_rate": 2e-05,
+      "loss": 0.05461916,
+      "step": 5449
+    },
+    {
+      "epoch": 10.9,
+      "grad_norm": 1.4648699760437012,
+      "learning_rate": 2e-05,
+      "loss": 0.05058068,
+      "step": 5450
+    },
+    {
+      "epoch": 10.902,
+      "grad_norm": 1.2384024858474731,
+      "learning_rate": 2e-05,
+      "loss": 0.04764497,
+      "step": 5451
+    },
+    {
+      "epoch": 10.904,
+      "grad_norm": 1.6182745695114136,
+      "learning_rate": 2e-05,
+      "loss": 0.05357751,
+      "step": 5452
+    },
+    {
+      "epoch": 10.906,
+      "grad_norm": 1.5405651330947876,
+      "learning_rate": 2e-05,
+      "loss": 0.0363545,
+      "step": 5453
+    },
+    {
+      "epoch": 10.908,
+      "grad_norm": 1.2049387693405151,
+      "learning_rate": 2e-05,
+      "loss": 0.0472638,
+      "step": 5454
+    },
+    {
+      "epoch": 10.91,
+      "grad_norm": 2.5308351516723633,
+      "learning_rate": 2e-05,
+      "loss": 0.04407152,
+      "step": 5455
+    },
+    {
+      "epoch": 10.912,
+      "grad_norm": 1.3933305740356445,
+      "learning_rate": 2e-05,
+      "loss": 0.04460661,
+      "step": 5456
+    },
+    {
+      "epoch": 10.914,
+      "grad_norm": 1.4004764556884766,
+      "learning_rate": 2e-05,
+      "loss": 0.04893076,
+      "step": 5457
+    },
+    {
+      "epoch": 10.916,
+      "grad_norm": 1.8205981254577637,
+      "learning_rate": 2e-05,
+      "loss": 0.04703292,
+      "step": 5458
+    },
+    {
+      "epoch": 10.918,
+      "grad_norm": 1.8013707399368286,
+      "learning_rate": 2e-05,
+      "loss": 0.055125,
+      "step": 5459
+    },
+    {
+      "epoch": 10.92,
+      "grad_norm": 1.451092004776001,
+      "learning_rate": 2e-05,
+      "loss": 0.04543486,
+      "step": 5460
+    },
+    {
+      "epoch": 10.922,
+      "grad_norm": 1.1553832292556763,
+      "learning_rate": 2e-05,
+      "loss": 0.04265869,
+      "step": 5461
+    },
+    {
+      "epoch": 10.924,
+      "grad_norm": 1.960945725440979,
+      "learning_rate": 2e-05,
+      "loss": 0.05424058,
+      "step": 5462
+    },
+    {
+      "epoch": 10.926,
+      "grad_norm": 1.4518814086914062,
+      "learning_rate": 2e-05,
+      "loss": 0.04993463,
+      "step": 5463
+    },
+    {
+      "epoch": 10.928,
+      "grad_norm": 2.5878188610076904,
+      "learning_rate": 2e-05,
+      "loss": 0.04935042,
+      "step": 5464
+    },
+    {
+      "epoch": 10.93,
+      "grad_norm": 1.3575770854949951,
+      "learning_rate": 2e-05,
+      "loss": 0.05218004,
+      "step": 5465
+    },
+    {
+      "epoch": 10.932,
+      "grad_norm": 1.4629638195037842,
+      "learning_rate": 2e-05,
+      "loss": 0.05007394,
+      "step": 5466
+    },
+    {
+      "epoch": 10.934,
+      "grad_norm": 1.2704107761383057,
+      "learning_rate": 2e-05,
+      "loss": 0.04607791,
+      "step": 5467
+    },
+    {
+      "epoch": 10.936,
+      "grad_norm": 1.5008957386016846,
+      "learning_rate": 2e-05,
+      "loss": 0.04755886,
+      "step": 5468
+    },
+    {
+      "epoch": 10.938,
+      "grad_norm": 1.6490647792816162,
+      "learning_rate": 2e-05,
+      "loss": 0.04356069,
+      "step": 5469
+    },
+    {
+      "epoch": 10.94,
+      "grad_norm": 1.9687342643737793,
+      "learning_rate": 2e-05,
+      "loss": 0.04906989,
+      "step": 5470
+    },
+    {
+      "epoch": 10.942,
+      "grad_norm": 2.3474905490875244,
+      "learning_rate": 2e-05,
+      "loss": 0.04697709,
+      "step": 5471
+    },
+    {
+      "epoch": 10.943999999999999,
+      "grad_norm": 1.7721339464187622,
+      "learning_rate": 2e-05,
+      "loss": 0.06682569,
+      "step": 5472
+    },
+    {
+      "epoch": 10.946,
+      "grad_norm": 1.4311152696609497,
+      "learning_rate": 2e-05,
+      "loss": 0.04959798,
+      "step": 5473
+    },
+    {
+      "epoch": 10.948,
+      "grad_norm": 1.3908685445785522,
+      "learning_rate": 2e-05,
+      "loss": 0.04007864,
+      "step": 5474
+    },
+    {
+      "epoch": 10.95,
+      "grad_norm": 1.1253409385681152,
+      "learning_rate": 2e-05,
+      "loss": 0.0392089,
+      "step": 5475
+    },
+    {
+      "epoch": 10.952,
+      "grad_norm": 1.5628206729888916,
+      "learning_rate": 2e-05,
+      "loss": 0.03705291,
+      "step": 5476
+    },
+    {
+      "epoch": 10.954,
+      "grad_norm": 2.015963077545166,
+      "learning_rate": 2e-05,
+      "loss": 0.05837575,
+      "step": 5477
+    },
+    {
+      "epoch": 10.956,
+      "grad_norm": 1.1806350946426392,
+      "learning_rate": 2e-05,
+      "loss": 0.04036895,
+      "step": 5478
+    },
+    {
+      "epoch": 10.958,
+      "grad_norm": 1.8110302686691284,
+      "learning_rate": 2e-05,
+      "loss": 0.05655763,
+      "step": 5479
+    },
+    {
+      "epoch": 10.96,
+      "grad_norm": 1.6625659465789795,
+      "learning_rate": 2e-05,
+      "loss": 0.04185361,
+      "step": 5480
+    },
+    {
+      "epoch": 10.962,
+      "grad_norm": 1.8857502937316895,
+      "learning_rate": 2e-05,
+      "loss": 0.05384925,
+      "step": 5481
+    },
+    {
+      "epoch": 10.964,
+      "grad_norm": 1.3210104703903198,
+      "learning_rate": 2e-05,
+      "loss": 0.0597494,
+      "step": 5482
+    },
+    {
+      "epoch": 10.966,
+      "grad_norm": 1.587018609046936,
+      "learning_rate": 2e-05,
+      "loss": 0.05472574,
+      "step": 5483
+    },
+    {
+      "epoch": 10.968,
+      "grad_norm": 1.3318067789077759,
+      "learning_rate": 2e-05,
+      "loss": 0.0541456,
+      "step": 5484
+    },
+    {
+      "epoch": 10.97,
+      "grad_norm": 1.4794989824295044,
+      "learning_rate": 2e-05,
+      "loss": 0.05960969,
+      "step": 5485
+    },
+    {
+      "epoch": 10.972,
+      "grad_norm": 3.501803398132324,
+      "learning_rate": 2e-05,
+      "loss": 0.0588432,
+      "step": 5486
+    },
+    {
+      "epoch": 10.974,
+      "grad_norm": 1.6838804483413696,
+      "learning_rate": 2e-05,
+      "loss": 0.05511768,
+      "step": 5487
+    },
+    {
+      "epoch": 10.975999999999999,
+      "grad_norm": 1.2759146690368652,
+      "learning_rate": 2e-05,
+      "loss": 0.02767239,
+      "step": 5488
+    },
+    {
+      "epoch": 10.978,
+      "grad_norm": 1.4586148262023926,
+      "learning_rate": 2e-05,
+      "loss": 0.05381429,
+      "step": 5489
+    },
+    {
+      "epoch": 10.98,
+      "grad_norm": 1.5755432844161987,
+      "learning_rate": 2e-05,
+      "loss": 0.03358914,
+      "step": 5490
+    },
+    {
+      "epoch": 10.982,
+      "grad_norm": 1.175610899925232,
+      "learning_rate": 2e-05,
+      "loss": 0.04267032,
+      "step": 5491
+    },
+    {
+      "epoch": 10.984,
+      "grad_norm": 1.1766899824142456,
+      "learning_rate": 2e-05,
+      "loss": 0.03875332,
+      "step": 5492
+    },
+    {
+      "epoch": 10.986,
+      "grad_norm": 2.118785858154297,
+      "learning_rate": 2e-05,
+      "loss": 0.05230087,
+      "step": 5493
+    },
+    {
+      "epoch": 10.988,
+      "grad_norm": 1.6943100690841675,
+      "learning_rate": 2e-05,
+      "loss": 0.05274594,
+      "step": 5494
+    },
+    {
+      "epoch": 10.99,
+      "grad_norm": 1.087948203086853,
+      "learning_rate": 2e-05,
+      "loss": 0.03773314,
+      "step": 5495
+    },
+    {
+      "epoch": 10.992,
+      "grad_norm": 1.6124333143234253,
+      "learning_rate": 2e-05,
+      "loss": 0.04908092,
+      "step": 5496
+    },
+    {
+      "epoch": 10.994,
+      "grad_norm": 1.851080298423767,
+      "learning_rate": 2e-05,
+      "loss": 0.05457485,
+      "step": 5497
+    },
+    {
+      "epoch": 10.996,
+      "grad_norm": 1.4189625978469849,
+      "learning_rate": 2e-05,
+      "loss": 0.04965939,
+      "step": 5498
+    },
+    {
+      "epoch": 10.998,
+      "grad_norm": 1.9393243789672852,
+      "learning_rate": 2e-05,
+      "loss": 0.05298079,
+      "step": 5499
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 1.709896445274353,
+      "learning_rate": 2e-05,
+      "loss": 0.05041164,
+      "step": 5500
+    },
+    {
+      "epoch": 11.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.998,
+        "AngleClassification_2": 0.996,
+        "AngleClassification_3": 0.9600798403193613,
+        "Equal_1": 0.994,
+        "Equal_2": 0.9660678642714571,
+        "Equal_3": 0.874251497005988,
+        "LineComparison_1": 0.996,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9740518962075848,
+        "Parallel_1": 0.9879759519038076,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.986,
+        "Perpendicular_1": 0.988,
+        "Perpendicular_2": 0.948,
+        "Perpendicular_3": 0.6513026052104208,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9956666666666667,
+        "PointLiesOnCircle_3": 0.9892666666666666,
+        "PointLiesOnLine_1": 0.9959919839679359,
+        "PointLiesOnLine_2": 0.9859719438877755,
+        "PointLiesOnLine_3": 0.9740518962075848
+      },
+      "eval_runtime": 319.8654,
+      "eval_samples_per_second": 32.826,
+      "eval_steps_per_second": 0.657,
+      "step": 5500
+    },
+    {
+      "epoch": 11.002,
+      "grad_norm": 1.7501477003097534,
+      "learning_rate": 2e-05,
+      "loss": 0.04067567,
+      "step": 5501
+    },
+    {
+      "epoch": 11.004,
+      "grad_norm": 2.942864179611206,
+      "learning_rate": 2e-05,
+      "loss": 0.05952626,
+      "step": 5502
+    },
+    {
+      "epoch": 11.006,
+      "grad_norm": 1.5498440265655518,
+      "learning_rate": 2e-05,
+      "loss": 0.03488794,
+      "step": 5503
+    },
+    {
+      "epoch": 11.008,
+      "grad_norm": 1.4755291938781738,
+      "learning_rate": 2e-05,
+      "loss": 0.05729527,
+      "step": 5504
+    },
+    {
+      "epoch": 11.01,
+      "grad_norm": 1.4485598802566528,
+      "learning_rate": 2e-05,
+      "loss": 0.05340921,
+      "step": 5505
+    },
+    {
+      "epoch": 11.012,
+      "grad_norm": 1.3793925046920776,
+      "learning_rate": 2e-05,
+      "loss": 0.04637571,
+      "step": 5506
+    },
+    {
+      "epoch": 11.014,
+      "grad_norm": 1.4103147983551025,
+      "learning_rate": 2e-05,
+      "loss": 0.04511734,
+      "step": 5507
+    },
+    {
+      "epoch": 11.016,
+      "grad_norm": 1.3534992933273315,
+      "learning_rate": 2e-05,
+      "loss": 0.04359635,
+      "step": 5508
+    },
+    {
+      "epoch": 11.018,
+      "grad_norm": 1.5268181562423706,
+      "learning_rate": 2e-05,
+      "loss": 0.07277807,
+      "step": 5509
+    },
+    {
+      "epoch": 11.02,
+      "grad_norm": 1.0915262699127197,
+      "learning_rate": 2e-05,
+      "loss": 0.05696312,
+      "step": 5510
+    },
+    {
+      "epoch": 11.022,
+      "grad_norm": 1.7027065753936768,
+      "learning_rate": 2e-05,
+      "loss": 0.05292184,
+      "step": 5511
+    },
+    {
+      "epoch": 11.024,
+      "grad_norm": 1.5341792106628418,
+      "learning_rate": 2e-05,
+      "loss": 0.05033635,
+      "step": 5512
+    },
+    {
+      "epoch": 11.026,
+      "grad_norm": 1.545507788658142,
+      "learning_rate": 2e-05,
+      "loss": 0.06413613,
+      "step": 5513
+    },
+    {
+      "epoch": 11.028,
+      "grad_norm": 1.227196455001831,
+      "learning_rate": 2e-05,
+      "loss": 0.04655745,
+      "step": 5514
+    },
+    {
+      "epoch": 11.03,
+      "grad_norm": 1.5401439666748047,
+      "learning_rate": 2e-05,
+      "loss": 0.05112302,
+      "step": 5515
+    },
+    {
+      "epoch": 11.032,
+      "grad_norm": 0.9862892627716064,
+      "learning_rate": 2e-05,
+      "loss": 0.03179028,
+      "step": 5516
+    },
+    {
+      "epoch": 11.034,
+      "grad_norm": 1.793843388557434,
+      "learning_rate": 2e-05,
+      "loss": 0.06312604,
+      "step": 5517
+    },
+    {
+      "epoch": 11.036,
+      "grad_norm": 2.1920554637908936,
+      "learning_rate": 2e-05,
+      "loss": 0.06224691,
+      "step": 5518
+    },
+    {
+      "epoch": 11.038,
+      "grad_norm": 1.6671439409255981,
+      "learning_rate": 2e-05,
+      "loss": 0.05866244,
+      "step": 5519
+    },
+    {
+      "epoch": 11.04,
+      "grad_norm": 1.550345540046692,
+      "learning_rate": 2e-05,
+      "loss": 0.04544277,
+      "step": 5520
+    },
+    {
+      "epoch": 11.042,
+      "grad_norm": 1.3246674537658691,
+      "learning_rate": 2e-05,
+      "loss": 0.03805883,
+      "step": 5521
+    },
+    {
+      "epoch": 11.044,
+      "grad_norm": 1.9008183479309082,
+      "learning_rate": 2e-05,
+      "loss": 0.04575905,
+      "step": 5522
+    },
+    {
+      "epoch": 11.046,
+      "grad_norm": 1.9028370380401611,
+      "learning_rate": 2e-05,
+      "loss": 0.05989815,
+      "step": 5523
+    },
+    {
+      "epoch": 11.048,
+      "grad_norm": 1.2603625059127808,
+      "learning_rate": 2e-05,
+      "loss": 0.03706544,
+      "step": 5524
+    },
+    {
+      "epoch": 11.05,
+      "grad_norm": 1.408327579498291,
+      "learning_rate": 2e-05,
+      "loss": 0.05105807,
+      "step": 5525
+    },
+    {
+      "epoch": 11.052,
+      "grad_norm": 1.762861728668213,
+      "learning_rate": 2e-05,
+      "loss": 0.03200347,
+      "step": 5526
+    },
+    {
+      "epoch": 11.054,
+      "grad_norm": 1.0934480428695679,
+      "learning_rate": 2e-05,
+      "loss": 0.04322592,
+      "step": 5527
+    },
+    {
+      "epoch": 11.056,
+      "grad_norm": 2.1491646766662598,
+      "learning_rate": 2e-05,
+      "loss": 0.04375915,
+      "step": 5528
+    },
+    {
+      "epoch": 11.058,
+      "grad_norm": 1.0617128610610962,
+      "learning_rate": 2e-05,
+      "loss": 0.03229889,
+      "step": 5529
+    },
+    {
+      "epoch": 11.06,
+      "grad_norm": 1.9952315092086792,
+      "learning_rate": 2e-05,
+      "loss": 0.0592179,
+      "step": 5530
+    },
+    {
+      "epoch": 11.062,
+      "grad_norm": 2.140627861022949,
+      "learning_rate": 2e-05,
+      "loss": 0.04644826,
+      "step": 5531
+    },
+    {
+      "epoch": 11.064,
+      "grad_norm": 1.437983512878418,
+      "learning_rate": 2e-05,
+      "loss": 0.03985,
+      "step": 5532
+    },
+    {
+      "epoch": 11.066,
+      "grad_norm": 2.316026210784912,
+      "learning_rate": 2e-05,
+      "loss": 0.05902199,
+      "step": 5533
+    },
+    {
+      "epoch": 11.068,
+      "grad_norm": 1.4913086891174316,
+      "learning_rate": 2e-05,
+      "loss": 0.03914908,
+      "step": 5534
+    },
+    {
+      "epoch": 11.07,
+      "grad_norm": 1.6400469541549683,
+      "learning_rate": 2e-05,
+      "loss": 0.04192843,
+      "step": 5535
+    },
+    {
+      "epoch": 11.072,
+      "grad_norm": 1.754105567932129,
+      "learning_rate": 2e-05,
+      "loss": 0.04989998,
+      "step": 5536
+    },
+    {
+      "epoch": 11.074,
+      "grad_norm": 2.085247039794922,
+      "learning_rate": 2e-05,
+      "loss": 0.04821778,
+      "step": 5537
+    },
+    {
+      "epoch": 11.076,
+      "grad_norm": 1.0525646209716797,
+      "learning_rate": 2e-05,
+      "loss": 0.03182067,
+      "step": 5538
+    },
+    {
+      "epoch": 11.078,
+      "grad_norm": 1.7224996089935303,
+      "learning_rate": 2e-05,
+      "loss": 0.04517365,
+      "step": 5539
+    },
+    {
+      "epoch": 11.08,
+      "grad_norm": 1.285914659500122,
+      "learning_rate": 2e-05,
+      "loss": 0.0277714,
+      "step": 5540
+    },
+    {
+      "epoch": 11.082,
+      "grad_norm": 1.9249039888381958,
+      "learning_rate": 2e-05,
+      "loss": 0.03731865,
+      "step": 5541
+    },
+    {
+      "epoch": 11.084,
+      "grad_norm": 3.5124082565307617,
+      "learning_rate": 2e-05,
+      "loss": 0.04889687,
+      "step": 5542
+    },
+    {
+      "epoch": 11.086,
+      "grad_norm": 1.3310890197753906,
+      "learning_rate": 2e-05,
+      "loss": 0.04091962,
+      "step": 5543
+    },
+    {
+      "epoch": 11.088,
+      "grad_norm": 1.5621048212051392,
+      "learning_rate": 2e-05,
+      "loss": 0.05734289,
+      "step": 5544
+    },
+    {
+      "epoch": 11.09,
+      "grad_norm": 2.3096556663513184,
+      "learning_rate": 2e-05,
+      "loss": 0.0544033,
+      "step": 5545
+    },
+    {
+      "epoch": 11.092,
+      "grad_norm": 2.0304203033447266,
+      "learning_rate": 2e-05,
+      "loss": 0.03412958,
+      "step": 5546
+    },
+    {
+      "epoch": 11.094,
+      "grad_norm": 3.0657801628112793,
+      "learning_rate": 2e-05,
+      "loss": 0.05514826,
+      "step": 5547
+    },
+    {
+      "epoch": 11.096,
+      "grad_norm": 1.4915958642959595,
+      "learning_rate": 2e-05,
+      "loss": 0.04357203,
+      "step": 5548
+    },
+    {
+      "epoch": 11.098,
+      "grad_norm": 2.101444959640503,
+      "learning_rate": 2e-05,
+      "loss": 0.05817562,
+      "step": 5549
+    },
+    {
+      "epoch": 11.1,
+      "grad_norm": 1.9840679168701172,
+      "learning_rate": 2e-05,
+      "loss": 0.05163846,
+      "step": 5550
+    },
+    {
+      "epoch": 11.102,
+      "grad_norm": 1.5036526918411255,
+      "learning_rate": 2e-05,
+      "loss": 0.03977584,
+      "step": 5551
+    },
+    {
+      "epoch": 11.104,
+      "grad_norm": 3.665069103240967,
+      "learning_rate": 2e-05,
+      "loss": 0.05804493,
+      "step": 5552
+    },
+    {
+      "epoch": 11.106,
+      "grad_norm": 6.136903762817383,
+      "learning_rate": 2e-05,
+      "loss": 0.05603513,
+      "step": 5553
+    },
+    {
+      "epoch": 11.108,
+      "grad_norm": 2.01459002494812,
+      "learning_rate": 2e-05,
+      "loss": 0.04759218,
+      "step": 5554
+    },
+    {
+      "epoch": 11.11,
+      "grad_norm": 1.221979022026062,
+      "learning_rate": 2e-05,
+      "loss": 0.04913111,
+      "step": 5555
+    },
+    {
+      "epoch": 11.112,
+      "grad_norm": 2.235368490219116,
+      "learning_rate": 2e-05,
+      "loss": 0.06046454,
+      "step": 5556
+    },
+    {
+      "epoch": 11.114,
+      "grad_norm": 2.291501998901367,
+      "learning_rate": 2e-05,
+      "loss": 0.05759356,
+      "step": 5557
+    },
+    {
+      "epoch": 11.116,
+      "grad_norm": 1.4968833923339844,
+      "learning_rate": 2e-05,
+      "loss": 0.06014612,
+      "step": 5558
+    },
+    {
+      "epoch": 11.118,
+      "grad_norm": 1.9257349967956543,
+      "learning_rate": 2e-05,
+      "loss": 0.0344676,
+      "step": 5559
+    },
+    {
+      "epoch": 11.12,
+      "grad_norm": 1.1876585483551025,
+      "learning_rate": 2e-05,
+      "loss": 0.03832501,
+      "step": 5560
+    },
+    {
+      "epoch": 11.122,
+      "grad_norm": 1.0107924938201904,
+      "learning_rate": 2e-05,
+      "loss": 0.03613226,
+      "step": 5561
+    },
+    {
+      "epoch": 11.124,
+      "grad_norm": 1.3432567119598389,
+      "learning_rate": 2e-05,
+      "loss": 0.0471199,
+      "step": 5562
+    },
+    {
+      "epoch": 11.126,
+      "grad_norm": 1.2003259658813477,
+      "learning_rate": 2e-05,
+      "loss": 0.04361069,
+      "step": 5563
+    },
+    {
+      "epoch": 11.128,
+      "grad_norm": 0.9868248701095581,
+      "learning_rate": 2e-05,
+      "loss": 0.03666285,
+      "step": 5564
+    },
+    {
+      "epoch": 11.13,
+      "grad_norm": 1.2956124544143677,
+      "learning_rate": 2e-05,
+      "loss": 0.04453753,
+      "step": 5565
+    },
+    {
+      "epoch": 11.132,
+      "grad_norm": 1.5581343173980713,
+      "learning_rate": 2e-05,
+      "loss": 0.04929563,
+      "step": 5566
+    },
+    {
+      "epoch": 11.134,
+      "grad_norm": 2.0106360912323,
+      "learning_rate": 2e-05,
+      "loss": 0.04062722,
+      "step": 5567
+    },
+    {
+      "epoch": 11.136,
+      "grad_norm": 1.4985320568084717,
+      "learning_rate": 2e-05,
+      "loss": 0.05832936,
+      "step": 5568
+    },
+    {
+      "epoch": 11.138,
+      "grad_norm": 1.6845251321792603,
+      "learning_rate": 2e-05,
+      "loss": 0.03393908,
+      "step": 5569
+    },
+    {
+      "epoch": 11.14,
+      "grad_norm": 1.7386165857315063,
+      "learning_rate": 2e-05,
+      "loss": 0.05278086,
+      "step": 5570
+    },
+    {
+      "epoch": 11.142,
+      "grad_norm": 1.629763126373291,
+      "learning_rate": 2e-05,
+      "loss": 0.05002792,
+      "step": 5571
+    },
+    {
+      "epoch": 11.144,
+      "grad_norm": 1.2016594409942627,
+      "learning_rate": 2e-05,
+      "loss": 0.05499132,
+      "step": 5572
+    },
+    {
+      "epoch": 11.146,
+      "grad_norm": 1.2156981229782104,
+      "learning_rate": 2e-05,
+      "loss": 0.04719136,
+      "step": 5573
+    },
+    {
+      "epoch": 11.148,
+      "grad_norm": 1.1108115911483765,
+      "learning_rate": 2e-05,
+      "loss": 0.04154814,
+      "step": 5574
+    },
+    {
+      "epoch": 11.15,
+      "grad_norm": 2.7240407466888428,
+      "learning_rate": 2e-05,
+      "loss": 0.07720132,
+      "step": 5575
+    },
+    {
+      "epoch": 11.152,
+      "grad_norm": 1.6736434698104858,
+      "learning_rate": 2e-05,
+      "loss": 0.0617914,
+      "step": 5576
+    },
+    {
+      "epoch": 11.154,
+      "grad_norm": 1.8674448728561401,
+      "learning_rate": 2e-05,
+      "loss": 0.05250496,
+      "step": 5577
+    },
+    {
+      "epoch": 11.156,
+      "grad_norm": 2.2997653484344482,
+      "learning_rate": 2e-05,
+      "loss": 0.08813807,
+      "step": 5578
+    },
+    {
+      "epoch": 11.158,
+      "grad_norm": 1.312296748161316,
+      "learning_rate": 2e-05,
+      "loss": 0.03612276,
+      "step": 5579
+    },
+    {
+      "epoch": 11.16,
+      "grad_norm": 1.4704864025115967,
+      "learning_rate": 2e-05,
+      "loss": 0.05287594,
+      "step": 5580
+    },
+    {
+      "epoch": 11.162,
+      "grad_norm": 0.9963863492012024,
+      "learning_rate": 2e-05,
+      "loss": 0.03753687,
+      "step": 5581
+    },
+    {
+      "epoch": 11.164,
+      "grad_norm": 2.919243097305298,
+      "learning_rate": 2e-05,
+      "loss": 0.05586375,
+      "step": 5582
+    },
+    {
+      "epoch": 11.166,
+      "grad_norm": 2.9407522678375244,
+      "learning_rate": 2e-05,
+      "loss": 0.04070038,
+      "step": 5583
+    },
+    {
+      "epoch": 11.168,
+      "grad_norm": 1.1414661407470703,
+      "learning_rate": 2e-05,
+      "loss": 0.03115014,
+      "step": 5584
+    },
+    {
+      "epoch": 11.17,
+      "grad_norm": 3.4734015464782715,
+      "learning_rate": 2e-05,
+      "loss": 0.05622058,
+      "step": 5585
+    },
+    {
+      "epoch": 11.172,
+      "grad_norm": 1.3070592880249023,
+      "learning_rate": 2e-05,
+      "loss": 0.05956008,
+      "step": 5586
+    },
+    {
+      "epoch": 11.174,
+      "grad_norm": 1.8945657014846802,
+      "learning_rate": 2e-05,
+      "loss": 0.05505826,
+      "step": 5587
+    },
+    {
+      "epoch": 11.176,
+      "grad_norm": 2.134795665740967,
+      "learning_rate": 2e-05,
+      "loss": 0.05037409,
+      "step": 5588
+    },
+    {
+      "epoch": 11.178,
+      "grad_norm": 1.618909239768982,
+      "learning_rate": 2e-05,
+      "loss": 0.04687338,
+      "step": 5589
+    },
+    {
+      "epoch": 11.18,
+      "grad_norm": 1.6525713205337524,
+      "learning_rate": 2e-05,
+      "loss": 0.03109693,
+      "step": 5590
+    },
+    {
+      "epoch": 11.182,
+      "grad_norm": 1.4128676652908325,
+      "learning_rate": 2e-05,
+      "loss": 0.04869646,
+      "step": 5591
+    },
+    {
+      "epoch": 11.184,
+      "grad_norm": 1.8516077995300293,
+      "learning_rate": 2e-05,
+      "loss": 0.05194153,
+      "step": 5592
+    },
+    {
+      "epoch": 11.186,
+      "grad_norm": 0.9598081111907959,
+      "learning_rate": 2e-05,
+      "loss": 0.03175935,
+      "step": 5593
+    },
+    {
+      "epoch": 11.188,
+      "grad_norm": 2.366396903991699,
+      "learning_rate": 2e-05,
+      "loss": 0.04697275,
+      "step": 5594
+    },
+    {
+      "epoch": 11.19,
+      "grad_norm": 1.4049228429794312,
+      "learning_rate": 2e-05,
+      "loss": 0.04712133,
+      "step": 5595
+    },
+    {
+      "epoch": 11.192,
+      "grad_norm": 1.392557978630066,
+      "learning_rate": 2e-05,
+      "loss": 0.04036619,
+      "step": 5596
+    },
+    {
+      "epoch": 11.194,
+      "grad_norm": 2.7289698123931885,
+      "learning_rate": 2e-05,
+      "loss": 0.05585587,
+      "step": 5597
+    },
+    {
+      "epoch": 11.196,
+      "grad_norm": 1.3307260274887085,
+      "learning_rate": 2e-05,
+      "loss": 0.04757424,
+      "step": 5598
+    },
+    {
+      "epoch": 11.198,
+      "grad_norm": 1.4445418119430542,
+      "learning_rate": 2e-05,
+      "loss": 0.04362855,
+      "step": 5599
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 1.3054895401000977,
+      "learning_rate": 2e-05,
+      "loss": 0.0476041,
+      "step": 5600
+    },
+    {
+      "epoch": 11.202,
+      "grad_norm": 1.1574424505233765,
+      "learning_rate": 2e-05,
+      "loss": 0.04523313,
+      "step": 5601
+    },
+    {
+      "epoch": 11.204,
+      "grad_norm": 1.6531107425689697,
+      "learning_rate": 2e-05,
+      "loss": 0.07256036,
+      "step": 5602
+    },
+    {
+      "epoch": 11.206,
+      "grad_norm": 2.6283984184265137,
+      "learning_rate": 2e-05,
+      "loss": 0.05259442,
+      "step": 5603
+    },
+    {
+      "epoch": 11.208,
+      "grad_norm": 1.1295219659805298,
+      "learning_rate": 2e-05,
+      "loss": 0.02722023,
+      "step": 5604
+    },
+    {
+      "epoch": 11.21,
+      "grad_norm": 1.7580029964447021,
+      "learning_rate": 2e-05,
+      "loss": 0.0378816,
+      "step": 5605
+    },
+    {
+      "epoch": 11.212,
+      "grad_norm": 1.1564671993255615,
+      "learning_rate": 2e-05,
+      "loss": 0.04453285,
+      "step": 5606
+    },
+    {
+      "epoch": 11.214,
+      "grad_norm": 2.0429930686950684,
+      "learning_rate": 2e-05,
+      "loss": 0.06690899,
+      "step": 5607
+    },
+    {
+      "epoch": 11.216,
+      "grad_norm": 3.2511401176452637,
+      "learning_rate": 2e-05,
+      "loss": 0.07473037,
+      "step": 5608
+    },
+    {
+      "epoch": 11.218,
+      "grad_norm": 1.5251747369766235,
+      "learning_rate": 2e-05,
+      "loss": 0.06034661,
+      "step": 5609
+    },
+    {
+      "epoch": 11.22,
+      "grad_norm": 1.3258527517318726,
+      "learning_rate": 2e-05,
+      "loss": 0.04761764,
+      "step": 5610
+    },
+    {
+      "epoch": 11.222,
+      "grad_norm": 1.9736108779907227,
+      "learning_rate": 2e-05,
+      "loss": 0.04002544,
+      "step": 5611
+    },
+    {
+      "epoch": 11.224,
+      "grad_norm": 1.7155126333236694,
+      "learning_rate": 2e-05,
+      "loss": 0.05041002,
+      "step": 5612
+    },
+    {
+      "epoch": 11.226,
+      "grad_norm": 1.2253735065460205,
+      "learning_rate": 2e-05,
+      "loss": 0.04136164,
+      "step": 5613
+    },
+    {
+      "epoch": 11.228,
+      "grad_norm": 1.5986520051956177,
+      "learning_rate": 2e-05,
+      "loss": 0.04651842,
+      "step": 5614
+    },
+    {
+      "epoch": 11.23,
+      "grad_norm": 1.277672529220581,
+      "learning_rate": 2e-05,
+      "loss": 0.04242517,
+      "step": 5615
+    },
+    {
+      "epoch": 11.232,
+      "grad_norm": 1.531924843788147,
+      "learning_rate": 2e-05,
+      "loss": 0.0391548,
+      "step": 5616
+    },
+    {
+      "epoch": 11.234,
+      "grad_norm": 1.1350847482681274,
+      "learning_rate": 2e-05,
+      "loss": 0.04663714,
+      "step": 5617
+    },
+    {
+      "epoch": 11.236,
+      "grad_norm": 1.4929993152618408,
+      "learning_rate": 2e-05,
+      "loss": 0.04786985,
+      "step": 5618
+    },
+    {
+      "epoch": 11.238,
+      "grad_norm": 1.8032375574111938,
+      "learning_rate": 2e-05,
+      "loss": 0.0391562,
+      "step": 5619
+    },
+    {
+      "epoch": 11.24,
+      "grad_norm": 1.310897946357727,
+      "learning_rate": 2e-05,
+      "loss": 0.0419879,
+      "step": 5620
+    },
+    {
+      "epoch": 11.242,
+      "grad_norm": 1.8344627618789673,
+      "learning_rate": 2e-05,
+      "loss": 0.05171813,
+      "step": 5621
+    },
+    {
+      "epoch": 11.244,
+      "grad_norm": 1.4439165592193604,
+      "learning_rate": 2e-05,
+      "loss": 0.04475713,
+      "step": 5622
+    },
+    {
+      "epoch": 11.246,
+      "grad_norm": 1.3868870735168457,
+      "learning_rate": 2e-05,
+      "loss": 0.0524424,
+      "step": 5623
+    },
+    {
+      "epoch": 11.248,
+      "grad_norm": 1.4693684577941895,
+      "learning_rate": 2e-05,
+      "loss": 0.03726972,
+      "step": 5624
+    },
+    {
+      "epoch": 11.25,
+      "grad_norm": 1.3319610357284546,
+      "learning_rate": 2e-05,
+      "loss": 0.0376013,
+      "step": 5625
+    },
+    {
+      "epoch": 11.252,
+      "grad_norm": 1.1958826780319214,
+      "learning_rate": 2e-05,
+      "loss": 0.03406876,
+      "step": 5626
+    },
+    {
+      "epoch": 11.254,
+      "grad_norm": 1.5152924060821533,
+      "learning_rate": 2e-05,
+      "loss": 0.05023153,
+      "step": 5627
+    },
+    {
+      "epoch": 11.256,
+      "grad_norm": 1.8399534225463867,
+      "learning_rate": 2e-05,
+      "loss": 0.05513574,
+      "step": 5628
+    },
+    {
+      "epoch": 11.258,
+      "grad_norm": 1.1258716583251953,
+      "learning_rate": 2e-05,
+      "loss": 0.02846676,
+      "step": 5629
+    },
+    {
+      "epoch": 11.26,
+      "grad_norm": 1.4945526123046875,
+      "learning_rate": 2e-05,
+      "loss": 0.04922044,
+      "step": 5630
+    },
+    {
+      "epoch": 11.262,
+      "grad_norm": 1.420878291130066,
+      "learning_rate": 2e-05,
+      "loss": 0.05598453,
+      "step": 5631
+    },
+    {
+      "epoch": 11.264,
+      "grad_norm": 1.74343740940094,
+      "learning_rate": 2e-05,
+      "loss": 0.05555171,
+      "step": 5632
+    },
+    {
+      "epoch": 11.266,
+      "grad_norm": 1.4211103916168213,
+      "learning_rate": 2e-05,
+      "loss": 0.04510403,
+      "step": 5633
+    },
+    {
+      "epoch": 11.268,
+      "grad_norm": 1.2378079891204834,
+      "learning_rate": 2e-05,
+      "loss": 0.04460875,
+      "step": 5634
+    },
+    {
+      "epoch": 11.27,
+      "grad_norm": 2.1717655658721924,
+      "learning_rate": 2e-05,
+      "loss": 0.04952989,
+      "step": 5635
+    },
+    {
+      "epoch": 11.272,
+      "grad_norm": 1.3453963994979858,
+      "learning_rate": 2e-05,
+      "loss": 0.04727048,
+      "step": 5636
+    },
+    {
+      "epoch": 11.274000000000001,
+      "grad_norm": 1.2908873558044434,
+      "learning_rate": 2e-05,
+      "loss": 0.04669511,
+      "step": 5637
+    },
+    {
+      "epoch": 11.276,
+      "grad_norm": 2.5825376510620117,
+      "learning_rate": 2e-05,
+      "loss": 0.07910918,
+      "step": 5638
+    },
+    {
+      "epoch": 11.278,
+      "grad_norm": 1.3286540508270264,
+      "learning_rate": 2e-05,
+      "loss": 0.05208978,
+      "step": 5639
+    },
+    {
+      "epoch": 11.28,
+      "grad_norm": 2.3290231227874756,
+      "learning_rate": 2e-05,
+      "loss": 0.0465233,
+      "step": 5640
+    },
+    {
+      "epoch": 11.282,
+      "grad_norm": 3.7450168132781982,
+      "learning_rate": 2e-05,
+      "loss": 0.04494961,
+      "step": 5641
+    },
+    {
+      "epoch": 11.284,
+      "grad_norm": 2.037266492843628,
+      "learning_rate": 2e-05,
+      "loss": 0.05497518,
+      "step": 5642
+    },
+    {
+      "epoch": 11.286,
+      "grad_norm": 1.5916911363601685,
+      "learning_rate": 2e-05,
+      "loss": 0.04094953,
+      "step": 5643
+    },
+    {
+      "epoch": 11.288,
+      "grad_norm": 1.6948169469833374,
+      "learning_rate": 2e-05,
+      "loss": 0.0476352,
+      "step": 5644
+    },
+    {
+      "epoch": 11.29,
+      "grad_norm": 2.1372830867767334,
+      "learning_rate": 2e-05,
+      "loss": 0.05398251,
+      "step": 5645
+    },
+    {
+      "epoch": 11.292,
+      "grad_norm": 1.3128230571746826,
+      "learning_rate": 2e-05,
+      "loss": 0.04471845,
+      "step": 5646
+    },
+    {
+      "epoch": 11.294,
+      "grad_norm": 1.647568941116333,
+      "learning_rate": 2e-05,
+      "loss": 0.04331946,
+      "step": 5647
+    },
+    {
+      "epoch": 11.296,
+      "grad_norm": 1.923680067062378,
+      "learning_rate": 2e-05,
+      "loss": 0.05029312,
+      "step": 5648
+    },
+    {
+      "epoch": 11.298,
+      "grad_norm": 1.4542899131774902,
+      "learning_rate": 2e-05,
+      "loss": 0.05089693,
+      "step": 5649
+    },
+    {
+      "epoch": 11.3,
+      "grad_norm": 1.5298106670379639,
+      "learning_rate": 2e-05,
+      "loss": 0.05881848,
+      "step": 5650
+    },
+    {
+      "epoch": 11.302,
+      "grad_norm": 1.7057805061340332,
+      "learning_rate": 2e-05,
+      "loss": 0.03789561,
+      "step": 5651
+    },
+    {
+      "epoch": 11.304,
+      "grad_norm": 1.3690357208251953,
+      "learning_rate": 2e-05,
+      "loss": 0.04559196,
+      "step": 5652
+    },
+    {
+      "epoch": 11.306,
+      "grad_norm": 1.7711060047149658,
+      "learning_rate": 2e-05,
+      "loss": 0.04960579,
+      "step": 5653
+    },
+    {
+      "epoch": 11.308,
+      "grad_norm": 2.154360294342041,
+      "learning_rate": 2e-05,
+      "loss": 0.04761767,
+      "step": 5654
+    },
+    {
+      "epoch": 11.31,
+      "grad_norm": 1.1826550960540771,
+      "learning_rate": 2e-05,
+      "loss": 0.04088816,
+      "step": 5655
+    },
+    {
+      "epoch": 11.312,
+      "grad_norm": 2.5155763626098633,
+      "learning_rate": 2e-05,
+      "loss": 0.04958711,
+      "step": 5656
+    },
+    {
+      "epoch": 11.314,
+      "grad_norm": 1.2392568588256836,
+      "learning_rate": 2e-05,
+      "loss": 0.05496053,
+      "step": 5657
+    },
+    {
+      "epoch": 11.316,
+      "grad_norm": 1.723213791847229,
+      "learning_rate": 2e-05,
+      "loss": 0.04695657,
+      "step": 5658
+    },
+    {
+      "epoch": 11.318,
+      "grad_norm": 1.6298481225967407,
+      "learning_rate": 2e-05,
+      "loss": 0.03813414,
+      "step": 5659
+    },
+    {
+      "epoch": 11.32,
+      "grad_norm": 1.4247621297836304,
+      "learning_rate": 2e-05,
+      "loss": 0.04710161,
+      "step": 5660
+    },
+    {
+      "epoch": 11.322,
+      "grad_norm": 1.5706455707550049,
+      "learning_rate": 2e-05,
+      "loss": 0.0463535,
+      "step": 5661
+    },
+    {
+      "epoch": 11.324,
+      "grad_norm": 2.462221622467041,
+      "learning_rate": 2e-05,
+      "loss": 0.05929623,
+      "step": 5662
+    },
+    {
+      "epoch": 11.326,
+      "grad_norm": 1.4576634168624878,
+      "learning_rate": 2e-05,
+      "loss": 0.05897439,
+      "step": 5663
+    },
+    {
+      "epoch": 11.328,
+      "grad_norm": 1.6032121181488037,
+      "learning_rate": 2e-05,
+      "loss": 0.04866658,
+      "step": 5664
+    },
+    {
+      "epoch": 11.33,
+      "grad_norm": 1.4402192831039429,
+      "learning_rate": 2e-05,
+      "loss": 0.03332867,
+      "step": 5665
+    },
+    {
+      "epoch": 11.332,
+      "grad_norm": 2.8734233379364014,
+      "learning_rate": 2e-05,
+      "loss": 0.06615589,
+      "step": 5666
+    },
+    {
+      "epoch": 11.334,
+      "grad_norm": 1.652347445487976,
+      "learning_rate": 2e-05,
+      "loss": 0.07197806,
+      "step": 5667
+    },
+    {
+      "epoch": 11.336,
+      "grad_norm": 1.8642592430114746,
+      "learning_rate": 2e-05,
+      "loss": 0.0569929,
+      "step": 5668
+    },
+    {
+      "epoch": 11.338,
+      "grad_norm": 1.2723792791366577,
+      "learning_rate": 2e-05,
+      "loss": 0.04522455,
+      "step": 5669
+    },
+    {
+      "epoch": 11.34,
+      "grad_norm": 1.5746574401855469,
+      "learning_rate": 2e-05,
+      "loss": 0.05542893,
+      "step": 5670
+    },
+    {
+      "epoch": 11.342,
+      "grad_norm": 1.4517905712127686,
+      "learning_rate": 2e-05,
+      "loss": 0.04733806,
+      "step": 5671
+    },
+    {
+      "epoch": 11.344,
+      "grad_norm": 1.3056635856628418,
+      "learning_rate": 2e-05,
+      "loss": 0.04615809,
+      "step": 5672
+    },
+    {
+      "epoch": 11.346,
+      "grad_norm": 2.0126426219940186,
+      "learning_rate": 2e-05,
+      "loss": 0.04903982,
+      "step": 5673
+    },
+    {
+      "epoch": 11.348,
+      "grad_norm": 2.094006299972534,
+      "learning_rate": 2e-05,
+      "loss": 0.0404023,
+      "step": 5674
+    },
+    {
+      "epoch": 11.35,
+      "grad_norm": 2.433847188949585,
+      "learning_rate": 2e-05,
+      "loss": 0.03881406,
+      "step": 5675
+    },
+    {
+      "epoch": 11.352,
+      "grad_norm": 1.2057197093963623,
+      "learning_rate": 2e-05,
+      "loss": 0.0405013,
+      "step": 5676
+    },
+    {
+      "epoch": 11.354,
+      "grad_norm": 1.8133881092071533,
+      "learning_rate": 2e-05,
+      "loss": 0.04236561,
+      "step": 5677
+    },
+    {
+      "epoch": 11.356,
+      "grad_norm": 2.568739891052246,
+      "learning_rate": 2e-05,
+      "loss": 0.05626953,
+      "step": 5678
+    },
+    {
+      "epoch": 11.358,
+      "grad_norm": 1.2331525087356567,
+      "learning_rate": 2e-05,
+      "loss": 0.05501791,
+      "step": 5679
+    },
+    {
+      "epoch": 11.36,
+      "grad_norm": 1.5900580883026123,
+      "learning_rate": 2e-05,
+      "loss": 0.03576569,
+      "step": 5680
+    },
+    {
+      "epoch": 11.362,
+      "grad_norm": 1.9113343954086304,
+      "learning_rate": 2e-05,
+      "loss": 0.06783809,
+      "step": 5681
+    },
+    {
+      "epoch": 11.364,
+      "grad_norm": 2.187023639678955,
+      "learning_rate": 2e-05,
+      "loss": 0.04357416,
+      "step": 5682
+    },
+    {
+      "epoch": 11.366,
+      "grad_norm": 1.2699072360992432,
+      "learning_rate": 2e-05,
+      "loss": 0.03943901,
+      "step": 5683
+    },
+    {
+      "epoch": 11.368,
+      "grad_norm": 1.9693132638931274,
+      "learning_rate": 2e-05,
+      "loss": 0.05115809,
+      "step": 5684
+    },
+    {
+      "epoch": 11.37,
+      "grad_norm": 1.3677825927734375,
+      "learning_rate": 2e-05,
+      "loss": 0.04716478,
+      "step": 5685
+    },
+    {
+      "epoch": 11.372,
+      "grad_norm": 2.4023549556732178,
+      "learning_rate": 2e-05,
+      "loss": 0.04504751,
+      "step": 5686
+    },
+    {
+      "epoch": 11.374,
+      "grad_norm": 2.511744737625122,
+      "learning_rate": 2e-05,
+      "loss": 0.04527961,
+      "step": 5687
+    },
+    {
+      "epoch": 11.376,
+      "grad_norm": 3.562417984008789,
+      "learning_rate": 2e-05,
+      "loss": 0.05556261,
+      "step": 5688
+    },
+    {
+      "epoch": 11.378,
+      "grad_norm": 1.733673095703125,
+      "learning_rate": 2e-05,
+      "loss": 0.0592088,
+      "step": 5689
+    },
+    {
+      "epoch": 11.38,
+      "grad_norm": 1.9149365425109863,
+      "learning_rate": 2e-05,
+      "loss": 0.04439528,
+      "step": 5690
+    },
+    {
+      "epoch": 11.382,
+      "grad_norm": 2.3991384506225586,
+      "learning_rate": 2e-05,
+      "loss": 0.06331401,
+      "step": 5691
+    },
+    {
+      "epoch": 11.384,
+      "grad_norm": 1.122333288192749,
+      "learning_rate": 2e-05,
+      "loss": 0.04532647,
+      "step": 5692
+    },
+    {
+      "epoch": 11.386,
+      "grad_norm": 1.6301887035369873,
+      "learning_rate": 2e-05,
+      "loss": 0.05703401,
+      "step": 5693
+    },
+    {
+      "epoch": 11.388,
+      "grad_norm": 1.636753797531128,
+      "learning_rate": 2e-05,
+      "loss": 0.05419588,
+      "step": 5694
+    },
+    {
+      "epoch": 11.39,
+      "grad_norm": 1.2946252822875977,
+      "learning_rate": 2e-05,
+      "loss": 0.05132918,
+      "step": 5695
+    },
+    {
+      "epoch": 11.392,
+      "grad_norm": 2.3236074447631836,
+      "learning_rate": 2e-05,
+      "loss": 0.05070593,
+      "step": 5696
+    },
+    {
+      "epoch": 11.394,
+      "grad_norm": 1.109092116355896,
+      "learning_rate": 2e-05,
+      "loss": 0.0335158,
+      "step": 5697
+    },
+    {
+      "epoch": 11.396,
+      "grad_norm": 1.7937777042388916,
+      "learning_rate": 2e-05,
+      "loss": 0.06048029,
+      "step": 5698
+    },
+    {
+      "epoch": 11.398,
+      "grad_norm": 1.0988600254058838,
+      "learning_rate": 2e-05,
+      "loss": 0.04137257,
+      "step": 5699
+    },
+    {
+      "epoch": 11.4,
+      "grad_norm": 1.4302674531936646,
+      "learning_rate": 2e-05,
+      "loss": 0.04906035,
+      "step": 5700
+    },
+    {
+      "epoch": 11.402,
+      "grad_norm": 1.5210566520690918,
+      "learning_rate": 2e-05,
+      "loss": 0.03527754,
+      "step": 5701
+    },
+    {
+      "epoch": 11.404,
+      "grad_norm": 2.1926465034484863,
+      "learning_rate": 2e-05,
+      "loss": 0.04686191,
+      "step": 5702
+    },
+    {
+      "epoch": 11.406,
+      "grad_norm": 1.6266396045684814,
+      "learning_rate": 2e-05,
+      "loss": 0.05651092,
+      "step": 5703
+    },
+    {
+      "epoch": 11.408,
+      "grad_norm": 2.796839952468872,
+      "learning_rate": 2e-05,
+      "loss": 0.07000058,
+      "step": 5704
+    },
+    {
+      "epoch": 11.41,
+      "grad_norm": 1.4611526727676392,
+      "learning_rate": 2e-05,
+      "loss": 0.05321472,
+      "step": 5705
+    },
+    {
+      "epoch": 11.412,
+      "grad_norm": 1.8903465270996094,
+      "learning_rate": 2e-05,
+      "loss": 0.05226947,
+      "step": 5706
+    },
+    {
+      "epoch": 11.414,
+      "grad_norm": 1.4791998863220215,
+      "learning_rate": 2e-05,
+      "loss": 0.05257621,
+      "step": 5707
+    },
+    {
+      "epoch": 11.416,
+      "grad_norm": 1.5878609418869019,
+      "learning_rate": 2e-05,
+      "loss": 0.06273852,
+      "step": 5708
+    },
+    {
+      "epoch": 11.418,
+      "grad_norm": 2.1732583045959473,
+      "learning_rate": 2e-05,
+      "loss": 0.05149956,
+      "step": 5709
+    },
+    {
+      "epoch": 11.42,
+      "grad_norm": 1.677105188369751,
+      "learning_rate": 2e-05,
+      "loss": 0.0615867,
+      "step": 5710
+    },
+    {
+      "epoch": 11.422,
+      "grad_norm": 1.29386305809021,
+      "learning_rate": 2e-05,
+      "loss": 0.04953182,
+      "step": 5711
+    },
+    {
+      "epoch": 11.424,
+      "grad_norm": 1.7477467060089111,
+      "learning_rate": 2e-05,
+      "loss": 0.06783712,
+      "step": 5712
+    },
+    {
+      "epoch": 11.426,
+      "grad_norm": 3.544480085372925,
+      "learning_rate": 2e-05,
+      "loss": 0.04404292,
+      "step": 5713
+    },
+    {
+      "epoch": 11.428,
+      "grad_norm": 2.753350019454956,
+      "learning_rate": 2e-05,
+      "loss": 0.07701689,
+      "step": 5714
+    },
+    {
+      "epoch": 11.43,
+      "grad_norm": 1.8815979957580566,
+      "learning_rate": 2e-05,
+      "loss": 0.05695759,
+      "step": 5715
+    },
+    {
+      "epoch": 11.432,
+      "grad_norm": 1.5264737606048584,
+      "learning_rate": 2e-05,
+      "loss": 0.05297385,
+      "step": 5716
+    },
+    {
+      "epoch": 11.434,
+      "grad_norm": 1.047830581665039,
+      "learning_rate": 2e-05,
+      "loss": 0.03878511,
+      "step": 5717
+    },
+    {
+      "epoch": 11.436,
+      "grad_norm": 1.0375819206237793,
+      "learning_rate": 2e-05,
+      "loss": 0.03545886,
+      "step": 5718
+    },
+    {
+      "epoch": 11.438,
+      "grad_norm": 1.4977059364318848,
+      "learning_rate": 2e-05,
+      "loss": 0.0493038,
+      "step": 5719
+    },
+    {
+      "epoch": 11.44,
+      "grad_norm": 1.3288768529891968,
+      "learning_rate": 2e-05,
+      "loss": 0.0505427,
+      "step": 5720
+    },
+    {
+      "epoch": 11.442,
+      "grad_norm": 1.954521894454956,
+      "learning_rate": 2e-05,
+      "loss": 0.06295772,
+      "step": 5721
+    },
+    {
+      "epoch": 11.444,
+      "grad_norm": 1.0402206182479858,
+      "learning_rate": 2e-05,
+      "loss": 0.03475888,
+      "step": 5722
+    },
+    {
+      "epoch": 11.446,
+      "grad_norm": 2.404123544692993,
+      "learning_rate": 2e-05,
+      "loss": 0.0562352,
+      "step": 5723
+    },
+    {
+      "epoch": 11.448,
+      "grad_norm": 1.7195630073547363,
+      "learning_rate": 2e-05,
+      "loss": 0.05285126,
+      "step": 5724
+    },
+    {
+      "epoch": 11.45,
+      "grad_norm": 1.2412029504776,
+      "learning_rate": 2e-05,
+      "loss": 0.05088343,
+      "step": 5725
+    },
+    {
+      "epoch": 11.452,
+      "grad_norm": 2.3141374588012695,
+      "learning_rate": 2e-05,
+      "loss": 0.0701945,
+      "step": 5726
+    },
+    {
+      "epoch": 11.454,
+      "grad_norm": 1.1392027139663696,
+      "learning_rate": 2e-05,
+      "loss": 0.03937323,
+      "step": 5727
+    },
+    {
+      "epoch": 11.456,
+      "grad_norm": 1.5960547924041748,
+      "learning_rate": 2e-05,
+      "loss": 0.04383022,
+      "step": 5728
+    },
+    {
+      "epoch": 11.458,
+      "grad_norm": 1.1272916793823242,
+      "learning_rate": 2e-05,
+      "loss": 0.04322441,
+      "step": 5729
+    },
+    {
+      "epoch": 11.46,
+      "grad_norm": 1.3566938638687134,
+      "learning_rate": 2e-05,
+      "loss": 0.04182547,
+      "step": 5730
+    },
+    {
+      "epoch": 11.462,
+      "grad_norm": 1.7132365703582764,
+      "learning_rate": 2e-05,
+      "loss": 0.04979304,
+      "step": 5731
+    },
+    {
+      "epoch": 11.464,
+      "grad_norm": 1.125922679901123,
+      "learning_rate": 2e-05,
+      "loss": 0.04207063,
+      "step": 5732
+    },
+    {
+      "epoch": 11.466,
+      "grad_norm": 1.6913458108901978,
+      "learning_rate": 2e-05,
+      "loss": 0.04294857,
+      "step": 5733
+    },
+    {
+      "epoch": 11.468,
+      "grad_norm": 1.0827990770339966,
+      "learning_rate": 2e-05,
+      "loss": 0.05028411,
+      "step": 5734
+    },
+    {
+      "epoch": 11.47,
+      "grad_norm": 1.2500609159469604,
+      "learning_rate": 2e-05,
+      "loss": 0.04735212,
+      "step": 5735
+    },
+    {
+      "epoch": 11.472,
+      "grad_norm": 1.5820868015289307,
+      "learning_rate": 2e-05,
+      "loss": 0.04795214,
+      "step": 5736
+    },
+    {
+      "epoch": 11.474,
+      "grad_norm": 4.20005464553833,
+      "learning_rate": 2e-05,
+      "loss": 0.05945217,
+      "step": 5737
+    },
+    {
+      "epoch": 11.475999999999999,
+      "grad_norm": 1.4440017938613892,
+      "learning_rate": 2e-05,
+      "loss": 0.0505971,
+      "step": 5738
+    },
+    {
+      "epoch": 11.478,
+      "grad_norm": 1.6714645624160767,
+      "learning_rate": 2e-05,
+      "loss": 0.04327673,
+      "step": 5739
+    },
+    {
+      "epoch": 11.48,
+      "grad_norm": 1.5207058191299438,
+      "learning_rate": 2e-05,
+      "loss": 0.03629073,
+      "step": 5740
+    },
+    {
+      "epoch": 11.482,
+      "grad_norm": 2.2845335006713867,
+      "learning_rate": 2e-05,
+      "loss": 0.04524094,
+      "step": 5741
+    },
+    {
+      "epoch": 11.484,
+      "grad_norm": 1.6629916429519653,
+      "learning_rate": 2e-05,
+      "loss": 0.03859144,
+      "step": 5742
+    },
+    {
+      "epoch": 11.486,
+      "grad_norm": 1.1069337129592896,
+      "learning_rate": 2e-05,
+      "loss": 0.03897121,
+      "step": 5743
+    },
+    {
+      "epoch": 11.488,
+      "grad_norm": 1.555224895477295,
+      "learning_rate": 2e-05,
+      "loss": 0.03899939,
+      "step": 5744
+    },
+    {
+      "epoch": 11.49,
+      "grad_norm": 2.02616548538208,
+      "learning_rate": 2e-05,
+      "loss": 0.05024801,
+      "step": 5745
+    },
+    {
+      "epoch": 11.492,
+      "grad_norm": 1.102475881576538,
+      "learning_rate": 2e-05,
+      "loss": 0.03961539,
+      "step": 5746
+    },
+    {
+      "epoch": 11.494,
+      "grad_norm": 1.6318930387496948,
+      "learning_rate": 2e-05,
+      "loss": 0.04914928,
+      "step": 5747
+    },
+    {
+      "epoch": 11.496,
+      "grad_norm": 1.7205697298049927,
+      "learning_rate": 2e-05,
+      "loss": 0.04520516,
+      "step": 5748
+    },
+    {
+      "epoch": 11.498,
+      "grad_norm": 2.896815538406372,
+      "learning_rate": 2e-05,
+      "loss": 0.06124,
+      "step": 5749
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 1.4367793798446655,
+      "learning_rate": 2e-05,
+      "loss": 0.03653619,
+      "step": 5750
+    },
+    {
+      "epoch": 11.502,
+      "grad_norm": 1.8099347352981567,
+      "learning_rate": 2e-05,
+      "loss": 0.04635201,
+      "step": 5751
+    },
+    {
+      "epoch": 11.504,
+      "grad_norm": 1.852498173713684,
+      "learning_rate": 2e-05,
+      "loss": 0.05065921,
+      "step": 5752
+    },
+    {
+      "epoch": 11.506,
+      "grad_norm": 1.1768702268600464,
+      "learning_rate": 2e-05,
+      "loss": 0.03448585,
+      "step": 5753
+    },
+    {
+      "epoch": 11.508,
+      "grad_norm": 1.5099678039550781,
+      "learning_rate": 2e-05,
+      "loss": 0.03958462,
+      "step": 5754
+    },
+    {
+      "epoch": 11.51,
+      "grad_norm": 3.429736375808716,
+      "learning_rate": 2e-05,
+      "loss": 0.04971122,
+      "step": 5755
+    },
+    {
+      "epoch": 11.512,
+      "grad_norm": 1.356461524963379,
+      "learning_rate": 2e-05,
+      "loss": 0.04755508,
+      "step": 5756
+    },
+    {
+      "epoch": 11.514,
+      "grad_norm": 2.2165658473968506,
+      "learning_rate": 2e-05,
+      "loss": 0.0588036,
+      "step": 5757
+    },
+    {
+      "epoch": 11.516,
+      "grad_norm": 1.2915750741958618,
+      "learning_rate": 2e-05,
+      "loss": 0.04737089,
+      "step": 5758
+    },
+    {
+      "epoch": 11.518,
+      "grad_norm": 1.303098440170288,
+      "learning_rate": 2e-05,
+      "loss": 0.04432829,
+      "step": 5759
+    },
+    {
+      "epoch": 11.52,
+      "grad_norm": 1.4214109182357788,
+      "learning_rate": 2e-05,
+      "loss": 0.04704283,
+      "step": 5760
+    },
+    {
+      "epoch": 11.522,
+      "grad_norm": 1.403090476989746,
+      "learning_rate": 2e-05,
+      "loss": 0.03694802,
+      "step": 5761
+    },
+    {
+      "epoch": 11.524000000000001,
+      "grad_norm": 4.832217216491699,
+      "learning_rate": 2e-05,
+      "loss": 0.05118045,
+      "step": 5762
+    },
+    {
+      "epoch": 11.526,
+      "grad_norm": 1.1842046976089478,
+      "learning_rate": 2e-05,
+      "loss": 0.0342038,
+      "step": 5763
+    },
+    {
+      "epoch": 11.528,
+      "grad_norm": 1.0605454444885254,
+      "learning_rate": 2e-05,
+      "loss": 0.03051936,
+      "step": 5764
+    },
+    {
+      "epoch": 11.53,
+      "grad_norm": 1.4966093301773071,
+      "learning_rate": 2e-05,
+      "loss": 0.05437325,
+      "step": 5765
+    },
+    {
+      "epoch": 11.532,
+      "grad_norm": 1.8924329280853271,
+      "learning_rate": 2e-05,
+      "loss": 0.0331119,
+      "step": 5766
+    },
+    {
+      "epoch": 11.534,
+      "grad_norm": 1.384433388710022,
+      "learning_rate": 2e-05,
+      "loss": 0.05945661,
+      "step": 5767
+    },
+    {
+      "epoch": 11.536,
+      "grad_norm": 1.6517621278762817,
+      "learning_rate": 2e-05,
+      "loss": 0.04912163,
+      "step": 5768
+    },
+    {
+      "epoch": 11.538,
+      "grad_norm": 1.414260983467102,
+      "learning_rate": 2e-05,
+      "loss": 0.05474976,
+      "step": 5769
+    },
+    {
+      "epoch": 11.54,
+      "grad_norm": 1.017541527748108,
+      "learning_rate": 2e-05,
+      "loss": 0.0367628,
+      "step": 5770
+    },
+    {
+      "epoch": 11.542,
+      "grad_norm": 1.0602508783340454,
+      "learning_rate": 2e-05,
+      "loss": 0.03636168,
+      "step": 5771
+    },
+    {
+      "epoch": 11.544,
+      "grad_norm": 1.6005361080169678,
+      "learning_rate": 2e-05,
+      "loss": 0.06516649,
+      "step": 5772
+    },
+    {
+      "epoch": 11.546,
+      "grad_norm": 1.2333847284317017,
+      "learning_rate": 2e-05,
+      "loss": 0.0381854,
+      "step": 5773
+    },
+    {
+      "epoch": 11.548,
+      "grad_norm": 1.438815951347351,
+      "learning_rate": 2e-05,
+      "loss": 0.04053817,
+      "step": 5774
+    },
+    {
+      "epoch": 11.55,
+      "grad_norm": 1.6449198722839355,
+      "learning_rate": 2e-05,
+      "loss": 0.04813613,
+      "step": 5775
+    },
+    {
+      "epoch": 11.552,
+      "grad_norm": 2.104741096496582,
+      "learning_rate": 2e-05,
+      "loss": 0.04231735,
+      "step": 5776
+    },
+    {
+      "epoch": 11.554,
+      "grad_norm": 1.2205615043640137,
+      "learning_rate": 2e-05,
+      "loss": 0.05345644,
+      "step": 5777
+    },
+    {
+      "epoch": 11.556000000000001,
+      "grad_norm": 1.7869209051132202,
+      "learning_rate": 2e-05,
+      "loss": 0.04566547,
+      "step": 5778
+    },
+    {
+      "epoch": 11.558,
+      "grad_norm": 2.4200775623321533,
+      "learning_rate": 2e-05,
+      "loss": 0.0442078,
+      "step": 5779
+    },
+    {
+      "epoch": 11.56,
+      "grad_norm": 1.3662333488464355,
+      "learning_rate": 2e-05,
+      "loss": 0.0457233,
+      "step": 5780
+    },
+    {
+      "epoch": 11.562,
+      "grad_norm": 1.3299016952514648,
+      "learning_rate": 2e-05,
+      "loss": 0.04756445,
+      "step": 5781
+    },
+    {
+      "epoch": 11.564,
+      "grad_norm": 1.2130399942398071,
+      "learning_rate": 2e-05,
+      "loss": 0.0316097,
+      "step": 5782
+    },
+    {
+      "epoch": 11.566,
+      "grad_norm": 1.805086612701416,
+      "learning_rate": 2e-05,
+      "loss": 0.060819,
+      "step": 5783
+    },
+    {
+      "epoch": 11.568,
+      "grad_norm": 1.9846824407577515,
+      "learning_rate": 2e-05,
+      "loss": 0.04482648,
+      "step": 5784
+    },
+    {
+      "epoch": 11.57,
+      "grad_norm": 2.4053966999053955,
+      "learning_rate": 2e-05,
+      "loss": 0.04749952,
+      "step": 5785
+    },
+    {
+      "epoch": 11.572,
+      "grad_norm": 1.3613585233688354,
+      "learning_rate": 2e-05,
+      "loss": 0.04724826,
+      "step": 5786
+    },
+    {
+      "epoch": 11.574,
+      "grad_norm": 2.5251240730285645,
+      "learning_rate": 2e-05,
+      "loss": 0.03613253,
+      "step": 5787
+    },
+    {
+      "epoch": 11.576,
+      "grad_norm": 2.14925217628479,
+      "learning_rate": 2e-05,
+      "loss": 0.04898756,
+      "step": 5788
+    },
+    {
+      "epoch": 11.578,
+      "grad_norm": 1.379772424697876,
+      "learning_rate": 2e-05,
+      "loss": 0.05590723,
+      "step": 5789
+    },
+    {
+      "epoch": 11.58,
+      "grad_norm": 2.1570491790771484,
+      "learning_rate": 2e-05,
+      "loss": 0.06367472,
+      "step": 5790
+    },
+    {
+      "epoch": 11.582,
+      "grad_norm": 1.7565891742706299,
+      "learning_rate": 2e-05,
+      "loss": 0.05175413,
+      "step": 5791
+    },
+    {
+      "epoch": 11.584,
+      "grad_norm": 1.842274785041809,
+      "learning_rate": 2e-05,
+      "loss": 0.05350547,
+      "step": 5792
+    },
+    {
+      "epoch": 11.586,
+      "grad_norm": 2.831634283065796,
+      "learning_rate": 2e-05,
+      "loss": 0.06426252,
+      "step": 5793
+    },
+    {
+      "epoch": 11.588,
+      "grad_norm": 1.5474140644073486,
+      "learning_rate": 2e-05,
+      "loss": 0.04553394,
+      "step": 5794
+    },
+    {
+      "epoch": 11.59,
+      "grad_norm": 2.2714931964874268,
+      "learning_rate": 2e-05,
+      "loss": 0.02121895,
+      "step": 5795
+    },
+    {
+      "epoch": 11.592,
+      "grad_norm": 1.8448201417922974,
+      "learning_rate": 2e-05,
+      "loss": 0.0600131,
+      "step": 5796
+    },
+    {
+      "epoch": 11.594,
+      "grad_norm": 1.1388860940933228,
+      "learning_rate": 2e-05,
+      "loss": 0.03663793,
+      "step": 5797
+    },
+    {
+      "epoch": 11.596,
+      "grad_norm": 1.9692730903625488,
+      "learning_rate": 2e-05,
+      "loss": 0.04725102,
+      "step": 5798
+    },
+    {
+      "epoch": 11.598,
+      "grad_norm": 1.9636507034301758,
+      "learning_rate": 2e-05,
+      "loss": 0.04308451,
+      "step": 5799
+    },
+    {
+      "epoch": 11.6,
+      "grad_norm": 1.0663782358169556,
+      "learning_rate": 2e-05,
+      "loss": 0.0288805,
+      "step": 5800
+    },
+    {
+      "epoch": 11.602,
+      "grad_norm": 1.4374663829803467,
+      "learning_rate": 2e-05,
+      "loss": 0.04004197,
+      "step": 5801
+    },
+    {
+      "epoch": 11.604,
+      "grad_norm": 1.7417012453079224,
+      "learning_rate": 2e-05,
+      "loss": 0.05123825,
+      "step": 5802
+    },
+    {
+      "epoch": 11.606,
+      "grad_norm": 2.1125712394714355,
+      "learning_rate": 2e-05,
+      "loss": 0.0459488,
+      "step": 5803
+    },
+    {
+      "epoch": 11.608,
+      "grad_norm": 1.3598778247833252,
+      "learning_rate": 2e-05,
+      "loss": 0.03898705,
+      "step": 5804
+    },
+    {
+      "epoch": 11.61,
+      "grad_norm": 1.2977287769317627,
+      "learning_rate": 2e-05,
+      "loss": 0.05425401,
+      "step": 5805
+    },
+    {
+      "epoch": 11.612,
+      "grad_norm": 1.863133430480957,
+      "learning_rate": 2e-05,
+      "loss": 0.05389511,
+      "step": 5806
+    },
+    {
+      "epoch": 11.614,
+      "grad_norm": 2.131502389907837,
+      "learning_rate": 2e-05,
+      "loss": 0.05331743,
+      "step": 5807
+    },
+    {
+      "epoch": 11.616,
+      "grad_norm": 2.4704604148864746,
+      "learning_rate": 2e-05,
+      "loss": 0.05750574,
+      "step": 5808
+    },
+    {
+      "epoch": 11.618,
+      "grad_norm": 1.9358444213867188,
+      "learning_rate": 2e-05,
+      "loss": 0.04222788,
+      "step": 5809
+    },
+    {
+      "epoch": 11.62,
+      "grad_norm": 1.9284040927886963,
+      "learning_rate": 2e-05,
+      "loss": 0.04325092,
+      "step": 5810
+    },
+    {
+      "epoch": 11.622,
+      "grad_norm": 2.1048789024353027,
+      "learning_rate": 2e-05,
+      "loss": 0.05008987,
+      "step": 5811
+    },
+    {
+      "epoch": 11.624,
+      "grad_norm": 2.60477614402771,
+      "learning_rate": 2e-05,
+      "loss": 0.04310272,
+      "step": 5812
+    },
+    {
+      "epoch": 11.626,
+      "grad_norm": 1.3271856307983398,
+      "learning_rate": 2e-05,
+      "loss": 0.03580116,
+      "step": 5813
+    },
+    {
+      "epoch": 11.628,
+      "grad_norm": 1.6767809391021729,
+      "learning_rate": 2e-05,
+      "loss": 0.03502236,
+      "step": 5814
+    },
+    {
+      "epoch": 11.63,
+      "grad_norm": 1.5544588565826416,
+      "learning_rate": 2e-05,
+      "loss": 0.05321189,
+      "step": 5815
+    },
+    {
+      "epoch": 11.632,
+      "grad_norm": 1.3371565341949463,
+      "learning_rate": 2e-05,
+      "loss": 0.04384248,
+      "step": 5816
+    },
+    {
+      "epoch": 11.634,
+      "grad_norm": 1.6395615339279175,
+      "learning_rate": 2e-05,
+      "loss": 0.04614739,
+      "step": 5817
+    },
+    {
+      "epoch": 11.636,
+      "grad_norm": 2.506648302078247,
+      "learning_rate": 2e-05,
+      "loss": 0.06705838,
+      "step": 5818
+    },
+    {
+      "epoch": 11.638,
+      "grad_norm": 1.640137791633606,
+      "learning_rate": 2e-05,
+      "loss": 0.05612247,
+      "step": 5819
+    },
+    {
+      "epoch": 11.64,
+      "grad_norm": 2.4301540851593018,
+      "learning_rate": 2e-05,
+      "loss": 0.05863875,
+      "step": 5820
+    },
+    {
+      "epoch": 11.642,
+      "grad_norm": 1.5058404207229614,
+      "learning_rate": 2e-05,
+      "loss": 0.04215011,
+      "step": 5821
+    },
+    {
+      "epoch": 11.644,
+      "grad_norm": 1.2542659044265747,
+      "learning_rate": 2e-05,
+      "loss": 0.04177867,
+      "step": 5822
+    },
+    {
+      "epoch": 11.646,
+      "grad_norm": 1.107443928718567,
+      "learning_rate": 2e-05,
+      "loss": 0.04239887,
+      "step": 5823
+    },
+    {
+      "epoch": 11.648,
+      "grad_norm": 1.2233611345291138,
+      "learning_rate": 2e-05,
+      "loss": 0.04678968,
+      "step": 5824
+    },
+    {
+      "epoch": 11.65,
+      "grad_norm": 1.8364397287368774,
+      "learning_rate": 2e-05,
+      "loss": 0.04483078,
+      "step": 5825
+    },
+    {
+      "epoch": 11.652,
+      "grad_norm": 1.9486418962478638,
+      "learning_rate": 2e-05,
+      "loss": 0.04975449,
+      "step": 5826
+    },
+    {
+      "epoch": 11.654,
+      "grad_norm": 1.555195689201355,
+      "learning_rate": 2e-05,
+      "loss": 0.04872251,
+      "step": 5827
+    },
+    {
+      "epoch": 11.656,
+      "grad_norm": 1.7315301895141602,
+      "learning_rate": 2e-05,
+      "loss": 0.04223534,
+      "step": 5828
+    },
+    {
+      "epoch": 11.658,
+      "grad_norm": 1.342836856842041,
+      "learning_rate": 2e-05,
+      "loss": 0.06062792,
+      "step": 5829
+    },
+    {
+      "epoch": 11.66,
+      "grad_norm": 1.3641505241394043,
+      "learning_rate": 2e-05,
+      "loss": 0.05880754,
+      "step": 5830
+    },
+    {
+      "epoch": 11.662,
+      "grad_norm": 1.6247472763061523,
+      "learning_rate": 2e-05,
+      "loss": 0.05718165,
+      "step": 5831
+    },
+    {
+      "epoch": 11.664,
+      "grad_norm": 1.6310220956802368,
+      "learning_rate": 2e-05,
+      "loss": 0.0669639,
+      "step": 5832
+    },
+    {
+      "epoch": 11.666,
+      "grad_norm": 1.7226531505584717,
+      "learning_rate": 2e-05,
+      "loss": 0.04655602,
+      "step": 5833
+    },
+    {
+      "epoch": 11.668,
+      "grad_norm": 1.3371641635894775,
+      "learning_rate": 2e-05,
+      "loss": 0.05352715,
+      "step": 5834
+    },
+    {
+      "epoch": 11.67,
+      "grad_norm": 1.9057135581970215,
+      "learning_rate": 2e-05,
+      "loss": 0.06148591,
+      "step": 5835
+    },
+    {
+      "epoch": 11.672,
+      "grad_norm": 2.0866620540618896,
+      "learning_rate": 2e-05,
+      "loss": 0.04995357,
+      "step": 5836
+    },
+    {
+      "epoch": 11.674,
+      "grad_norm": 2.5532031059265137,
+      "learning_rate": 2e-05,
+      "loss": 0.07622005,
+      "step": 5837
+    },
+    {
+      "epoch": 11.676,
+      "grad_norm": 2.778372287750244,
+      "learning_rate": 2e-05,
+      "loss": 0.06307095,
+      "step": 5838
+    },
+    {
+      "epoch": 11.678,
+      "grad_norm": 1.43252432346344,
+      "learning_rate": 2e-05,
+      "loss": 0.04999834,
+      "step": 5839
+    },
+    {
+      "epoch": 11.68,
+      "grad_norm": 1.3844120502471924,
+      "learning_rate": 2e-05,
+      "loss": 0.05112729,
+      "step": 5840
+    },
+    {
+      "epoch": 11.682,
+      "grad_norm": 1.2430758476257324,
+      "learning_rate": 2e-05,
+      "loss": 0.04392055,
+      "step": 5841
+    },
+    {
+      "epoch": 11.684,
+      "grad_norm": 1.273295521736145,
+      "learning_rate": 2e-05,
+      "loss": 0.04273405,
+      "step": 5842
+    },
+    {
+      "epoch": 11.686,
+      "grad_norm": 1.3100574016571045,
+      "learning_rate": 2e-05,
+      "loss": 0.05339495,
+      "step": 5843
+    },
+    {
+      "epoch": 11.688,
+      "grad_norm": 1.267080545425415,
+      "learning_rate": 2e-05,
+      "loss": 0.0397922,
+      "step": 5844
+    },
+    {
+      "epoch": 11.69,
+      "grad_norm": 1.1280689239501953,
+      "learning_rate": 2e-05,
+      "loss": 0.04549348,
+      "step": 5845
+    },
+    {
+      "epoch": 11.692,
+      "grad_norm": 1.0839488506317139,
+      "learning_rate": 2e-05,
+      "loss": 0.03936727,
+      "step": 5846
+    },
+    {
+      "epoch": 11.693999999999999,
+      "grad_norm": 1.552826166152954,
+      "learning_rate": 2e-05,
+      "loss": 0.04919448,
+      "step": 5847
+    },
+    {
+      "epoch": 11.696,
+      "grad_norm": 1.6215623617172241,
+      "learning_rate": 2e-05,
+      "loss": 0.03126223,
+      "step": 5848
+    },
+    {
+      "epoch": 11.698,
+      "grad_norm": 1.1549961566925049,
+      "learning_rate": 2e-05,
+      "loss": 0.03589753,
+      "step": 5849
+    },
+    {
+      "epoch": 11.7,
+      "grad_norm": 1.782952070236206,
+      "learning_rate": 2e-05,
+      "loss": 0.0509676,
+      "step": 5850
+    },
+    {
+      "epoch": 11.702,
+      "grad_norm": 1.7008768320083618,
+      "learning_rate": 2e-05,
+      "loss": 0.07416837,
+      "step": 5851
+    },
+    {
+      "epoch": 11.704,
+      "grad_norm": 1.7450941801071167,
+      "learning_rate": 2e-05,
+      "loss": 0.04925626,
+      "step": 5852
+    },
+    {
+      "epoch": 11.706,
+      "grad_norm": 2.7231221199035645,
+      "learning_rate": 2e-05,
+      "loss": 0.06783025,
+      "step": 5853
+    },
+    {
+      "epoch": 11.708,
+      "grad_norm": 1.7462410926818848,
+      "learning_rate": 2e-05,
+      "loss": 0.05293039,
+      "step": 5854
+    },
+    {
+      "epoch": 11.71,
+      "grad_norm": 1.803440809249878,
+      "learning_rate": 2e-05,
+      "loss": 0.05988405,
+      "step": 5855
+    },
+    {
+      "epoch": 11.712,
+      "grad_norm": 1.4763165712356567,
+      "learning_rate": 2e-05,
+      "loss": 0.04566108,
+      "step": 5856
+    },
+    {
+      "epoch": 11.714,
+      "grad_norm": 1.44134521484375,
+      "learning_rate": 2e-05,
+      "loss": 0.04696288,
+      "step": 5857
+    },
+    {
+      "epoch": 11.716,
+      "grad_norm": 1.0796514749526978,
+      "learning_rate": 2e-05,
+      "loss": 0.03282233,
+      "step": 5858
+    },
+    {
+      "epoch": 11.718,
+      "grad_norm": 1.3339581489562988,
+      "learning_rate": 2e-05,
+      "loss": 0.04751251,
+      "step": 5859
+    },
+    {
+      "epoch": 11.72,
+      "grad_norm": 1.997147560119629,
+      "learning_rate": 2e-05,
+      "loss": 0.04504245,
+      "step": 5860
+    },
+    {
+      "epoch": 11.722,
+      "grad_norm": 1.335020899772644,
+      "learning_rate": 2e-05,
+      "loss": 0.04243895,
+      "step": 5861
+    },
+    {
+      "epoch": 11.724,
+      "grad_norm": 2.0907578468322754,
+      "learning_rate": 2e-05,
+      "loss": 0.04816663,
+      "step": 5862
+    },
+    {
+      "epoch": 11.725999999999999,
+      "grad_norm": 1.7296607494354248,
+      "learning_rate": 2e-05,
+      "loss": 0.04865206,
+      "step": 5863
+    },
+    {
+      "epoch": 11.728,
+      "grad_norm": 1.610846996307373,
+      "learning_rate": 2e-05,
+      "loss": 0.0545282,
+      "step": 5864
+    },
+    {
+      "epoch": 11.73,
+      "grad_norm": 1.4066860675811768,
+      "learning_rate": 2e-05,
+      "loss": 0.047715,
+      "step": 5865
+    },
+    {
+      "epoch": 11.732,
+      "grad_norm": 1.3445024490356445,
+      "learning_rate": 2e-05,
+      "loss": 0.04617265,
+      "step": 5866
+    },
+    {
+      "epoch": 11.734,
+      "grad_norm": 2.7535033226013184,
+      "learning_rate": 2e-05,
+      "loss": 0.04723052,
+      "step": 5867
+    },
+    {
+      "epoch": 11.736,
+      "grad_norm": 1.5304971933364868,
+      "learning_rate": 2e-05,
+      "loss": 0.03638636,
+      "step": 5868
+    },
+    {
+      "epoch": 11.738,
+      "grad_norm": 1.0276997089385986,
+      "learning_rate": 2e-05,
+      "loss": 0.03411288,
+      "step": 5869
+    },
+    {
+      "epoch": 11.74,
+      "grad_norm": 1.4574671983718872,
+      "learning_rate": 2e-05,
+      "loss": 0.05382932,
+      "step": 5870
+    },
+    {
+      "epoch": 11.742,
+      "grad_norm": 3.5009210109710693,
+      "learning_rate": 2e-05,
+      "loss": 0.06978729,
+      "step": 5871
+    },
+    {
+      "epoch": 11.744,
+      "grad_norm": 2.0347979068756104,
+      "learning_rate": 2e-05,
+      "loss": 0.0829041,
+      "step": 5872
+    },
+    {
+      "epoch": 11.746,
+      "grad_norm": 2.443451404571533,
+      "learning_rate": 2e-05,
+      "loss": 0.06375959,
+      "step": 5873
+    },
+    {
+      "epoch": 11.748,
+      "grad_norm": 2.023491859436035,
+      "learning_rate": 2e-05,
+      "loss": 0.06856895,
+      "step": 5874
+    },
+    {
+      "epoch": 11.75,
+      "grad_norm": 1.3537729978561401,
+      "learning_rate": 2e-05,
+      "loss": 0.04351505,
+      "step": 5875
+    },
+    {
+      "epoch": 11.752,
+      "grad_norm": 1.664186716079712,
+      "learning_rate": 2e-05,
+      "loss": 0.03763783,
+      "step": 5876
+    },
+    {
+      "epoch": 11.754,
+      "grad_norm": 1.5077100992202759,
+      "learning_rate": 2e-05,
+      "loss": 0.03997686,
+      "step": 5877
+    },
+    {
+      "epoch": 11.756,
+      "grad_norm": 2.3517305850982666,
+      "learning_rate": 2e-05,
+      "loss": 0.03420842,
+      "step": 5878
+    },
+    {
+      "epoch": 11.758,
+      "grad_norm": 1.4957901239395142,
+      "learning_rate": 2e-05,
+      "loss": 0.04347527,
+      "step": 5879
+    },
+    {
+      "epoch": 11.76,
+      "grad_norm": 1.5121381282806396,
+      "learning_rate": 2e-05,
+      "loss": 0.05459063,
+      "step": 5880
+    },
+    {
+      "epoch": 11.762,
+      "grad_norm": 1.957458734512329,
+      "learning_rate": 2e-05,
+      "loss": 0.05225588,
+      "step": 5881
+    },
+    {
+      "epoch": 11.764,
+      "grad_norm": 1.498754858970642,
+      "learning_rate": 2e-05,
+      "loss": 0.03588563,
+      "step": 5882
+    },
+    {
+      "epoch": 11.766,
+      "grad_norm": 1.5608701705932617,
+      "learning_rate": 2e-05,
+      "loss": 0.04653958,
+      "step": 5883
+    },
+    {
+      "epoch": 11.768,
+      "grad_norm": 2.1723520755767822,
+      "learning_rate": 2e-05,
+      "loss": 0.05516068,
+      "step": 5884
+    },
+    {
+      "epoch": 11.77,
+      "grad_norm": 2.607795000076294,
+      "learning_rate": 2e-05,
+      "loss": 0.05331718,
+      "step": 5885
+    },
+    {
+      "epoch": 11.772,
+      "grad_norm": 1.2783042192459106,
+      "learning_rate": 2e-05,
+      "loss": 0.04722802,
+      "step": 5886
+    },
+    {
+      "epoch": 11.774000000000001,
+      "grad_norm": 1.131067156791687,
+      "learning_rate": 2e-05,
+      "loss": 0.04217054,
+      "step": 5887
+    },
+    {
+      "epoch": 11.776,
+      "grad_norm": 2.056450366973877,
+      "learning_rate": 2e-05,
+      "loss": 0.04825941,
+      "step": 5888
+    },
+    {
+      "epoch": 11.778,
+      "grad_norm": 1.3246713876724243,
+      "learning_rate": 2e-05,
+      "loss": 0.0532944,
+      "step": 5889
+    },
+    {
+      "epoch": 11.78,
+      "grad_norm": 1.6754448413848877,
+      "learning_rate": 2e-05,
+      "loss": 0.05571581,
+      "step": 5890
+    },
+    {
+      "epoch": 11.782,
+      "grad_norm": 1.6864081621170044,
+      "learning_rate": 2e-05,
+      "loss": 0.04703726,
+      "step": 5891
+    },
+    {
+      "epoch": 11.784,
+      "grad_norm": 1.1856135129928589,
+      "learning_rate": 2e-05,
+      "loss": 0.03456947,
+      "step": 5892
+    },
+    {
+      "epoch": 11.786,
+      "grad_norm": 1.1907005310058594,
+      "learning_rate": 2e-05,
+      "loss": 0.03893632,
+      "step": 5893
+    },
+    {
+      "epoch": 11.788,
+      "grad_norm": 1.570899248123169,
+      "learning_rate": 2e-05,
+      "loss": 0.05421094,
+      "step": 5894
+    },
+    {
+      "epoch": 11.79,
+      "grad_norm": 2.6174049377441406,
+      "learning_rate": 2e-05,
+      "loss": 0.05405217,
+      "step": 5895
+    },
+    {
+      "epoch": 11.792,
+      "grad_norm": 1.9722440242767334,
+      "learning_rate": 2e-05,
+      "loss": 0.04381296,
+      "step": 5896
+    },
+    {
+      "epoch": 11.794,
+      "grad_norm": 1.0976735353469849,
+      "learning_rate": 2e-05,
+      "loss": 0.02918898,
+      "step": 5897
+    },
+    {
+      "epoch": 11.796,
+      "grad_norm": 1.4373865127563477,
+      "learning_rate": 2e-05,
+      "loss": 0.05200865,
+      "step": 5898
+    },
+    {
+      "epoch": 11.798,
+      "grad_norm": 1.3867244720458984,
+      "learning_rate": 2e-05,
+      "loss": 0.03751809,
+      "step": 5899
+    },
+    {
+      "epoch": 11.8,
+      "grad_norm": 1.449845314025879,
+      "learning_rate": 2e-05,
+      "loss": 0.04667427,
+      "step": 5900
+    },
+    {
+      "epoch": 11.802,
+      "grad_norm": 1.7023156881332397,
+      "learning_rate": 2e-05,
+      "loss": 0.03977741,
+      "step": 5901
+    },
+    {
+      "epoch": 11.804,
+      "grad_norm": 1.2221662998199463,
+      "learning_rate": 2e-05,
+      "loss": 0.04769532,
+      "step": 5902
+    },
+    {
+      "epoch": 11.806000000000001,
+      "grad_norm": 3.443532943725586,
+      "learning_rate": 2e-05,
+      "loss": 0.06289086,
+      "step": 5903
+    },
+    {
+      "epoch": 11.808,
+      "grad_norm": 1.2141717672348022,
+      "learning_rate": 2e-05,
+      "loss": 0.0353791,
+      "step": 5904
+    },
+    {
+      "epoch": 11.81,
+      "grad_norm": 1.126047134399414,
+      "learning_rate": 2e-05,
+      "loss": 0.03271578,
+      "step": 5905
+    },
+    {
+      "epoch": 11.812,
+      "grad_norm": 1.419371485710144,
+      "learning_rate": 2e-05,
+      "loss": 0.0543963,
+      "step": 5906
+    },
+    {
+      "epoch": 11.814,
+      "grad_norm": 2.132117748260498,
+      "learning_rate": 2e-05,
+      "loss": 0.06309062,
+      "step": 5907
+    },
+    {
+      "epoch": 11.816,
+      "grad_norm": 1.5774179697036743,
+      "learning_rate": 2e-05,
+      "loss": 0.05101613,
+      "step": 5908
+    },
+    {
+      "epoch": 11.818,
+      "grad_norm": 1.1032048463821411,
+      "learning_rate": 2e-05,
+      "loss": 0.03164658,
+      "step": 5909
+    },
+    {
+      "epoch": 11.82,
+      "grad_norm": 1.8389184474945068,
+      "learning_rate": 2e-05,
+      "loss": 0.05535435,
+      "step": 5910
+    },
+    {
+      "epoch": 11.822,
+      "grad_norm": 1.3901278972625732,
+      "learning_rate": 2e-05,
+      "loss": 0.04581359,
+      "step": 5911
+    },
+    {
+      "epoch": 11.824,
+      "grad_norm": 1.5085663795471191,
+      "learning_rate": 2e-05,
+      "loss": 0.04446094,
+      "step": 5912
+    },
+    {
+      "epoch": 11.826,
+      "grad_norm": 1.282586932182312,
+      "learning_rate": 2e-05,
+      "loss": 0.04857932,
+      "step": 5913
+    },
+    {
+      "epoch": 11.828,
+      "grad_norm": 1.1354501247406006,
+      "learning_rate": 2e-05,
+      "loss": 0.04103622,
+      "step": 5914
+    },
+    {
+      "epoch": 11.83,
+      "grad_norm": 1.4101886749267578,
+      "learning_rate": 2e-05,
+      "loss": 0.03275158,
+      "step": 5915
+    },
+    {
+      "epoch": 11.832,
+      "grad_norm": 1.4309121370315552,
+      "learning_rate": 2e-05,
+      "loss": 0.04146826,
+      "step": 5916
+    },
+    {
+      "epoch": 11.834,
+      "grad_norm": 1.0470571517944336,
+      "learning_rate": 2e-05,
+      "loss": 0.03013843,
+      "step": 5917
+    },
+    {
+      "epoch": 11.836,
+      "grad_norm": 2.17950177192688,
+      "learning_rate": 2e-05,
+      "loss": 0.06400843,
+      "step": 5918
+    },
+    {
+      "epoch": 11.838,
+      "grad_norm": 1.5850987434387207,
+      "learning_rate": 2e-05,
+      "loss": 0.04890703,
+      "step": 5919
+    },
+    {
+      "epoch": 11.84,
+      "grad_norm": 2.1424400806427,
+      "learning_rate": 2e-05,
+      "loss": 0.04672211,
+      "step": 5920
+    },
+    {
+      "epoch": 11.842,
+      "grad_norm": 1.6447819471359253,
+      "learning_rate": 2e-05,
+      "loss": 0.05776293,
+      "step": 5921
+    },
+    {
+      "epoch": 11.844,
+      "grad_norm": 1.2918273210525513,
+      "learning_rate": 2e-05,
+      "loss": 0.04792162,
+      "step": 5922
+    },
+    {
+      "epoch": 11.846,
+      "grad_norm": 1.787546157836914,
+      "learning_rate": 2e-05,
+      "loss": 0.04592175,
+      "step": 5923
+    },
+    {
+      "epoch": 11.848,
+      "grad_norm": 2.0610523223876953,
+      "learning_rate": 2e-05,
+      "loss": 0.0371214,
+      "step": 5924
+    },
+    {
+      "epoch": 11.85,
+      "grad_norm": 1.302527666091919,
+      "learning_rate": 2e-05,
+      "loss": 0.0393346,
+      "step": 5925
+    },
+    {
+      "epoch": 11.852,
+      "grad_norm": 1.6477103233337402,
+      "learning_rate": 2e-05,
+      "loss": 0.05504544,
+      "step": 5926
+    },
+    {
+      "epoch": 11.854,
+      "grad_norm": 1.8336002826690674,
+      "learning_rate": 2e-05,
+      "loss": 0.05120177,
+      "step": 5927
+    },
+    {
+      "epoch": 11.856,
+      "grad_norm": 1.318718671798706,
+      "learning_rate": 2e-05,
+      "loss": 0.05148219,
+      "step": 5928
+    },
+    {
+      "epoch": 11.858,
+      "grad_norm": 1.6090713739395142,
+      "learning_rate": 2e-05,
+      "loss": 0.05758989,
+      "step": 5929
+    },
+    {
+      "epoch": 11.86,
+      "grad_norm": 1.4209413528442383,
+      "learning_rate": 2e-05,
+      "loss": 0.03771012,
+      "step": 5930
+    },
+    {
+      "epoch": 11.862,
+      "grad_norm": 1.147481083869934,
+      "learning_rate": 2e-05,
+      "loss": 0.05305842,
+      "step": 5931
+    },
+    {
+      "epoch": 11.864,
+      "grad_norm": 5.9689130783081055,
+      "learning_rate": 2e-05,
+      "loss": 0.03842753,
+      "step": 5932
+    },
+    {
+      "epoch": 11.866,
+      "grad_norm": 2.704475164413452,
+      "learning_rate": 2e-05,
+      "loss": 0.04317836,
+      "step": 5933
+    },
+    {
+      "epoch": 11.868,
+      "grad_norm": 13.987378120422363,
+      "learning_rate": 2e-05,
+      "loss": 0.04295823,
+      "step": 5934
+    },
+    {
+      "epoch": 11.87,
+      "grad_norm": 2.0759031772613525,
+      "learning_rate": 2e-05,
+      "loss": 0.05019844,
+      "step": 5935
+    },
+    {
+      "epoch": 11.872,
+      "grad_norm": 2.631631851196289,
+      "learning_rate": 2e-05,
+      "loss": 0.04314917,
+      "step": 5936
+    },
+    {
+      "epoch": 11.874,
+      "grad_norm": 1.2975753545761108,
+      "learning_rate": 2e-05,
+      "loss": 0.0499692,
+      "step": 5937
+    },
+    {
+      "epoch": 11.876,
+      "grad_norm": 1.5606701374053955,
+      "learning_rate": 2e-05,
+      "loss": 0.04614861,
+      "step": 5938
+    },
+    {
+      "epoch": 11.878,
+      "grad_norm": 1.3647533655166626,
+      "learning_rate": 2e-05,
+      "loss": 0.03900645,
+      "step": 5939
+    },
+    {
+      "epoch": 11.88,
+      "grad_norm": 1.462010145187378,
+      "learning_rate": 2e-05,
+      "loss": 0.05590525,
+      "step": 5940
+    },
+    {
+      "epoch": 11.882,
+      "grad_norm": 2.939069986343384,
+      "learning_rate": 2e-05,
+      "loss": 0.06512713,
+      "step": 5941
+    },
+    {
+      "epoch": 11.884,
+      "grad_norm": 2.307407855987549,
+      "learning_rate": 2e-05,
+      "loss": 0.04771993,
+      "step": 5942
+    },
+    {
+      "epoch": 11.886,
+      "grad_norm": 1.784456491470337,
+      "learning_rate": 2e-05,
+      "loss": 0.05009575,
+      "step": 5943
+    },
+    {
+      "epoch": 11.888,
+      "grad_norm": 1.886611819267273,
+      "learning_rate": 2e-05,
+      "loss": 0.05641106,
+      "step": 5944
+    },
+    {
+      "epoch": 11.89,
+      "grad_norm": 1.1477409601211548,
+      "learning_rate": 2e-05,
+      "loss": 0.0363906,
+      "step": 5945
+    },
+    {
+      "epoch": 11.892,
+      "grad_norm": 1.2578458786010742,
+      "learning_rate": 2e-05,
+      "loss": 0.04756481,
+      "step": 5946
+    },
+    {
+      "epoch": 11.894,
+      "grad_norm": 1.938241958618164,
+      "learning_rate": 2e-05,
+      "loss": 0.0651107,
+      "step": 5947
+    },
+    {
+      "epoch": 11.896,
+      "grad_norm": 2.048572063446045,
+      "learning_rate": 2e-05,
+      "loss": 0.05753088,
+      "step": 5948
+    },
+    {
+      "epoch": 11.898,
+      "grad_norm": 1.759401559829712,
+      "learning_rate": 2e-05,
+      "loss": 0.0439103,
+      "step": 5949
+    },
+    {
+      "epoch": 11.9,
+      "grad_norm": 1.6436753273010254,
+      "learning_rate": 2e-05,
+      "loss": 0.04948621,
+      "step": 5950
+    },
+    {
+      "epoch": 11.902,
+      "grad_norm": 1.4181915521621704,
+      "learning_rate": 2e-05,
+      "loss": 0.0447426,
+      "step": 5951
+    },
+    {
+      "epoch": 11.904,
+      "grad_norm": 1.2882474660873413,
+      "learning_rate": 2e-05,
+      "loss": 0.05199945,
+      "step": 5952
+    },
+    {
+      "epoch": 11.906,
+      "grad_norm": 1.0682871341705322,
+      "learning_rate": 2e-05,
+      "loss": 0.03927372,
+      "step": 5953
+    },
+    {
+      "epoch": 11.908,
+      "grad_norm": 1.2409359216690063,
+      "learning_rate": 2e-05,
+      "loss": 0.05247876,
+      "step": 5954
+    },
+    {
+      "epoch": 11.91,
+      "grad_norm": 1.4313156604766846,
+      "learning_rate": 2e-05,
+      "loss": 0.04420187,
+      "step": 5955
+    },
+    {
+      "epoch": 11.912,
+      "grad_norm": 1.0682390928268433,
+      "learning_rate": 2e-05,
+      "loss": 0.03606921,
+      "step": 5956
+    },
+    {
+      "epoch": 11.914,
+      "grad_norm": 1.4402589797973633,
+      "learning_rate": 2e-05,
+      "loss": 0.0467343,
+      "step": 5957
+    },
+    {
+      "epoch": 11.916,
+      "grad_norm": 1.4133102893829346,
+      "learning_rate": 2e-05,
+      "loss": 0.04049528,
+      "step": 5958
+    },
+    {
+      "epoch": 11.918,
+      "grad_norm": 1.3268779516220093,
+      "learning_rate": 2e-05,
+      "loss": 0.06124356,
+      "step": 5959
+    },
+    {
+      "epoch": 11.92,
+      "grad_norm": 1.5016638040542603,
+      "learning_rate": 2e-05,
+      "loss": 0.04653203,
+      "step": 5960
+    },
+    {
+      "epoch": 11.922,
+      "grad_norm": 1.280086874961853,
+      "learning_rate": 2e-05,
+      "loss": 0.04292293,
+      "step": 5961
+    },
+    {
+      "epoch": 11.924,
+      "grad_norm": 1.5201990604400635,
+      "learning_rate": 2e-05,
+      "loss": 0.05881717,
+      "step": 5962
+    },
+    {
+      "epoch": 11.926,
+      "grad_norm": 2.3044865131378174,
+      "learning_rate": 2e-05,
+      "loss": 0.06573568,
+      "step": 5963
+    },
+    {
+      "epoch": 11.928,
+      "grad_norm": 1.8041393756866455,
+      "learning_rate": 2e-05,
+      "loss": 0.05998692,
+      "step": 5964
+    },
+    {
+      "epoch": 11.93,
+      "grad_norm": 1.9705407619476318,
+      "learning_rate": 2e-05,
+      "loss": 0.05393011,
+      "step": 5965
+    },
+    {
+      "epoch": 11.932,
+      "grad_norm": 1.5527033805847168,
+      "learning_rate": 2e-05,
+      "loss": 0.05403456,
+      "step": 5966
+    },
+    {
+      "epoch": 11.934,
+      "grad_norm": 1.115352749824524,
+      "learning_rate": 2e-05,
+      "loss": 0.03984626,
+      "step": 5967
+    },
+    {
+      "epoch": 11.936,
+      "grad_norm": 2.1990714073181152,
+      "learning_rate": 2e-05,
+      "loss": 0.05849539,
+      "step": 5968
+    },
+    {
+      "epoch": 11.938,
+      "grad_norm": 2.753941535949707,
+      "learning_rate": 2e-05,
+      "loss": 0.05485374,
+      "step": 5969
+    },
+    {
+      "epoch": 11.94,
+      "grad_norm": 1.1634223461151123,
+      "learning_rate": 2e-05,
+      "loss": 0.04611646,
+      "step": 5970
+    },
+    {
+      "epoch": 11.942,
+      "grad_norm": 1.3184393644332886,
+      "learning_rate": 2e-05,
+      "loss": 0.04198894,
+      "step": 5971
+    },
+    {
+      "epoch": 11.943999999999999,
+      "grad_norm": 1.2988426685333252,
+      "learning_rate": 2e-05,
+      "loss": 0.05265917,
+      "step": 5972
+    },
+    {
+      "epoch": 11.946,
+      "grad_norm": 1.7585318088531494,
+      "learning_rate": 2e-05,
+      "loss": 0.05654293,
+      "step": 5973
+    },
+    {
+      "epoch": 11.948,
+      "grad_norm": 1.6478968858718872,
+      "learning_rate": 2e-05,
+      "loss": 0.03530508,
+      "step": 5974
+    },
+    {
+      "epoch": 11.95,
+      "grad_norm": 2.3206682205200195,
+      "learning_rate": 2e-05,
+      "loss": 0.04617747,
+      "step": 5975
+    },
+    {
+      "epoch": 11.952,
+      "grad_norm": 1.2929635047912598,
+      "learning_rate": 2e-05,
+      "loss": 0.03368628,
+      "step": 5976
+    },
+    {
+      "epoch": 11.954,
+      "grad_norm": 1.047960877418518,
+      "learning_rate": 2e-05,
+      "loss": 0.03170536,
+      "step": 5977
+    },
+    {
+      "epoch": 11.956,
+      "grad_norm": 1.0071247816085815,
+      "learning_rate": 2e-05,
+      "loss": 0.02869906,
+      "step": 5978
+    },
+    {
+      "epoch": 11.958,
+      "grad_norm": 2.0578083992004395,
+      "learning_rate": 2e-05,
+      "loss": 0.04879463,
+      "step": 5979
+    },
+    {
+      "epoch": 11.96,
+      "grad_norm": 1.6822428703308105,
+      "learning_rate": 2e-05,
+      "loss": 0.0515471,
+      "step": 5980
+    },
+    {
+      "epoch": 11.962,
+      "grad_norm": 1.1344988346099854,
+      "learning_rate": 2e-05,
+      "loss": 0.03721001,
+      "step": 5981
+    },
+    {
+      "epoch": 11.964,
+      "grad_norm": 1.0933568477630615,
+      "learning_rate": 2e-05,
+      "loss": 0.03774923,
+      "step": 5982
+    },
+    {
+      "epoch": 11.966,
+      "grad_norm": 1.6839399337768555,
+      "learning_rate": 2e-05,
+      "loss": 0.03457904,
+      "step": 5983
+    },
+    {
+      "epoch": 11.968,
+      "grad_norm": 1.0689489841461182,
+      "learning_rate": 2e-05,
+      "loss": 0.0464288,
+      "step": 5984
+    },
+    {
+      "epoch": 11.97,
+      "grad_norm": 1.0576542615890503,
+      "learning_rate": 2e-05,
+      "loss": 0.03478286,
+      "step": 5985
+    },
+    {
+      "epoch": 11.972,
+      "grad_norm": 1.7853341102600098,
+      "learning_rate": 2e-05,
+      "loss": 0.05685054,
+      "step": 5986
+    },
+    {
+      "epoch": 11.974,
+      "grad_norm": 1.4036171436309814,
+      "learning_rate": 2e-05,
+      "loss": 0.04735323,
+      "step": 5987
+    },
+    {
+      "epoch": 11.975999999999999,
+      "grad_norm": 2.2815001010894775,
+      "learning_rate": 2e-05,
+      "loss": 0.06094185,
+      "step": 5988
+    },
+    {
+      "epoch": 11.978,
+      "grad_norm": 1.4660301208496094,
+      "learning_rate": 2e-05,
+      "loss": 0.04432792,
+      "step": 5989
+    },
+    {
+      "epoch": 11.98,
+      "grad_norm": 2.011078119277954,
+      "learning_rate": 2e-05,
+      "loss": 0.05223164,
+      "step": 5990
+    },
+    {
+      "epoch": 11.982,
+      "grad_norm": 1.2670975923538208,
+      "learning_rate": 2e-05,
+      "loss": 0.03780036,
+      "step": 5991
+    },
+    {
+      "epoch": 11.984,
+      "grad_norm": 1.5381466150283813,
+      "learning_rate": 2e-05,
+      "loss": 0.04062825,
+      "step": 5992
+    },
+    {
+      "epoch": 11.986,
+      "grad_norm": 1.1946245431900024,
+      "learning_rate": 2e-05,
+      "loss": 0.03976649,
+      "step": 5993
+    },
+    {
+      "epoch": 11.988,
+      "grad_norm": 2.6480839252471924,
+      "learning_rate": 2e-05,
+      "loss": 0.04942157,
+      "step": 5994
+    },
+    {
+      "epoch": 11.99,
+      "grad_norm": 1.3954662084579468,
+      "learning_rate": 2e-05,
+      "loss": 0.04232785,
+      "step": 5995
+    },
+    {
+      "epoch": 11.992,
+      "grad_norm": 1.0573922395706177,
+      "learning_rate": 2e-05,
+      "loss": 0.03846195,
+      "step": 5996
+    },
+    {
+      "epoch": 11.994,
+      "grad_norm": 1.5346850156784058,
+      "learning_rate": 2e-05,
+      "loss": 0.04501946,
+      "step": 5997
+    },
+    {
+      "epoch": 11.996,
+      "grad_norm": 1.567658543586731,
+      "learning_rate": 2e-05,
+      "loss": 0.04125125,
+      "step": 5998
+    },
+    {
+      "epoch": 11.998,
+      "grad_norm": 1.4279097318649292,
+      "learning_rate": 2e-05,
+      "loss": 0.03719701,
+      "step": 5999
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 1.4664994478225708,
+      "learning_rate": 2e-05,
+      "loss": 0.06088963,
+      "step": 6000
+    },
+    {
+      "epoch": 12.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.978,
+        "AngleClassification_2": 0.992,
+        "AngleClassification_3": 0.9540918163672655,
+        "Equal_1": 0.998,
+        "Equal_2": 0.9800399201596807,
+        "Equal_3": 0.906187624750499,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 0.998003992015968,
+        "LineComparison_3": 0.9920159680638723,
+        "Parallel_1": 0.9919839679358717,
+        "Parallel_2": 0.9979959919839679,
+        "Parallel_3": 0.992,
+        "Perpendicular_1": 0.986,
+        "Perpendicular_2": 0.946,
+        "Perpendicular_3": 0.6663326653306614,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 0.9976666666666667,
+        "PointLiesOnCircle_3": 0.9856666666666667,
+        "PointLiesOnLine_1": 0.9959919839679359,
+        "PointLiesOnLine_2": 0.9979959919839679,
+        "PointLiesOnLine_3": 0.9680638722554891
+      },
+      "eval_runtime": 319.0536,
+      "eval_samples_per_second": 32.91,
+      "eval_steps_per_second": 0.658,
+      "step": 6000
+    },
+    {
+      "epoch": 12.002,
+      "grad_norm": 2.38565731048584,
+      "learning_rate": 2e-05,
+      "loss": 0.0569401,
+      "step": 6001
+    },
+    {
+      "epoch": 12.004,
+      "grad_norm": 2.440852403640747,
+      "learning_rate": 2e-05,
+      "loss": 0.05189714,
+      "step": 6002
+    },
+    {
+      "epoch": 12.006,
+      "grad_norm": 1.0071848630905151,
+      "learning_rate": 2e-05,
+      "loss": 0.03481928,
+      "step": 6003
+    },
+    {
+      "epoch": 12.008,
+      "grad_norm": 1.0588632822036743,
+      "learning_rate": 2e-05,
+      "loss": 0.03610472,
+      "step": 6004
+    },
+    {
+      "epoch": 12.01,
+      "grad_norm": 1.5201109647750854,
+      "learning_rate": 2e-05,
+      "loss": 0.04367625,
+      "step": 6005
+    },
+    {
+      "epoch": 12.012,
+      "grad_norm": 1.7423279285430908,
+      "learning_rate": 2e-05,
+      "loss": 0.03819691,
+      "step": 6006
+    },
+    {
+      "epoch": 12.014,
+      "grad_norm": 1.3542839288711548,
+      "learning_rate": 2e-05,
+      "loss": 0.0486775,
+      "step": 6007
+    },
+    {
+      "epoch": 12.016,
+      "grad_norm": 1.939989447593689,
+      "learning_rate": 2e-05,
+      "loss": 0.05491658,
+      "step": 6008
+    },
+    {
+      "epoch": 12.018,
+      "grad_norm": 2.302846670150757,
+      "learning_rate": 2e-05,
+      "loss": 0.04849474,
+      "step": 6009
+    },
+    {
+      "epoch": 12.02,
+      "grad_norm": 1.8097832202911377,
+      "learning_rate": 2e-05,
+      "loss": 0.04880159,
+      "step": 6010
+    },
+    {
+      "epoch": 12.022,
+      "grad_norm": 0.9699602723121643,
+      "learning_rate": 2e-05,
+      "loss": 0.03032275,
+      "step": 6011
+    },
+    {
+      "epoch": 12.024,
+      "grad_norm": 1.2547293901443481,
+      "learning_rate": 2e-05,
+      "loss": 0.03023205,
+      "step": 6012
+    },
+    {
+      "epoch": 12.026,
+      "grad_norm": 1.7490140199661255,
+      "learning_rate": 2e-05,
+      "loss": 0.03436543,
+      "step": 6013
+    },
+    {
+      "epoch": 12.028,
+      "grad_norm": 1.5913435220718384,
+      "learning_rate": 2e-05,
+      "loss": 0.03803439,
+      "step": 6014
+    },
+    {
+      "epoch": 12.03,
+      "grad_norm": 1.5961428880691528,
+      "learning_rate": 2e-05,
+      "loss": 0.04362968,
+      "step": 6015
+    },
+    {
+      "epoch": 12.032,
+      "grad_norm": 1.9875961542129517,
+      "learning_rate": 2e-05,
+      "loss": 0.0427319,
+      "step": 6016
+    },
+    {
+      "epoch": 12.034,
+      "grad_norm": 1.6006630659103394,
+      "learning_rate": 2e-05,
+      "loss": 0.04221214,
+      "step": 6017
+    },
+    {
+      "epoch": 12.036,
+      "grad_norm": 1.985170602798462,
+      "learning_rate": 2e-05,
+      "loss": 0.04991886,
+      "step": 6018
+    },
+    {
+      "epoch": 12.038,
+      "grad_norm": 1.3990066051483154,
+      "learning_rate": 2e-05,
+      "loss": 0.0360059,
+      "step": 6019
+    },
+    {
+      "epoch": 12.04,
+      "grad_norm": 1.732647180557251,
+      "learning_rate": 2e-05,
+      "loss": 0.03687097,
+      "step": 6020
+    },
+    {
+      "epoch": 12.042,
+      "grad_norm": 1.6655960083007812,
+      "learning_rate": 2e-05,
+      "loss": 0.05822805,
+      "step": 6021
+    },
+    {
+      "epoch": 12.044,
+      "grad_norm": 1.6695098876953125,
+      "learning_rate": 2e-05,
+      "loss": 0.04767611,
+      "step": 6022
+    },
+    {
+      "epoch": 12.046,
+      "grad_norm": 1.7412924766540527,
+      "learning_rate": 2e-05,
+      "loss": 0.04930716,
+      "step": 6023
+    },
+    {
+      "epoch": 12.048,
+      "grad_norm": 1.7823171615600586,
+      "learning_rate": 2e-05,
+      "loss": 0.03755423,
+      "step": 6024
+    },
+    {
+      "epoch": 12.05,
+      "grad_norm": 1.2226248979568481,
+      "learning_rate": 2e-05,
+      "loss": 0.04179034,
+      "step": 6025
+    },
+    {
+      "epoch": 12.052,
+      "grad_norm": 1.5548951625823975,
+      "learning_rate": 2e-05,
+      "loss": 0.03161469,
+      "step": 6026
+    },
+    {
+      "epoch": 12.054,
+      "grad_norm": 1.9556974172592163,
+      "learning_rate": 2e-05,
+      "loss": 0.04450768,
+      "step": 6027
+    },
+    {
+      "epoch": 12.056,
+      "grad_norm": 1.177003026008606,
+      "learning_rate": 2e-05,
+      "loss": 0.04270407,
+      "step": 6028
+    },
+    {
+      "epoch": 12.058,
+      "grad_norm": 1.3655465841293335,
+      "learning_rate": 2e-05,
+      "loss": 0.05215304,
+      "step": 6029
+    },
+    {
+      "epoch": 12.06,
+      "grad_norm": 3.6705446243286133,
+      "learning_rate": 2e-05,
+      "loss": 0.05514611,
+      "step": 6030
+    },
+    {
+      "epoch": 12.062,
+      "grad_norm": 7.220978260040283,
+      "learning_rate": 2e-05,
+      "loss": 0.05021281,
+      "step": 6031
+    },
+    {
+      "epoch": 12.064,
+      "grad_norm": 1.2202095985412598,
+      "learning_rate": 2e-05,
+      "loss": 0.03365953,
+      "step": 6032
+    },
+    {
+      "epoch": 12.066,
+      "grad_norm": 1.2675375938415527,
+      "learning_rate": 2e-05,
+      "loss": 0.03978067,
+      "step": 6033
+    },
+    {
+      "epoch": 12.068,
+      "grad_norm": 1.05901300907135,
+      "learning_rate": 2e-05,
+      "loss": 0.03192661,
+      "step": 6034
+    },
+    {
+      "epoch": 12.07,
+      "grad_norm": 1.8929624557495117,
+      "learning_rate": 2e-05,
+      "loss": 0.03948138,
+      "step": 6035
+    },
+    {
+      "epoch": 12.072,
+      "grad_norm": 1.3291373252868652,
+      "learning_rate": 2e-05,
+      "loss": 0.04005025,
+      "step": 6036
+    },
+    {
+      "epoch": 12.074,
+      "grad_norm": 2.3758695125579834,
+      "learning_rate": 2e-05,
+      "loss": 0.04173733,
+      "step": 6037
+    },
+    {
+      "epoch": 12.076,
+      "grad_norm": 2.500800848007202,
+      "learning_rate": 2e-05,
+      "loss": 0.05645321,
+      "step": 6038
+    },
+    {
+      "epoch": 12.078,
+      "grad_norm": 1.4360809326171875,
+      "learning_rate": 2e-05,
+      "loss": 0.04782884,
+      "step": 6039
+    },
+    {
+      "epoch": 12.08,
+      "grad_norm": 2.3848273754119873,
+      "learning_rate": 2e-05,
+      "loss": 0.04527626,
+      "step": 6040
+    },
+    {
+      "epoch": 12.082,
+      "grad_norm": 1.747944951057434,
+      "learning_rate": 2e-05,
+      "loss": 0.07185665,
+      "step": 6041
+    },
+    {
+      "epoch": 12.084,
+      "grad_norm": 3.546881675720215,
+      "learning_rate": 2e-05,
+      "loss": 0.05210949,
+      "step": 6042
+    },
+    {
+      "epoch": 12.086,
+      "grad_norm": 2.6596226692199707,
+      "learning_rate": 2e-05,
+      "loss": 0.04689426,
+      "step": 6043
+    },
+    {
+      "epoch": 12.088,
+      "grad_norm": 2.573843479156494,
+      "learning_rate": 2e-05,
+      "loss": 0.05877293,
+      "step": 6044
+    },
+    {
+      "epoch": 12.09,
+      "grad_norm": 2.237309694290161,
+      "learning_rate": 2e-05,
+      "loss": 0.03652829,
+      "step": 6045
+    },
+    {
+      "epoch": 12.092,
+      "grad_norm": 1.193651795387268,
+      "learning_rate": 2e-05,
+      "loss": 0.03118789,
+      "step": 6046
+    },
+    {
+      "epoch": 12.094,
+      "grad_norm": 2.136948823928833,
+      "learning_rate": 2e-05,
+      "loss": 0.03762619,
+      "step": 6047
+    },
+    {
+      "epoch": 12.096,
+      "grad_norm": 1.9296658039093018,
+      "learning_rate": 2e-05,
+      "loss": 0.05173202,
+      "step": 6048
+    },
+    {
+      "epoch": 12.098,
+      "grad_norm": 1.0394606590270996,
+      "learning_rate": 2e-05,
+      "loss": 0.03266693,
+      "step": 6049
+    },
+    {
+      "epoch": 12.1,
+      "grad_norm": 1.249806523323059,
+      "learning_rate": 2e-05,
+      "loss": 0.03473232,
+      "step": 6050
+    },
+    {
+      "epoch": 12.102,
+      "grad_norm": 1.4066314697265625,
+      "learning_rate": 2e-05,
+      "loss": 0.04402903,
+      "step": 6051
+    },
+    {
+      "epoch": 12.104,
+      "grad_norm": 1.833040475845337,
+      "learning_rate": 2e-05,
+      "loss": 0.03956625,
+      "step": 6052
+    },
+    {
+      "epoch": 12.106,
+      "grad_norm": 1.421608805656433,
+      "learning_rate": 2e-05,
+      "loss": 0.04748692,
+      "step": 6053
+    },
+    {
+      "epoch": 12.108,
+      "grad_norm": 2.0671706199645996,
+      "learning_rate": 2e-05,
+      "loss": 0.04925867,
+      "step": 6054
+    },
+    {
+      "epoch": 12.11,
+      "grad_norm": 1.106614589691162,
+      "learning_rate": 2e-05,
+      "loss": 0.04091745,
+      "step": 6055
+    },
+    {
+      "epoch": 12.112,
+      "grad_norm": 1.2263094186782837,
+      "learning_rate": 2e-05,
+      "loss": 0.04056298,
+      "step": 6056
+    },
+    {
+      "epoch": 12.114,
+      "grad_norm": 3.7029409408569336,
+      "learning_rate": 2e-05,
+      "loss": 0.05836046,
+      "step": 6057
+    },
+    {
+      "epoch": 12.116,
+      "grad_norm": 1.413335919380188,
+      "learning_rate": 2e-05,
+      "loss": 0.05062287,
+      "step": 6058
+    },
+    {
+      "epoch": 12.118,
+      "grad_norm": 3.4559438228607178,
+      "learning_rate": 2e-05,
+      "loss": 0.0437443,
+      "step": 6059
+    },
+    {
+      "epoch": 12.12,
+      "grad_norm": 1.7302478551864624,
+      "learning_rate": 2e-05,
+      "loss": 0.05316646,
+      "step": 6060
+    },
+    {
+      "epoch": 12.122,
+      "grad_norm": 1.5175634622573853,
+      "learning_rate": 2e-05,
+      "loss": 0.04142622,
+      "step": 6061
+    },
+    {
+      "epoch": 12.124,
+      "grad_norm": 1.6398677825927734,
+      "learning_rate": 2e-05,
+      "loss": 0.05954774,
+      "step": 6062
+    },
+    {
+      "epoch": 12.126,
+      "grad_norm": 1.5120245218276978,
+      "learning_rate": 2e-05,
+      "loss": 0.05301241,
+      "step": 6063
+    },
+    {
+      "epoch": 12.128,
+      "grad_norm": 1.1674778461456299,
+      "learning_rate": 2e-05,
+      "loss": 0.04489746,
+      "step": 6064
+    },
+    {
+      "epoch": 12.13,
+      "grad_norm": 1.299045205116272,
+      "learning_rate": 2e-05,
+      "loss": 0.03792151,
+      "step": 6065
+    },
+    {
+      "epoch": 12.132,
+      "grad_norm": 0.9443352222442627,
+      "learning_rate": 2e-05,
+      "loss": 0.02998193,
+      "step": 6066
+    },
+    {
+      "epoch": 12.134,
+      "grad_norm": 1.6018176078796387,
+      "learning_rate": 2e-05,
+      "loss": 0.04062822,
+      "step": 6067
+    },
+    {
+      "epoch": 12.136,
+      "grad_norm": 2.1333489418029785,
+      "learning_rate": 2e-05,
+      "loss": 0.03400786,
+      "step": 6068
+    },
+    {
+      "epoch": 12.138,
+      "grad_norm": 1.8581465482711792,
+      "learning_rate": 2e-05,
+      "loss": 0.05270678,
+      "step": 6069
+    },
+    {
+      "epoch": 12.14,
+      "grad_norm": 1.1669179201126099,
+      "learning_rate": 2e-05,
+      "loss": 0.04782745,
+      "step": 6070
+    },
+    {
+      "epoch": 12.142,
+      "grad_norm": 2.276296615600586,
+      "learning_rate": 2e-05,
+      "loss": 0.03568403,
+      "step": 6071
+    },
+    {
+      "epoch": 12.144,
+      "grad_norm": 1.1378567218780518,
+      "learning_rate": 2e-05,
+      "loss": 0.03819509,
+      "step": 6072
+    },
+    {
+      "epoch": 12.146,
+      "grad_norm": 2.1486968994140625,
+      "learning_rate": 2e-05,
+      "loss": 0.04405282,
+      "step": 6073
+    },
+    {
+      "epoch": 12.148,
+      "grad_norm": 1.2601561546325684,
+      "learning_rate": 2e-05,
+      "loss": 0.03810599,
+      "step": 6074
+    },
+    {
+      "epoch": 12.15,
+      "grad_norm": 1.7045177221298218,
+      "learning_rate": 2e-05,
+      "loss": 0.04431679,
+      "step": 6075
+    },
+    {
+      "epoch": 12.152,
+      "grad_norm": 1.4039207696914673,
+      "learning_rate": 2e-05,
+      "loss": 0.05113788,
+      "step": 6076
+    },
+    {
+      "epoch": 12.154,
+      "grad_norm": 1.8916300535202026,
+      "learning_rate": 2e-05,
+      "loss": 0.0379601,
+      "step": 6077
+    },
+    {
+      "epoch": 12.156,
+      "grad_norm": 1.337270736694336,
+      "learning_rate": 2e-05,
+      "loss": 0.04206997,
+      "step": 6078
+    },
+    {
+      "epoch": 12.158,
+      "grad_norm": 1.3554550409317017,
+      "learning_rate": 2e-05,
+      "loss": 0.04288577,
+      "step": 6079
+    },
+    {
+      "epoch": 12.16,
+      "grad_norm": 1.9474183320999146,
+      "learning_rate": 2e-05,
+      "loss": 0.04348049,
+      "step": 6080
+    },
+    {
+      "epoch": 12.162,
+      "grad_norm": 0.8301889300346375,
+      "learning_rate": 2e-05,
+      "loss": 0.02466136,
+      "step": 6081
+    },
+    {
+      "epoch": 12.164,
+      "grad_norm": 1.5788853168487549,
+      "learning_rate": 2e-05,
+      "loss": 0.05005291,
+      "step": 6082
+    },
+    {
+      "epoch": 12.166,
+      "grad_norm": 2.1936469078063965,
+      "learning_rate": 2e-05,
+      "loss": 0.05456689,
+      "step": 6083
+    },
+    {
+      "epoch": 12.168,
+      "grad_norm": 1.8480530977249146,
+      "learning_rate": 2e-05,
+      "loss": 0.05015424,
+      "step": 6084
+    },
+    {
+      "epoch": 12.17,
+      "grad_norm": 1.3784325122833252,
+      "learning_rate": 2e-05,
+      "loss": 0.03511147,
+      "step": 6085
+    },
+    {
+      "epoch": 12.172,
+      "grad_norm": 1.4679827690124512,
+      "learning_rate": 2e-05,
+      "loss": 0.028723,
+      "step": 6086
+    },
+    {
+      "epoch": 12.174,
+      "grad_norm": 1.677220344543457,
+      "learning_rate": 2e-05,
+      "loss": 0.04889953,
+      "step": 6087
+    },
+    {
+      "epoch": 12.176,
+      "grad_norm": 1.5468508005142212,
+      "learning_rate": 2e-05,
+      "loss": 0.04408753,
+      "step": 6088
+    },
+    {
+      "epoch": 12.178,
+      "grad_norm": 1.36641526222229,
+      "learning_rate": 2e-05,
+      "loss": 0.03904226,
+      "step": 6089
+    },
+    {
+      "epoch": 12.18,
+      "grad_norm": 2.0277597904205322,
+      "learning_rate": 2e-05,
+      "loss": 0.03122667,
+      "step": 6090
+    },
+    {
+      "epoch": 12.182,
+      "grad_norm": 2.095195770263672,
+      "learning_rate": 2e-05,
+      "loss": 0.04639265,
+      "step": 6091
+    },
+    {
+      "epoch": 12.184,
+      "grad_norm": 1.4929248094558716,
+      "learning_rate": 2e-05,
+      "loss": 0.04378875,
+      "step": 6092
+    },
+    {
+      "epoch": 12.186,
+      "grad_norm": 1.040937066078186,
+      "learning_rate": 2e-05,
+      "loss": 0.02706816,
+      "step": 6093
+    },
+    {
+      "epoch": 12.188,
+      "grad_norm": 1.5594635009765625,
+      "learning_rate": 2e-05,
+      "loss": 0.02982716,
+      "step": 6094
+    },
+    {
+      "epoch": 12.19,
+      "grad_norm": 1.0690845251083374,
+      "learning_rate": 2e-05,
+      "loss": 0.02932034,
+      "step": 6095
+    },
+    {
+      "epoch": 12.192,
+      "grad_norm": 1.3992670774459839,
+      "learning_rate": 2e-05,
+      "loss": 0.04999903,
+      "step": 6096
+    },
+    {
+      "epoch": 12.194,
+      "grad_norm": 1.2843531370162964,
+      "learning_rate": 2e-05,
+      "loss": 0.02511716,
+      "step": 6097
+    },
+    {
+      "epoch": 12.196,
+      "grad_norm": 1.8418145179748535,
+      "learning_rate": 2e-05,
+      "loss": 0.04045618,
+      "step": 6098
+    },
+    {
+      "epoch": 12.198,
+      "grad_norm": 1.3255176544189453,
+      "learning_rate": 2e-05,
+      "loss": 0.04162053,
+      "step": 6099
+    },
+    {
+      "epoch": 12.2,
+      "grad_norm": 1.1222180128097534,
+      "learning_rate": 2e-05,
+      "loss": 0.02524855,
+      "step": 6100
+    },
+    {
+      "epoch": 12.202,
+      "grad_norm": 1.5370876789093018,
+      "learning_rate": 2e-05,
+      "loss": 0.04771005,
+      "step": 6101
+    },
+    {
+      "epoch": 12.204,
+      "grad_norm": 1.1293209791183472,
+      "learning_rate": 2e-05,
+      "loss": 0.03034871,
+      "step": 6102
+    },
+    {
+      "epoch": 12.206,
+      "grad_norm": 1.13023042678833,
+      "learning_rate": 2e-05,
+      "loss": 0.02627375,
+      "step": 6103
+    },
+    {
+      "epoch": 12.208,
+      "grad_norm": 1.3612079620361328,
+      "learning_rate": 2e-05,
+      "loss": 0.04360874,
+      "step": 6104
+    },
+    {
+      "epoch": 12.21,
+      "grad_norm": 3.503488779067993,
+      "learning_rate": 2e-05,
+      "loss": 0.03669972,
+      "step": 6105
+    },
+    {
+      "epoch": 12.212,
+      "grad_norm": 2.4305996894836426,
+      "learning_rate": 2e-05,
+      "loss": 0.05017614,
+      "step": 6106
+    },
+    {
+      "epoch": 12.214,
+      "grad_norm": 1.4109909534454346,
+      "learning_rate": 2e-05,
+      "loss": 0.03262326,
+      "step": 6107
+    },
+    {
+      "epoch": 12.216,
+      "grad_norm": 1.8456965684890747,
+      "learning_rate": 2e-05,
+      "loss": 0.0493413,
+      "step": 6108
+    },
+    {
+      "epoch": 12.218,
+      "grad_norm": 1.6295368671417236,
+      "learning_rate": 2e-05,
+      "loss": 0.04406694,
+      "step": 6109
+    },
+    {
+      "epoch": 12.22,
+      "grad_norm": 1.2703592777252197,
+      "learning_rate": 2e-05,
+      "loss": 0.04533841,
+      "step": 6110
+    },
+    {
+      "epoch": 12.222,
+      "grad_norm": 3.6067559719085693,
+      "learning_rate": 2e-05,
+      "loss": 0.05607805,
+      "step": 6111
+    },
+    {
+      "epoch": 12.224,
+      "grad_norm": 1.2701184749603271,
+      "learning_rate": 2e-05,
+      "loss": 0.04785171,
+      "step": 6112
+    },
+    {
+      "epoch": 12.226,
+      "grad_norm": 1.1288963556289673,
+      "learning_rate": 2e-05,
+      "loss": 0.04206613,
+      "step": 6113
+    },
+    {
+      "epoch": 12.228,
+      "grad_norm": 1.6207784414291382,
+      "learning_rate": 2e-05,
+      "loss": 0.04109996,
+      "step": 6114
+    },
+    {
+      "epoch": 12.23,
+      "grad_norm": 1.5829488039016724,
+      "learning_rate": 2e-05,
+      "loss": 0.03511421,
+      "step": 6115
+    },
+    {
+      "epoch": 12.232,
+      "grad_norm": 1.4869654178619385,
+      "learning_rate": 2e-05,
+      "loss": 0.04815879,
+      "step": 6116
+    },
+    {
+      "epoch": 12.234,
+      "grad_norm": 1.412255048751831,
+      "learning_rate": 2e-05,
+      "loss": 0.0408961,
+      "step": 6117
+    },
+    {
+      "epoch": 12.236,
+      "grad_norm": 0.9086294174194336,
+      "learning_rate": 2e-05,
+      "loss": 0.03307983,
+      "step": 6118
+    },
+    {
+      "epoch": 12.238,
+      "grad_norm": 1.4410475492477417,
+      "learning_rate": 2e-05,
+      "loss": 0.04874083,
+      "step": 6119
+    },
+    {
+      "epoch": 12.24,
+      "grad_norm": 2.1542203426361084,
+      "learning_rate": 2e-05,
+      "loss": 0.03492495,
+      "step": 6120
+    },
+    {
+      "epoch": 12.242,
+      "grad_norm": 1.4290752410888672,
+      "learning_rate": 2e-05,
+      "loss": 0.03959789,
+      "step": 6121
+    },
+    {
+      "epoch": 12.244,
+      "grad_norm": 3.0469701290130615,
+      "learning_rate": 2e-05,
+      "loss": 0.07812107,
+      "step": 6122
+    },
+    {
+      "epoch": 12.246,
+      "grad_norm": 1.6953679323196411,
+      "learning_rate": 2e-05,
+      "loss": 0.02141916,
+      "step": 6123
+    },
+    {
+      "epoch": 12.248,
+      "grad_norm": 1.8195632696151733,
+      "learning_rate": 2e-05,
+      "loss": 0.0405052,
+      "step": 6124
+    },
+    {
+      "epoch": 12.25,
+      "grad_norm": 2.004796266555786,
+      "learning_rate": 2e-05,
+      "loss": 0.03984334,
+      "step": 6125
+    },
+    {
+      "epoch": 12.252,
+      "grad_norm": 0.8493677973747253,
+      "learning_rate": 2e-05,
+      "loss": 0.03030949,
+      "step": 6126
+    },
+    {
+      "epoch": 12.254,
+      "grad_norm": 1.603154182434082,
+      "learning_rate": 2e-05,
+      "loss": 0.03452389,
+      "step": 6127
+    },
+    {
+      "epoch": 12.256,
+      "grad_norm": 1.4318041801452637,
+      "learning_rate": 2e-05,
+      "loss": 0.05320135,
+      "step": 6128
+    },
+    {
+      "epoch": 12.258,
+      "grad_norm": 1.305381178855896,
+      "learning_rate": 2e-05,
+      "loss": 0.04058621,
+      "step": 6129
+    },
+    {
+      "epoch": 12.26,
+      "grad_norm": 1.0062555074691772,
+      "learning_rate": 2e-05,
+      "loss": 0.03081638,
+      "step": 6130
+    },
+    {
+      "epoch": 12.262,
+      "grad_norm": 1.51310133934021,
+      "learning_rate": 2e-05,
+      "loss": 0.03336924,
+      "step": 6131
+    },
+    {
+      "epoch": 12.264,
+      "grad_norm": 1.28446626663208,
+      "learning_rate": 2e-05,
+      "loss": 0.03799494,
+      "step": 6132
+    },
+    {
+      "epoch": 12.266,
+      "grad_norm": 2.550501823425293,
+      "learning_rate": 2e-05,
+      "loss": 0.05581835,
+      "step": 6133
+    },
+    {
+      "epoch": 12.268,
+      "grad_norm": 1.476683259010315,
+      "learning_rate": 2e-05,
+      "loss": 0.04637709,
+      "step": 6134
+    },
+    {
+      "epoch": 12.27,
+      "grad_norm": 2.046898126602173,
+      "learning_rate": 2e-05,
+      "loss": 0.04047813,
+      "step": 6135
+    },
+    {
+      "epoch": 12.272,
+      "grad_norm": 1.6879256963729858,
+      "learning_rate": 2e-05,
+      "loss": 0.050128,
+      "step": 6136
+    },
+    {
+      "epoch": 12.274000000000001,
+      "grad_norm": 1.1848565340042114,
+      "learning_rate": 2e-05,
+      "loss": 0.03365932,
+      "step": 6137
+    },
+    {
+      "epoch": 12.276,
+      "grad_norm": 1.4317147731781006,
+      "learning_rate": 2e-05,
+      "loss": 0.04208836,
+      "step": 6138
+    },
+    {
+      "epoch": 12.278,
+      "grad_norm": 0.8532785773277283,
+      "learning_rate": 2e-05,
+      "loss": 0.02701728,
+      "step": 6139
+    },
+    {
+      "epoch": 12.28,
+      "grad_norm": 2.347813129425049,
+      "learning_rate": 2e-05,
+      "loss": 0.05657571,
+      "step": 6140
+    },
+    {
+      "epoch": 12.282,
+      "grad_norm": 1.616212248802185,
+      "learning_rate": 2e-05,
+      "loss": 0.0320894,
+      "step": 6141
+    },
+    {
+      "epoch": 12.284,
+      "grad_norm": 1.4436291456222534,
+      "learning_rate": 2e-05,
+      "loss": 0.03790341,
+      "step": 6142
+    },
+    {
+      "epoch": 12.286,
+      "grad_norm": 1.915105938911438,
+      "learning_rate": 2e-05,
+      "loss": 0.04711257,
+      "step": 6143
+    },
+    {
+      "epoch": 12.288,
+      "grad_norm": 1.6643471717834473,
+      "learning_rate": 2e-05,
+      "loss": 0.02922626,
+      "step": 6144
+    },
+    {
+      "epoch": 12.29,
+      "grad_norm": 1.9651648998260498,
+      "learning_rate": 2e-05,
+      "loss": 0.06149007,
+      "step": 6145
+    },
+    {
+      "epoch": 12.292,
+      "grad_norm": 1.5634087324142456,
+      "learning_rate": 2e-05,
+      "loss": 0.03441588,
+      "step": 6146
+    },
+    {
+      "epoch": 12.294,
+      "grad_norm": 3.6139161586761475,
+      "learning_rate": 2e-05,
+      "loss": 0.0543981,
+      "step": 6147
+    },
+    {
+      "epoch": 12.296,
+      "grad_norm": 1.771533489227295,
+      "learning_rate": 2e-05,
+      "loss": 0.04886418,
+      "step": 6148
+    },
+    {
+      "epoch": 12.298,
+      "grad_norm": 1.4337700605392456,
+      "learning_rate": 2e-05,
+      "loss": 0.02946185,
+      "step": 6149
+    },
+    {
+      "epoch": 12.3,
+      "grad_norm": 1.214970350265503,
+      "learning_rate": 2e-05,
+      "loss": 0.03541949,
+      "step": 6150
+    },
+    {
+      "epoch": 12.302,
+      "grad_norm": 1.8696826696395874,
+      "learning_rate": 2e-05,
+      "loss": 0.04219283,
+      "step": 6151
+    },
+    {
+      "epoch": 12.304,
+      "grad_norm": 1.1375106573104858,
+      "learning_rate": 2e-05,
+      "loss": 0.02643434,
+      "step": 6152
+    },
+    {
+      "epoch": 12.306,
+      "grad_norm": 1.9617277383804321,
+      "learning_rate": 2e-05,
+      "loss": 0.04668794,
+      "step": 6153
+    },
+    {
+      "epoch": 12.308,
+      "grad_norm": 1.219461441040039,
+      "learning_rate": 2e-05,
+      "loss": 0.02614435,
+      "step": 6154
+    },
+    {
+      "epoch": 12.31,
+      "grad_norm": 2.0026473999023438,
+      "learning_rate": 2e-05,
+      "loss": 0.05046513,
+      "step": 6155
+    },
+    {
+      "epoch": 12.312,
+      "grad_norm": 2.2148213386535645,
+      "learning_rate": 2e-05,
+      "loss": 0.03219437,
+      "step": 6156
+    },
+    {
+      "epoch": 12.314,
+      "grad_norm": 1.3021752834320068,
+      "learning_rate": 2e-05,
+      "loss": 0.03959469,
+      "step": 6157
+    },
+    {
+      "epoch": 12.316,
+      "grad_norm": 2.0185019969940186,
+      "learning_rate": 2e-05,
+      "loss": 0.04784639,
+      "step": 6158
+    },
+    {
+      "epoch": 12.318,
+      "grad_norm": 1.835060954093933,
+      "learning_rate": 2e-05,
+      "loss": 0.04800402,
+      "step": 6159
+    },
+    {
+      "epoch": 12.32,
+      "grad_norm": 1.1471303701400757,
+      "learning_rate": 2e-05,
+      "loss": 0.04062325,
+      "step": 6160
+    },
+    {
+      "epoch": 12.322,
+      "grad_norm": 1.5545427799224854,
+      "learning_rate": 2e-05,
+      "loss": 0.05014087,
+      "step": 6161
+    },
+    {
+      "epoch": 12.324,
+      "grad_norm": 1.3972580432891846,
+      "learning_rate": 2e-05,
+      "loss": 0.03715761,
+      "step": 6162
+    },
+    {
+      "epoch": 12.326,
+      "grad_norm": 1.1503256559371948,
+      "learning_rate": 2e-05,
+      "loss": 0.03644323,
+      "step": 6163
+    },
+    {
+      "epoch": 12.328,
+      "grad_norm": 2.5401759147644043,
+      "learning_rate": 2e-05,
+      "loss": 0.04062868,
+      "step": 6164
+    },
+    {
+      "epoch": 12.33,
+      "grad_norm": 1.3734387159347534,
+      "learning_rate": 2e-05,
+      "loss": 0.04200601,
+      "step": 6165
+    },
+    {
+      "epoch": 12.332,
+      "grad_norm": 1.1165435314178467,
+      "learning_rate": 2e-05,
+      "loss": 0.03473271,
+      "step": 6166
+    },
+    {
+      "epoch": 12.334,
+      "grad_norm": 1.3993453979492188,
+      "learning_rate": 2e-05,
+      "loss": 0.0432322,
+      "step": 6167
+    },
+    {
+      "epoch": 12.336,
+      "grad_norm": 1.486961841583252,
+      "learning_rate": 2e-05,
+      "loss": 0.04133502,
+      "step": 6168
+    },
+    {
+      "epoch": 12.338,
+      "grad_norm": 1.8763487339019775,
+      "learning_rate": 2e-05,
+      "loss": 0.04380414,
+      "step": 6169
+    },
+    {
+      "epoch": 12.34,
+      "grad_norm": 3.696859121322632,
+      "learning_rate": 2e-05,
+      "loss": 0.05829802,
+      "step": 6170
+    },
+    {
+      "epoch": 12.342,
+      "grad_norm": 1.087334156036377,
+      "learning_rate": 2e-05,
+      "loss": 0.03167205,
+      "step": 6171
+    },
+    {
+      "epoch": 12.344,
+      "grad_norm": 1.7672123908996582,
+      "learning_rate": 2e-05,
+      "loss": 0.0549523,
+      "step": 6172
+    },
+    {
+      "epoch": 12.346,
+      "grad_norm": 1.4898287057876587,
+      "learning_rate": 2e-05,
+      "loss": 0.05321936,
+      "step": 6173
+    },
+    {
+      "epoch": 12.348,
+      "grad_norm": 1.1989002227783203,
+      "learning_rate": 2e-05,
+      "loss": 0.03441871,
+      "step": 6174
+    },
+    {
+      "epoch": 12.35,
+      "grad_norm": 1.5749406814575195,
+      "learning_rate": 2e-05,
+      "loss": 0.04303573,
+      "step": 6175
+    },
+    {
+      "epoch": 12.352,
+      "grad_norm": 1.9614742994308472,
+      "learning_rate": 2e-05,
+      "loss": 0.06837121,
+      "step": 6176
+    },
+    {
+      "epoch": 12.354,
+      "grad_norm": 2.225933313369751,
+      "learning_rate": 2e-05,
+      "loss": 0.04865203,
+      "step": 6177
+    },
+    {
+      "epoch": 12.356,
+      "grad_norm": 2.307185173034668,
+      "learning_rate": 2e-05,
+      "loss": 0.05228138,
+      "step": 6178
+    },
+    {
+      "epoch": 12.358,
+      "grad_norm": 1.664549708366394,
+      "learning_rate": 2e-05,
+      "loss": 0.04851029,
+      "step": 6179
+    },
+    {
+      "epoch": 12.36,
+      "grad_norm": 1.9192323684692383,
+      "learning_rate": 2e-05,
+      "loss": 0.03343064,
+      "step": 6180
+    },
+    {
+      "epoch": 12.362,
+      "grad_norm": 1.0249700546264648,
+      "learning_rate": 2e-05,
+      "loss": 0.02511407,
+      "step": 6181
+    },
+    {
+      "epoch": 12.364,
+      "grad_norm": 1.6150821447372437,
+      "learning_rate": 2e-05,
+      "loss": 0.0319114,
+      "step": 6182
+    },
+    {
+      "epoch": 12.366,
+      "grad_norm": 1.0298298597335815,
+      "learning_rate": 2e-05,
+      "loss": 0.02553884,
+      "step": 6183
+    },
+    {
+      "epoch": 12.368,
+      "grad_norm": 2.103635787963867,
+      "learning_rate": 2e-05,
+      "loss": 0.05173234,
+      "step": 6184
+    },
+    {
+      "epoch": 12.37,
+      "grad_norm": 3.9606614112854004,
+      "learning_rate": 2e-05,
+      "loss": 0.04179762,
+      "step": 6185
+    },
+    {
+      "epoch": 12.372,
+      "grad_norm": 1.0706768035888672,
+      "learning_rate": 2e-05,
+      "loss": 0.03646227,
+      "step": 6186
+    },
+    {
+      "epoch": 12.374,
+      "grad_norm": 1.6147788763046265,
+      "learning_rate": 2e-05,
+      "loss": 0.03327171,
+      "step": 6187
+    },
+    {
+      "epoch": 12.376,
+      "grad_norm": 2.25844407081604,
+      "learning_rate": 2e-05,
+      "loss": 0.05061604,
+      "step": 6188
+    },
+    {
+      "epoch": 12.378,
+      "grad_norm": 1.574654459953308,
+      "learning_rate": 2e-05,
+      "loss": 0.04572042,
+      "step": 6189
+    },
+    {
+      "epoch": 12.38,
+      "grad_norm": 1.6320792436599731,
+      "learning_rate": 2e-05,
+      "loss": 0.04378324,
+      "step": 6190
+    },
+    {
+      "epoch": 12.382,
+      "grad_norm": 2.3261327743530273,
+      "learning_rate": 2e-05,
+      "loss": 0.04833006,
+      "step": 6191
+    },
+    {
+      "epoch": 12.384,
+      "grad_norm": 1.5708943605422974,
+      "learning_rate": 2e-05,
+      "loss": 0.03576447,
+      "step": 6192
+    },
+    {
+      "epoch": 12.386,
+      "grad_norm": 1.543707251548767,
+      "learning_rate": 2e-05,
+      "loss": 0.03951041,
+      "step": 6193
+    },
+    {
+      "epoch": 12.388,
+      "grad_norm": 1.1679742336273193,
+      "learning_rate": 2e-05,
+      "loss": 0.03864168,
+      "step": 6194
+    },
+    {
+      "epoch": 12.39,
+      "grad_norm": 2.291689157485962,
+      "learning_rate": 2e-05,
+      "loss": 0.04182493,
+      "step": 6195
+    },
+    {
+      "epoch": 12.392,
+      "grad_norm": 1.7392657995224,
+      "learning_rate": 2e-05,
+      "loss": 0.04270639,
+      "step": 6196
+    },
+    {
+      "epoch": 12.394,
+      "grad_norm": 1.5574685335159302,
+      "learning_rate": 2e-05,
+      "loss": 0.03201068,
+      "step": 6197
+    },
+    {
+      "epoch": 12.396,
+      "grad_norm": 1.4459373950958252,
+      "learning_rate": 2e-05,
+      "loss": 0.05233443,
+      "step": 6198
+    },
+    {
+      "epoch": 12.398,
+      "grad_norm": 1.459169626235962,
+      "learning_rate": 2e-05,
+      "loss": 0.0575183,
+      "step": 6199
+    },
+    {
+      "epoch": 12.4,
+      "grad_norm": 1.6888347864151,
+      "learning_rate": 2e-05,
+      "loss": 0.03846366,
+      "step": 6200
+    },
+    {
+      "epoch": 12.402,
+      "grad_norm": 2.6593565940856934,
+      "learning_rate": 2e-05,
+      "loss": 0.03823034,
+      "step": 6201
+    },
+    {
+      "epoch": 12.404,
+      "grad_norm": 1.6569043397903442,
+      "learning_rate": 2e-05,
+      "loss": 0.04339511,
+      "step": 6202
+    },
+    {
+      "epoch": 12.406,
+      "grad_norm": 0.9647795557975769,
+      "learning_rate": 2e-05,
+      "loss": 0.03080498,
+      "step": 6203
+    },
+    {
+      "epoch": 12.408,
+      "grad_norm": 2.6168878078460693,
+      "learning_rate": 2e-05,
+      "loss": 0.04123443,
+      "step": 6204
+    },
+    {
+      "epoch": 12.41,
+      "grad_norm": 1.4847103357315063,
+      "learning_rate": 2e-05,
+      "loss": 0.04044367,
+      "step": 6205
+    },
+    {
+      "epoch": 12.412,
+      "grad_norm": 2.002779483795166,
+      "learning_rate": 2e-05,
+      "loss": 0.06478892,
+      "step": 6206
+    },
+    {
+      "epoch": 12.414,
+      "grad_norm": 1.692628026008606,
+      "learning_rate": 2e-05,
+      "loss": 0.04533452,
+      "step": 6207
+    },
+    {
+      "epoch": 12.416,
+      "grad_norm": 2.04227876663208,
+      "learning_rate": 2e-05,
+      "loss": 0.05172842,
+      "step": 6208
+    },
+    {
+      "epoch": 12.418,
+      "grad_norm": 1.5471951961517334,
+      "learning_rate": 2e-05,
+      "loss": 0.04024774,
+      "step": 6209
+    },
+    {
+      "epoch": 12.42,
+      "grad_norm": 3.0287656784057617,
+      "learning_rate": 2e-05,
+      "loss": 0.07643662,
+      "step": 6210
+    },
+    {
+      "epoch": 12.422,
+      "grad_norm": 1.17807137966156,
+      "learning_rate": 2e-05,
+      "loss": 0.03349002,
+      "step": 6211
+    },
+    {
+      "epoch": 12.424,
+      "grad_norm": 1.3760405778884888,
+      "learning_rate": 2e-05,
+      "loss": 0.04548774,
+      "step": 6212
+    },
+    {
+      "epoch": 12.426,
+      "grad_norm": 1.2264763116836548,
+      "learning_rate": 2e-05,
+      "loss": 0.03662616,
+      "step": 6213
+    },
+    {
+      "epoch": 12.428,
+      "grad_norm": 1.1853759288787842,
+      "learning_rate": 2e-05,
+      "loss": 0.03420371,
+      "step": 6214
+    },
+    {
+      "epoch": 12.43,
+      "grad_norm": 1.136852502822876,
+      "learning_rate": 2e-05,
+      "loss": 0.04038435,
+      "step": 6215
+    },
+    {
+      "epoch": 12.432,
+      "grad_norm": 1.1933867931365967,
+      "learning_rate": 2e-05,
+      "loss": 0.04216532,
+      "step": 6216
+    },
+    {
+      "epoch": 12.434,
+      "grad_norm": 4.264523506164551,
+      "learning_rate": 2e-05,
+      "loss": 0.0426877,
+      "step": 6217
+    },
+    {
+      "epoch": 12.436,
+      "grad_norm": 1.2900351285934448,
+      "learning_rate": 2e-05,
+      "loss": 0.03366522,
+      "step": 6218
+    },
+    {
+      "epoch": 12.438,
+      "grad_norm": 2.727051258087158,
+      "learning_rate": 2e-05,
+      "loss": 0.05351751,
+      "step": 6219
+    },
+    {
+      "epoch": 12.44,
+      "grad_norm": 1.5386077165603638,
+      "learning_rate": 2e-05,
+      "loss": 0.05369819,
+      "step": 6220
+    },
+    {
+      "epoch": 12.442,
+      "grad_norm": 1.445359468460083,
+      "learning_rate": 2e-05,
+      "loss": 0.04263123,
+      "step": 6221
+    },
+    {
+      "epoch": 12.444,
+      "grad_norm": 1.123322606086731,
+      "learning_rate": 2e-05,
+      "loss": 0.04013848,
+      "step": 6222
+    },
+    {
+      "epoch": 12.446,
+      "grad_norm": 0.9369562268257141,
+      "learning_rate": 2e-05,
+      "loss": 0.02055129,
+      "step": 6223
+    },
+    {
+      "epoch": 12.448,
+      "grad_norm": 1.31740140914917,
+      "learning_rate": 2e-05,
+      "loss": 0.04068302,
+      "step": 6224
+    },
+    {
+      "epoch": 12.45,
+      "grad_norm": 1.307610034942627,
+      "learning_rate": 2e-05,
+      "loss": 0.05438361,
+      "step": 6225
+    },
+    {
+      "epoch": 12.452,
+      "grad_norm": 1.2957713603973389,
+      "learning_rate": 2e-05,
+      "loss": 0.02663429,
+      "step": 6226
+    },
+    {
+      "epoch": 12.454,
+      "grad_norm": 1.6599071025848389,
+      "learning_rate": 2e-05,
+      "loss": 0.04203624,
+      "step": 6227
+    },
+    {
+      "epoch": 12.456,
+      "grad_norm": 1.427505373954773,
+      "learning_rate": 2e-05,
+      "loss": 0.04064513,
+      "step": 6228
+    },
+    {
+      "epoch": 12.458,
+      "grad_norm": 1.9157131910324097,
+      "learning_rate": 2e-05,
+      "loss": 0.05884474,
+      "step": 6229
+    },
+    {
+      "epoch": 12.46,
+      "grad_norm": 1.3565418720245361,
+      "learning_rate": 2e-05,
+      "loss": 0.04177081,
+      "step": 6230
+    },
+    {
+      "epoch": 12.462,
+      "grad_norm": 1.1073006391525269,
+      "learning_rate": 2e-05,
+      "loss": 0.03636095,
+      "step": 6231
+    },
+    {
+      "epoch": 12.464,
+      "grad_norm": 1.0298948287963867,
+      "learning_rate": 2e-05,
+      "loss": 0.03898697,
+      "step": 6232
+    },
+    {
+      "epoch": 12.466,
+      "grad_norm": 1.1140865087509155,
+      "learning_rate": 2e-05,
+      "loss": 0.03568536,
+      "step": 6233
+    },
+    {
+      "epoch": 12.468,
+      "grad_norm": 1.7985575199127197,
+      "learning_rate": 2e-05,
+      "loss": 0.04029547,
+      "step": 6234
+    },
+    {
+      "epoch": 12.47,
+      "grad_norm": 1.2943682670593262,
+      "learning_rate": 2e-05,
+      "loss": 0.03818323,
+      "step": 6235
+    },
+    {
+      "epoch": 12.472,
+      "grad_norm": 1.2726134061813354,
+      "learning_rate": 2e-05,
+      "loss": 0.04815301,
+      "step": 6236
+    },
+    {
+      "epoch": 12.474,
+      "grad_norm": 1.8308722972869873,
+      "learning_rate": 2e-05,
+      "loss": 0.05556687,
+      "step": 6237
+    },
+    {
+      "epoch": 12.475999999999999,
+      "grad_norm": 1.4734604358673096,
+      "learning_rate": 2e-05,
+      "loss": 0.05400109,
+      "step": 6238
+    },
+    {
+      "epoch": 12.478,
+      "grad_norm": 1.909253478050232,
+      "learning_rate": 2e-05,
+      "loss": 0.04208795,
+      "step": 6239
+    },
+    {
+      "epoch": 12.48,
+      "grad_norm": 1.3197088241577148,
+      "learning_rate": 2e-05,
+      "loss": 0.03566179,
+      "step": 6240
+    },
+    {
+      "epoch": 12.482,
+      "grad_norm": 1.8218048810958862,
+      "learning_rate": 2e-05,
+      "loss": 0.05417576,
+      "step": 6241
+    },
+    {
+      "epoch": 12.484,
+      "grad_norm": 1.8446364402770996,
+      "learning_rate": 2e-05,
+      "loss": 0.04411314,
+      "step": 6242
+    },
+    {
+      "epoch": 12.486,
+      "grad_norm": 1.5696487426757812,
+      "learning_rate": 2e-05,
+      "loss": 0.04941075,
+      "step": 6243
+    },
+    {
+      "epoch": 12.488,
+      "grad_norm": 3.357449531555176,
+      "learning_rate": 2e-05,
+      "loss": 0.05497446,
+      "step": 6244
+    },
+    {
+      "epoch": 12.49,
+      "grad_norm": 3.2609894275665283,
+      "learning_rate": 2e-05,
+      "loss": 0.06280646,
+      "step": 6245
+    },
+    {
+      "epoch": 12.492,
+      "grad_norm": 1.8737690448760986,
+      "learning_rate": 2e-05,
+      "loss": 0.03448436,
+      "step": 6246
+    },
+    {
+      "epoch": 12.494,
+      "grad_norm": 1.4557076692581177,
+      "learning_rate": 2e-05,
+      "loss": 0.05713212,
+      "step": 6247
+    },
+    {
+      "epoch": 12.496,
+      "grad_norm": 1.4441951513290405,
+      "learning_rate": 2e-05,
+      "loss": 0.04061989,
+      "step": 6248
+    },
+    {
+      "epoch": 12.498,
+      "grad_norm": 1.1646778583526611,
+      "learning_rate": 2e-05,
+      "loss": 0.02681767,
+      "step": 6249
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 1.1207839250564575,
+      "learning_rate": 2e-05,
+      "loss": 0.03733048,
+      "step": 6250
+    },
+    {
+      "epoch": 12.502,
+      "grad_norm": 2.0258734226226807,
+      "learning_rate": 2e-05,
+      "loss": 0.05226999,
+      "step": 6251
+    },
+    {
+      "epoch": 12.504,
+      "grad_norm": 2.436753034591675,
+      "learning_rate": 2e-05,
+      "loss": 0.06572986,
+      "step": 6252
+    },
+    {
+      "epoch": 12.506,
+      "grad_norm": 1.5565096139907837,
+      "learning_rate": 2e-05,
+      "loss": 0.04375243,
+      "step": 6253
+    },
+    {
+      "epoch": 12.508,
+      "grad_norm": 2.353769302368164,
+      "learning_rate": 2e-05,
+      "loss": 0.06041701,
+      "step": 6254
+    },
+    {
+      "epoch": 12.51,
+      "grad_norm": 1.1774601936340332,
+      "learning_rate": 2e-05,
+      "loss": 0.04639146,
+      "step": 6255
+    },
+    {
+      "epoch": 12.512,
+      "grad_norm": 2.4573864936828613,
+      "learning_rate": 2e-05,
+      "loss": 0.04883677,
+      "step": 6256
+    },
+    {
+      "epoch": 12.514,
+      "grad_norm": 2.6108806133270264,
+      "learning_rate": 2e-05,
+      "loss": 0.06272373,
+      "step": 6257
+    },
+    {
+      "epoch": 12.516,
+      "grad_norm": 1.03660249710083,
+      "learning_rate": 2e-05,
+      "loss": 0.02872313,
+      "step": 6258
+    },
+    {
+      "epoch": 12.518,
+      "grad_norm": 2.0671160221099854,
+      "learning_rate": 2e-05,
+      "loss": 0.03815708,
+      "step": 6259
+    },
+    {
+      "epoch": 12.52,
+      "grad_norm": 2.072272777557373,
+      "learning_rate": 2e-05,
+      "loss": 0.03585609,
+      "step": 6260
+    },
+    {
+      "epoch": 12.522,
+      "grad_norm": 1.6502577066421509,
+      "learning_rate": 2e-05,
+      "loss": 0.05801409,
+      "step": 6261
+    },
+    {
+      "epoch": 12.524000000000001,
+      "grad_norm": 4.274342060089111,
+      "learning_rate": 2e-05,
+      "loss": 0.04877973,
+      "step": 6262
+    },
+    {
+      "epoch": 12.526,
+      "grad_norm": 1.2209831476211548,
+      "learning_rate": 2e-05,
+      "loss": 0.03942978,
+      "step": 6263
+    },
+    {
+      "epoch": 12.528,
+      "grad_norm": 1.6523079872131348,
+      "learning_rate": 2e-05,
+      "loss": 0.0488749,
+      "step": 6264
+    },
+    {
+      "epoch": 12.53,
+      "grad_norm": 2.2305943965911865,
+      "learning_rate": 2e-05,
+      "loss": 0.05138541,
+      "step": 6265
+    },
+    {
+      "epoch": 12.532,
+      "grad_norm": 1.3322969675064087,
+      "learning_rate": 2e-05,
+      "loss": 0.04032963,
+      "step": 6266
+    },
+    {
+      "epoch": 12.534,
+      "grad_norm": 2.0621213912963867,
+      "learning_rate": 2e-05,
+      "loss": 0.05739632,
+      "step": 6267
+    },
+    {
+      "epoch": 12.536,
+      "grad_norm": 0.9136289358139038,
+      "learning_rate": 2e-05,
+      "loss": 0.03144271,
+      "step": 6268
+    },
+    {
+      "epoch": 12.538,
+      "grad_norm": 1.2017995119094849,
+      "learning_rate": 2e-05,
+      "loss": 0.0261726,
+      "step": 6269
+    },
+    {
+      "epoch": 12.54,
+      "grad_norm": 1.861670732498169,
+      "learning_rate": 2e-05,
+      "loss": 0.05213568,
+      "step": 6270
+    },
+    {
+      "epoch": 12.542,
+      "grad_norm": 3.512301206588745,
+      "learning_rate": 2e-05,
+      "loss": 0.10029002,
+      "step": 6271
+    },
+    {
+      "epoch": 12.544,
+      "grad_norm": 1.3964629173278809,
+      "learning_rate": 2e-05,
+      "loss": 0.0487773,
+      "step": 6272
+    },
+    {
+      "epoch": 12.546,
+      "grad_norm": 1.554492473602295,
+      "learning_rate": 2e-05,
+      "loss": 0.04269033,
+      "step": 6273
+    },
+    {
+      "epoch": 12.548,
+      "grad_norm": 1.0274790525436401,
+      "learning_rate": 2e-05,
+      "loss": 0.02957164,
+      "step": 6274
+    },
+    {
+      "epoch": 12.55,
+      "grad_norm": 1.4311920404434204,
+      "learning_rate": 2e-05,
+      "loss": 0.03769149,
+      "step": 6275
+    },
+    {
+      "epoch": 12.552,
+      "grad_norm": 1.9928914308547974,
+      "learning_rate": 2e-05,
+      "loss": 0.04190875,
+      "step": 6276
+    },
+    {
+      "epoch": 12.554,
+      "grad_norm": 1.5183537006378174,
+      "learning_rate": 2e-05,
+      "loss": 0.04576535,
+      "step": 6277
+    },
+    {
+      "epoch": 12.556000000000001,
+      "grad_norm": 2.0962085723876953,
+      "learning_rate": 2e-05,
+      "loss": 0.04964176,
+      "step": 6278
+    },
+    {
+      "epoch": 12.558,
+      "grad_norm": 2.3136391639709473,
+      "learning_rate": 2e-05,
+      "loss": 0.03684264,
+      "step": 6279
+    },
+    {
+      "epoch": 12.56,
+      "grad_norm": 2.110243558883667,
+      "learning_rate": 2e-05,
+      "loss": 0.05677021,
+      "step": 6280
+    },
+    {
+      "epoch": 12.562,
+      "grad_norm": 3.119753360748291,
+      "learning_rate": 2e-05,
+      "loss": 0.04740483,
+      "step": 6281
+    },
+    {
+      "epoch": 12.564,
+      "grad_norm": 1.0421844720840454,
+      "learning_rate": 2e-05,
+      "loss": 0.03042347,
+      "step": 6282
+    },
+    {
+      "epoch": 12.566,
+      "grad_norm": 1.7684757709503174,
+      "learning_rate": 2e-05,
+      "loss": 0.04349629,
+      "step": 6283
+    },
+    {
+      "epoch": 12.568,
+      "grad_norm": 1.600565791130066,
+      "learning_rate": 2e-05,
+      "loss": 0.04671545,
+      "step": 6284
+    },
+    {
+      "epoch": 12.57,
+      "grad_norm": 1.4540395736694336,
+      "learning_rate": 2e-05,
+      "loss": 0.04411473,
+      "step": 6285
+    },
+    {
+      "epoch": 12.572,
+      "grad_norm": 1.092216968536377,
+      "learning_rate": 2e-05,
+      "loss": 0.03701637,
+      "step": 6286
+    },
+    {
+      "epoch": 12.574,
+      "grad_norm": 1.4555920362472534,
+      "learning_rate": 2e-05,
+      "loss": 0.05183559,
+      "step": 6287
+    },
+    {
+      "epoch": 12.576,
+      "grad_norm": 2.1019439697265625,
+      "learning_rate": 2e-05,
+      "loss": 0.06192996,
+      "step": 6288
+    },
+    {
+      "epoch": 12.578,
+      "grad_norm": 1.7618634700775146,
+      "learning_rate": 2e-05,
+      "loss": 0.03472544,
+      "step": 6289
+    },
+    {
+      "epoch": 12.58,
+      "grad_norm": 1.7783581018447876,
+      "learning_rate": 2e-05,
+      "loss": 0.04450977,
+      "step": 6290
+    },
+    {
+      "epoch": 12.582,
+      "grad_norm": 1.32155179977417,
+      "learning_rate": 2e-05,
+      "loss": 0.04913367,
+      "step": 6291
+    },
+    {
+      "epoch": 12.584,
+      "grad_norm": 1.7167394161224365,
+      "learning_rate": 2e-05,
+      "loss": 0.03525726,
+      "step": 6292
+    },
+    {
+      "epoch": 12.586,
+      "grad_norm": 1.5468295812606812,
+      "learning_rate": 2e-05,
+      "loss": 0.04274698,
+      "step": 6293
+    },
+    {
+      "epoch": 12.588,
+      "grad_norm": 1.0729691982269287,
+      "learning_rate": 2e-05,
+      "loss": 0.04672535,
+      "step": 6294
+    },
+    {
+      "epoch": 12.59,
+      "grad_norm": 1.9018189907073975,
+      "learning_rate": 2e-05,
+      "loss": 0.04148772,
+      "step": 6295
+    },
+    {
+      "epoch": 12.592,
+      "grad_norm": 1.4016549587249756,
+      "learning_rate": 2e-05,
+      "loss": 0.04215005,
+      "step": 6296
+    },
+    {
+      "epoch": 12.594,
+      "grad_norm": 1.5528912544250488,
+      "learning_rate": 2e-05,
+      "loss": 0.03888253,
+      "step": 6297
+    },
+    {
+      "epoch": 12.596,
+      "grad_norm": 1.093119740486145,
+      "learning_rate": 2e-05,
+      "loss": 0.03328623,
+      "step": 6298
+    },
+    {
+      "epoch": 12.598,
+      "grad_norm": 1.691413402557373,
+      "learning_rate": 2e-05,
+      "loss": 0.04086442,
+      "step": 6299
+    },
+    {
+      "epoch": 12.6,
+      "grad_norm": 1.854809284210205,
+      "learning_rate": 2e-05,
+      "loss": 0.03869913,
+      "step": 6300
+    },
+    {
+      "epoch": 12.602,
+      "grad_norm": 3.5889766216278076,
+      "learning_rate": 2e-05,
+      "loss": 0.03339052,
+      "step": 6301
+    },
+    {
+      "epoch": 12.604,
+      "grad_norm": 1.0106459856033325,
+      "learning_rate": 2e-05,
+      "loss": 0.03362305,
+      "step": 6302
+    },
+    {
+      "epoch": 12.606,
+      "grad_norm": 1.3339776992797852,
+      "learning_rate": 2e-05,
+      "loss": 0.03432526,
+      "step": 6303
+    },
+    {
+      "epoch": 12.608,
+      "grad_norm": 1.3809486627578735,
+      "learning_rate": 2e-05,
+      "loss": 0.0381794,
+      "step": 6304
+    },
+    {
+      "epoch": 12.61,
+      "grad_norm": 4.405636787414551,
+      "learning_rate": 2e-05,
+      "loss": 0.05852031,
+      "step": 6305
+    },
+    {
+      "epoch": 12.612,
+      "grad_norm": 2.022235870361328,
+      "learning_rate": 2e-05,
+      "loss": 0.04848814,
+      "step": 6306
+    },
+    {
+      "epoch": 12.614,
+      "grad_norm": 2.3099355697631836,
+      "learning_rate": 2e-05,
+      "loss": 0.05684424,
+      "step": 6307
+    },
+    {
+      "epoch": 12.616,
+      "grad_norm": 1.1495301723480225,
+      "learning_rate": 2e-05,
+      "loss": 0.02855151,
+      "step": 6308
+    },
+    {
+      "epoch": 12.618,
+      "grad_norm": 1.6715397834777832,
+      "learning_rate": 2e-05,
+      "loss": 0.03850598,
+      "step": 6309
+    },
+    {
+      "epoch": 12.62,
+      "grad_norm": 1.773354172706604,
+      "learning_rate": 2e-05,
+      "loss": 0.04509133,
+      "step": 6310
+    },
+    {
+      "epoch": 12.622,
+      "grad_norm": 1.419111967086792,
+      "learning_rate": 2e-05,
+      "loss": 0.03800103,
+      "step": 6311
+    },
+    {
+      "epoch": 12.624,
+      "grad_norm": 1.9277119636535645,
+      "learning_rate": 2e-05,
+      "loss": 0.05758407,
+      "step": 6312
+    },
+    {
+      "epoch": 12.626,
+      "grad_norm": 2.3455944061279297,
+      "learning_rate": 2e-05,
+      "loss": 0.05955437,
+      "step": 6313
+    },
+    {
+      "epoch": 12.628,
+      "grad_norm": 1.5641530752182007,
+      "learning_rate": 2e-05,
+      "loss": 0.05085967,
+      "step": 6314
+    },
+    {
+      "epoch": 12.63,
+      "grad_norm": 4.214975833892822,
+      "learning_rate": 2e-05,
+      "loss": 0.04824074,
+      "step": 6315
+    },
+    {
+      "epoch": 12.632,
+      "grad_norm": 2.380524158477783,
+      "learning_rate": 2e-05,
+      "loss": 0.0585499,
+      "step": 6316
+    },
+    {
+      "epoch": 12.634,
+      "grad_norm": 1.3643995523452759,
+      "learning_rate": 2e-05,
+      "loss": 0.04846994,
+      "step": 6317
+    },
+    {
+      "epoch": 12.636,
+      "grad_norm": 1.191190242767334,
+      "learning_rate": 2e-05,
+      "loss": 0.03240199,
+      "step": 6318
+    },
+    {
+      "epoch": 12.638,
+      "grad_norm": 2.9533886909484863,
+      "learning_rate": 2e-05,
+      "loss": 0.0543364,
+      "step": 6319
+    },
+    {
+      "epoch": 12.64,
+      "grad_norm": 1.2633097171783447,
+      "learning_rate": 2e-05,
+      "loss": 0.03655366,
+      "step": 6320
+    },
+    {
+      "epoch": 12.642,
+      "grad_norm": 1.2036083936691284,
+      "learning_rate": 2e-05,
+      "loss": 0.03070158,
+      "step": 6321
+    },
+    {
+      "epoch": 12.644,
+      "grad_norm": 2.5942792892456055,
+      "learning_rate": 2e-05,
+      "loss": 0.04942968,
+      "step": 6322
+    },
+    {
+      "epoch": 12.646,
+      "grad_norm": 1.0475519895553589,
+      "learning_rate": 2e-05,
+      "loss": 0.03110964,
+      "step": 6323
+    },
+    {
+      "epoch": 12.648,
+      "grad_norm": 1.921798825263977,
+      "learning_rate": 2e-05,
+      "loss": 0.06349019,
+      "step": 6324
+    },
+    {
+      "epoch": 12.65,
+      "grad_norm": 1.2762218713760376,
+      "learning_rate": 2e-05,
+      "loss": 0.04748666,
+      "step": 6325
+    },
+    {
+      "epoch": 12.652,
+      "grad_norm": 1.2782567739486694,
+      "learning_rate": 2e-05,
+      "loss": 0.04201008,
+      "step": 6326
+    },
+    {
+      "epoch": 12.654,
+      "grad_norm": 0.7562774419784546,
+      "learning_rate": 2e-05,
+      "loss": 0.02028156,
+      "step": 6327
+    },
+    {
+      "epoch": 12.656,
+      "grad_norm": 1.9511414766311646,
+      "learning_rate": 2e-05,
+      "loss": 0.04364496,
+      "step": 6328
+    },
+    {
+      "epoch": 12.658,
+      "grad_norm": 1.7642337083816528,
+      "learning_rate": 2e-05,
+      "loss": 0.05323292,
+      "step": 6329
+    },
+    {
+      "epoch": 12.66,
+      "grad_norm": 2.2341959476470947,
+      "learning_rate": 2e-05,
+      "loss": 0.03851413,
+      "step": 6330
+    },
+    {
+      "epoch": 12.662,
+      "grad_norm": 1.5995410680770874,
+      "learning_rate": 2e-05,
+      "loss": 0.0516777,
+      "step": 6331
+    },
+    {
+      "epoch": 12.664,
+      "grad_norm": 1.8847028017044067,
+      "learning_rate": 2e-05,
+      "loss": 0.04388159,
+      "step": 6332
+    },
+    {
+      "epoch": 12.666,
+      "grad_norm": 2.132364511489868,
+      "learning_rate": 2e-05,
+      "loss": 0.04297321,
+      "step": 6333
+    },
+    {
+      "epoch": 12.668,
+      "grad_norm": 2.32963228225708,
+      "learning_rate": 2e-05,
+      "loss": 0.03326223,
+      "step": 6334
+    },
+    {
+      "epoch": 12.67,
+      "grad_norm": 1.9717942476272583,
+      "learning_rate": 2e-05,
+      "loss": 0.04060198,
+      "step": 6335
+    },
+    {
+      "epoch": 12.672,
+      "grad_norm": 1.5272952318191528,
+      "learning_rate": 2e-05,
+      "loss": 0.05175653,
+      "step": 6336
+    },
+    {
+      "epoch": 12.674,
+      "grad_norm": 1.090294361114502,
+      "learning_rate": 2e-05,
+      "loss": 0.05121554,
+      "step": 6337
+    },
+    {
+      "epoch": 12.676,
+      "grad_norm": 0.9349586367607117,
+      "learning_rate": 2e-05,
+      "loss": 0.03004047,
+      "step": 6338
+    },
+    {
+      "epoch": 12.678,
+      "grad_norm": 3.6723544597625732,
+      "learning_rate": 2e-05,
+      "loss": 0.05083375,
+      "step": 6339
+    },
+    {
+      "epoch": 12.68,
+      "grad_norm": 1.8542966842651367,
+      "learning_rate": 2e-05,
+      "loss": 0.05500766,
+      "step": 6340
+    },
+    {
+      "epoch": 12.682,
+      "grad_norm": 1.6407077312469482,
+      "learning_rate": 2e-05,
+      "loss": 0.0521637,
+      "step": 6341
+    },
+    {
+      "epoch": 12.684,
+      "grad_norm": 1.1183030605316162,
+      "learning_rate": 2e-05,
+      "loss": 0.03896394,
+      "step": 6342
+    },
+    {
+      "epoch": 12.686,
+      "grad_norm": 1.987824559211731,
+      "learning_rate": 2e-05,
+      "loss": 0.05758013,
+      "step": 6343
+    },
+    {
+      "epoch": 12.688,
+      "grad_norm": 1.398348331451416,
+      "learning_rate": 2e-05,
+      "loss": 0.04054468,
+      "step": 6344
+    },
+    {
+      "epoch": 12.69,
+      "grad_norm": 1.3957326412200928,
+      "learning_rate": 2e-05,
+      "loss": 0.03729579,
+      "step": 6345
+    },
+    {
+      "epoch": 12.692,
+      "grad_norm": 1.2762089967727661,
+      "learning_rate": 2e-05,
+      "loss": 0.03696339,
+      "step": 6346
+    },
+    {
+      "epoch": 12.693999999999999,
+      "grad_norm": 2.0938408374786377,
+      "learning_rate": 2e-05,
+      "loss": 0.03400442,
+      "step": 6347
+    },
+    {
+      "epoch": 12.696,
+      "grad_norm": 1.4813979864120483,
+      "learning_rate": 2e-05,
+      "loss": 0.03928123,
+      "step": 6348
+    },
+    {
+      "epoch": 12.698,
+      "grad_norm": 1.0480759143829346,
+      "learning_rate": 2e-05,
+      "loss": 0.03268198,
+      "step": 6349
+    },
+    {
+      "epoch": 12.7,
+      "grad_norm": 1.0556561946868896,
+      "learning_rate": 2e-05,
+      "loss": 0.03014071,
+      "step": 6350
+    },
+    {
+      "epoch": 12.702,
+      "grad_norm": 1.721622347831726,
+      "learning_rate": 2e-05,
+      "loss": 0.04010597,
+      "step": 6351
+    },
+    {
+      "epoch": 12.704,
+      "grad_norm": 1.3392608165740967,
+      "learning_rate": 2e-05,
+      "loss": 0.04958211,
+      "step": 6352
+    },
+    {
+      "epoch": 12.706,
+      "grad_norm": 2.3841845989227295,
+      "learning_rate": 2e-05,
+      "loss": 0.06484979,
+      "step": 6353
+    },
+    {
+      "epoch": 12.708,
+      "grad_norm": 1.1281391382217407,
+      "learning_rate": 2e-05,
+      "loss": 0.03068016,
+      "step": 6354
+    },
+    {
+      "epoch": 12.71,
+      "grad_norm": 1.1351035833358765,
+      "learning_rate": 2e-05,
+      "loss": 0.03658585,
+      "step": 6355
+    },
+    {
+      "epoch": 12.712,
+      "grad_norm": 2.5402114391326904,
+      "learning_rate": 2e-05,
+      "loss": 0.04548267,
+      "step": 6356
+    },
+    {
+      "epoch": 12.714,
+      "grad_norm": 2.3668107986450195,
+      "learning_rate": 2e-05,
+      "loss": 0.03807458,
+      "step": 6357
+    },
+    {
+      "epoch": 12.716,
+      "grad_norm": 1.4179434776306152,
+      "learning_rate": 2e-05,
+      "loss": 0.03555714,
+      "step": 6358
+    },
+    {
+      "epoch": 12.718,
+      "grad_norm": 1.4804911613464355,
+      "learning_rate": 2e-05,
+      "loss": 0.04498214,
+      "step": 6359
+    },
+    {
+      "epoch": 12.72,
+      "grad_norm": 2.23551344871521,
+      "learning_rate": 2e-05,
+      "loss": 0.05175463,
+      "step": 6360
+    },
+    {
+      "epoch": 12.722,
+      "grad_norm": 1.7548456192016602,
+      "learning_rate": 2e-05,
+      "loss": 0.06676908,
+      "step": 6361
+    },
+    {
+      "epoch": 12.724,
+      "grad_norm": 1.852115511894226,
+      "learning_rate": 2e-05,
+      "loss": 0.05606053,
+      "step": 6362
+    },
+    {
+      "epoch": 12.725999999999999,
+      "grad_norm": 1.3961422443389893,
+      "learning_rate": 2e-05,
+      "loss": 0.04076095,
+      "step": 6363
+    },
+    {
+      "epoch": 12.728,
+      "grad_norm": 2.0832810401916504,
+      "learning_rate": 2e-05,
+      "loss": 0.06329206,
+      "step": 6364
+    },
+    {
+      "epoch": 12.73,
+      "grad_norm": 2.148437976837158,
+      "learning_rate": 2e-05,
+      "loss": 0.06662284,
+      "step": 6365
+    },
+    {
+      "epoch": 12.732,
+      "grad_norm": 1.9320807456970215,
+      "learning_rate": 2e-05,
+      "loss": 0.04591805,
+      "step": 6366
+    },
+    {
+      "epoch": 12.734,
+      "grad_norm": 1.0681511163711548,
+      "learning_rate": 2e-05,
+      "loss": 0.03428921,
+      "step": 6367
+    },
+    {
+      "epoch": 12.736,
+      "grad_norm": 1.1943943500518799,
+      "learning_rate": 2e-05,
+      "loss": 0.0293221,
+      "step": 6368
+    },
+    {
+      "epoch": 12.738,
+      "grad_norm": 1.1118965148925781,
+      "learning_rate": 2e-05,
+      "loss": 0.03196052,
+      "step": 6369
+    },
+    {
+      "epoch": 12.74,
+      "grad_norm": 1.5805182456970215,
+      "learning_rate": 2e-05,
+      "loss": 0.04865031,
+      "step": 6370
+    },
+    {
+      "epoch": 12.742,
+      "grad_norm": 3.262758255004883,
+      "learning_rate": 2e-05,
+      "loss": 0.05576367,
+      "step": 6371
+    },
+    {
+      "epoch": 12.744,
+      "grad_norm": 1.1225249767303467,
+      "learning_rate": 2e-05,
+      "loss": 0.03483301,
+      "step": 6372
+    },
+    {
+      "epoch": 12.746,
+      "grad_norm": 1.2867493629455566,
+      "learning_rate": 2e-05,
+      "loss": 0.04086279,
+      "step": 6373
+    },
+    {
+      "epoch": 12.748,
+      "grad_norm": 1.298018217086792,
+      "learning_rate": 2e-05,
+      "loss": 0.04258086,
+      "step": 6374
+    },
+    {
+      "epoch": 12.75,
+      "grad_norm": 1.216958999633789,
+      "learning_rate": 2e-05,
+      "loss": 0.04485769,
+      "step": 6375
+    },
+    {
+      "epoch": 12.752,
+      "grad_norm": 1.4479326009750366,
+      "learning_rate": 2e-05,
+      "loss": 0.04460638,
+      "step": 6376
+    },
+    {
+      "epoch": 12.754,
+      "grad_norm": 1.5058059692382812,
+      "learning_rate": 2e-05,
+      "loss": 0.0504184,
+      "step": 6377
+    },
+    {
+      "epoch": 12.756,
+      "grad_norm": 1.393099308013916,
+      "learning_rate": 2e-05,
+      "loss": 0.03610604,
+      "step": 6378
+    },
+    {
+      "epoch": 12.758,
+      "grad_norm": 1.2349895238876343,
+      "learning_rate": 2e-05,
+      "loss": 0.04152373,
+      "step": 6379
+    },
+    {
+      "epoch": 12.76,
+      "grad_norm": 2.2473106384277344,
+      "learning_rate": 2e-05,
+      "loss": 0.06257766,
+      "step": 6380
+    },
+    {
+      "epoch": 12.762,
+      "grad_norm": 4.311264514923096,
+      "learning_rate": 2e-05,
+      "loss": 0.04196393,
+      "step": 6381
+    },
+    {
+      "epoch": 12.764,
+      "grad_norm": 1.3708561658859253,
+      "learning_rate": 2e-05,
+      "loss": 0.03996716,
+      "step": 6382
+    },
+    {
+      "epoch": 12.766,
+      "grad_norm": 1.4658493995666504,
+      "learning_rate": 2e-05,
+      "loss": 0.03841384,
+      "step": 6383
+    },
+    {
+      "epoch": 12.768,
+      "grad_norm": 1.5609458684921265,
+      "learning_rate": 2e-05,
+      "loss": 0.04608671,
+      "step": 6384
+    },
+    {
+      "epoch": 12.77,
+      "grad_norm": 1.023798942565918,
+      "learning_rate": 2e-05,
+      "loss": 0.03047959,
+      "step": 6385
+    },
+    {
+      "epoch": 12.772,
+      "grad_norm": 2.012105703353882,
+      "learning_rate": 2e-05,
+      "loss": 0.04641195,
+      "step": 6386
+    },
+    {
+      "epoch": 12.774000000000001,
+      "grad_norm": 1.3531718254089355,
+      "learning_rate": 2e-05,
+      "loss": 0.04071771,
+      "step": 6387
+    },
+    {
+      "epoch": 12.776,
+      "grad_norm": 1.2917602062225342,
+      "learning_rate": 2e-05,
+      "loss": 0.03840836,
+      "step": 6388
+    },
+    {
+      "epoch": 12.778,
+      "grad_norm": 1.412302851676941,
+      "learning_rate": 2e-05,
+      "loss": 0.0444256,
+      "step": 6389
+    },
+    {
+      "epoch": 12.78,
+      "grad_norm": 2.208261489868164,
+      "learning_rate": 2e-05,
+      "loss": 0.03917177,
+      "step": 6390
+    },
+    {
+      "epoch": 12.782,
+      "grad_norm": 2.6841533184051514,
+      "learning_rate": 2e-05,
+      "loss": 0.04568644,
+      "step": 6391
+    },
+    {
+      "epoch": 12.784,
+      "grad_norm": 2.0989646911621094,
+      "learning_rate": 2e-05,
+      "loss": 0.04213623,
+      "step": 6392
+    },
+    {
+      "epoch": 12.786,
+      "grad_norm": 1.1875020265579224,
+      "learning_rate": 2e-05,
+      "loss": 0.03316896,
+      "step": 6393
+    },
+    {
+      "epoch": 12.788,
+      "grad_norm": 1.4427107572555542,
+      "learning_rate": 2e-05,
+      "loss": 0.04718685,
+      "step": 6394
+    },
+    {
+      "epoch": 12.79,
+      "grad_norm": 2.2982850074768066,
+      "learning_rate": 2e-05,
+      "loss": 0.06501245,
+      "step": 6395
+    },
+    {
+      "epoch": 12.792,
+      "grad_norm": 1.4120808839797974,
+      "learning_rate": 2e-05,
+      "loss": 0.04840774,
+      "step": 6396
+    },
+    {
+      "epoch": 12.794,
+      "grad_norm": 1.1496741771697998,
+      "learning_rate": 2e-05,
+      "loss": 0.02569684,
+      "step": 6397
+    },
+    {
+      "epoch": 12.796,
+      "grad_norm": 1.4502733945846558,
+      "learning_rate": 2e-05,
+      "loss": 0.0588194,
+      "step": 6398
+    },
+    {
+      "epoch": 12.798,
+      "grad_norm": 1.6323798894882202,
+      "learning_rate": 2e-05,
+      "loss": 0.04662085,
+      "step": 6399
+    },
+    {
+      "epoch": 12.8,
+      "grad_norm": 1.459981083869934,
+      "learning_rate": 2e-05,
+      "loss": 0.0337071,
+      "step": 6400
+    },
+    {
+      "epoch": 12.802,
+      "grad_norm": 1.2611836194992065,
+      "learning_rate": 2e-05,
+      "loss": 0.03413685,
+      "step": 6401
+    },
+    {
+      "epoch": 12.804,
+      "grad_norm": 1.0556434392929077,
+      "learning_rate": 2e-05,
+      "loss": 0.03022856,
+      "step": 6402
+    },
+    {
+      "epoch": 12.806000000000001,
+      "grad_norm": 1.1885935068130493,
+      "learning_rate": 2e-05,
+      "loss": 0.03547438,
+      "step": 6403
+    },
+    {
+      "epoch": 12.808,
+      "grad_norm": 1.4913263320922852,
+      "learning_rate": 2e-05,
+      "loss": 0.04301264,
+      "step": 6404
+    },
+    {
+      "epoch": 12.81,
+      "grad_norm": 1.678415298461914,
+      "learning_rate": 2e-05,
+      "loss": 0.0440227,
+      "step": 6405
+    },
+    {
+      "epoch": 12.812,
+      "grad_norm": 2.0774550437927246,
+      "learning_rate": 2e-05,
+      "loss": 0.03051781,
+      "step": 6406
+    },
+    {
+      "epoch": 12.814,
+      "grad_norm": 1.5015497207641602,
+      "learning_rate": 2e-05,
+      "loss": 0.04394282,
+      "step": 6407
+    },
+    {
+      "epoch": 12.816,
+      "grad_norm": 1.5424638986587524,
+      "learning_rate": 2e-05,
+      "loss": 0.0474889,
+      "step": 6408
+    },
+    {
+      "epoch": 12.818,
+      "grad_norm": 1.2423845529556274,
+      "learning_rate": 2e-05,
+      "loss": 0.03328826,
+      "step": 6409
+    },
+    {
+      "epoch": 12.82,
+      "grad_norm": 1.3648911714553833,
+      "learning_rate": 2e-05,
+      "loss": 0.05593203,
+      "step": 6410
+    },
+    {
+      "epoch": 12.822,
+      "grad_norm": 1.6377047300338745,
+      "learning_rate": 2e-05,
+      "loss": 0.04586127,
+      "step": 6411
+    },
+    {
+      "epoch": 12.824,
+      "grad_norm": 1.2745712995529175,
+      "learning_rate": 2e-05,
+      "loss": 0.02756302,
+      "step": 6412
+    },
+    {
+      "epoch": 12.826,
+      "grad_norm": 1.0557851791381836,
+      "learning_rate": 2e-05,
+      "loss": 0.02719074,
+      "step": 6413
+    },
+    {
+      "epoch": 12.828,
+      "grad_norm": 1.3740254640579224,
+      "learning_rate": 2e-05,
+      "loss": 0.03186225,
+      "step": 6414
+    },
+    {
+      "epoch": 12.83,
+      "grad_norm": 2.600973129272461,
+      "learning_rate": 2e-05,
+      "loss": 0.04746537,
+      "step": 6415
+    },
+    {
+      "epoch": 12.832,
+      "grad_norm": 1.1148313283920288,
+      "learning_rate": 2e-05,
+      "loss": 0.02911303,
+      "step": 6416
+    },
+    {
+      "epoch": 12.834,
+      "grad_norm": 1.085671067237854,
+      "learning_rate": 2e-05,
+      "loss": 0.03521315,
+      "step": 6417
+    },
+    {
+      "epoch": 12.836,
+      "grad_norm": 2.3671929836273193,
+      "learning_rate": 2e-05,
+      "loss": 0.04562847,
+      "step": 6418
+    },
+    {
+      "epoch": 12.838,
+      "grad_norm": 2.1059203147888184,
+      "learning_rate": 2e-05,
+      "loss": 0.05967949,
+      "step": 6419
+    },
+    {
+      "epoch": 12.84,
+      "grad_norm": 2.0058083534240723,
+      "learning_rate": 2e-05,
+      "loss": 0.03334425,
+      "step": 6420
+    },
+    {
+      "epoch": 12.842,
+      "grad_norm": 1.4104883670806885,
+      "learning_rate": 2e-05,
+      "loss": 0.03960668,
+      "step": 6421
+    },
+    {
+      "epoch": 12.844,
+      "grad_norm": 1.3945311307907104,
+      "learning_rate": 2e-05,
+      "loss": 0.05327337,
+      "step": 6422
+    },
+    {
+      "epoch": 12.846,
+      "grad_norm": 1.1917275190353394,
+      "learning_rate": 2e-05,
+      "loss": 0.03146118,
+      "step": 6423
+    },
+    {
+      "epoch": 12.848,
+      "grad_norm": 1.3188499212265015,
+      "learning_rate": 2e-05,
+      "loss": 0.03480065,
+      "step": 6424
+    },
+    {
+      "epoch": 12.85,
+      "grad_norm": 1.2820628881454468,
+      "learning_rate": 2e-05,
+      "loss": 0.02856812,
+      "step": 6425
+    },
+    {
+      "epoch": 12.852,
+      "grad_norm": 1.3507776260375977,
+      "learning_rate": 2e-05,
+      "loss": 0.0419952,
+      "step": 6426
+    },
+    {
+      "epoch": 12.854,
+      "grad_norm": 1.5551594495773315,
+      "learning_rate": 2e-05,
+      "loss": 0.06390089,
+      "step": 6427
+    },
+    {
+      "epoch": 12.856,
+      "grad_norm": 2.567291259765625,
+      "learning_rate": 2e-05,
+      "loss": 0.04246731,
+      "step": 6428
+    },
+    {
+      "epoch": 12.858,
+      "grad_norm": 1.707859754562378,
+      "learning_rate": 2e-05,
+      "loss": 0.05553305,
+      "step": 6429
+    },
+    {
+      "epoch": 12.86,
+      "grad_norm": 1.1673024892807007,
+      "learning_rate": 2e-05,
+      "loss": 0.04302656,
+      "step": 6430
+    },
+    {
+      "epoch": 12.862,
+      "grad_norm": 1.1114625930786133,
+      "learning_rate": 2e-05,
+      "loss": 0.03706741,
+      "step": 6431
+    },
+    {
+      "epoch": 12.864,
+      "grad_norm": 1.044904112815857,
+      "learning_rate": 2e-05,
+      "loss": 0.03192709,
+      "step": 6432
+    },
+    {
+      "epoch": 12.866,
+      "grad_norm": 1.69417142868042,
+      "learning_rate": 2e-05,
+      "loss": 0.05506884,
+      "step": 6433
+    },
+    {
+      "epoch": 12.868,
+      "grad_norm": 1.2619975805282593,
+      "learning_rate": 2e-05,
+      "loss": 0.03792794,
+      "step": 6434
+    },
+    {
+      "epoch": 12.87,
+      "grad_norm": 2.1460132598876953,
+      "learning_rate": 2e-05,
+      "loss": 0.05722627,
+      "step": 6435
+    },
+    {
+      "epoch": 12.872,
+      "grad_norm": 1.2090710401535034,
+      "learning_rate": 2e-05,
+      "loss": 0.05044911,
+      "step": 6436
+    },
+    {
+      "epoch": 12.874,
+      "grad_norm": 1.7237330675125122,
+      "learning_rate": 2e-05,
+      "loss": 0.0452129,
+      "step": 6437
+    },
+    {
+      "epoch": 12.876,
+      "grad_norm": 1.0395435094833374,
+      "learning_rate": 2e-05,
+      "loss": 0.02874579,
+      "step": 6438
+    },
+    {
+      "epoch": 12.878,
+      "grad_norm": 1.4696030616760254,
+      "learning_rate": 2e-05,
+      "loss": 0.04806498,
+      "step": 6439
+    },
+    {
+      "epoch": 12.88,
+      "grad_norm": 1.6383066177368164,
+      "learning_rate": 2e-05,
+      "loss": 0.03882147,
+      "step": 6440
+    },
+    {
+      "epoch": 12.882,
+      "grad_norm": 1.9339408874511719,
+      "learning_rate": 2e-05,
+      "loss": 0.06159985,
+      "step": 6441
+    },
+    {
+      "epoch": 12.884,
+      "grad_norm": 3.267814874649048,
+      "learning_rate": 2e-05,
+      "loss": 0.05857627,
+      "step": 6442
+    },
+    {
+      "epoch": 12.886,
+      "grad_norm": 2.2561628818511963,
+      "learning_rate": 2e-05,
+      "loss": 0.05107491,
+      "step": 6443
+    },
+    {
+      "epoch": 12.888,
+      "grad_norm": 1.9300895929336548,
+      "learning_rate": 2e-05,
+      "loss": 0.05386838,
+      "step": 6444
+    },
+    {
+      "epoch": 12.89,
+      "grad_norm": 3.2582273483276367,
+      "learning_rate": 2e-05,
+      "loss": 0.04117388,
+      "step": 6445
+    },
+    {
+      "epoch": 12.892,
+      "grad_norm": 1.428175449371338,
+      "learning_rate": 2e-05,
+      "loss": 0.03108019,
+      "step": 6446
+    },
+    {
+      "epoch": 12.894,
+      "grad_norm": 2.0222373008728027,
+      "learning_rate": 2e-05,
+      "loss": 0.03365955,
+      "step": 6447
+    },
+    {
+      "epoch": 12.896,
+      "grad_norm": 2.43520188331604,
+      "learning_rate": 2e-05,
+      "loss": 0.04152467,
+      "step": 6448
+    },
+    {
+      "epoch": 12.898,
+      "grad_norm": 1.3054020404815674,
+      "learning_rate": 2e-05,
+      "loss": 0.03886139,
+      "step": 6449
+    },
+    {
+      "epoch": 12.9,
+      "grad_norm": 1.3858981132507324,
+      "learning_rate": 2e-05,
+      "loss": 0.03895619,
+      "step": 6450
+    },
+    {
+      "epoch": 12.902,
+      "grad_norm": 0.9143564105033875,
+      "learning_rate": 2e-05,
+      "loss": 0.01918302,
+      "step": 6451
+    },
+    {
+      "epoch": 12.904,
+      "grad_norm": 1.4373931884765625,
+      "learning_rate": 2e-05,
+      "loss": 0.04344804,
+      "step": 6452
+    },
+    {
+      "epoch": 12.906,
+      "grad_norm": 1.0674983263015747,
+      "learning_rate": 2e-05,
+      "loss": 0.03236082,
+      "step": 6453
+    },
+    {
+      "epoch": 12.908,
+      "grad_norm": 1.2760518789291382,
+      "learning_rate": 2e-05,
+      "loss": 0.04398597,
+      "step": 6454
+    },
+    {
+      "epoch": 12.91,
+      "grad_norm": 2.119486093521118,
+      "learning_rate": 2e-05,
+      "loss": 0.03598646,
+      "step": 6455
+    },
+    {
+      "epoch": 12.912,
+      "grad_norm": 1.0327049493789673,
+      "learning_rate": 2e-05,
+      "loss": 0.02299444,
+      "step": 6456
+    },
+    {
+      "epoch": 12.914,
+      "grad_norm": 1.3598545789718628,
+      "learning_rate": 2e-05,
+      "loss": 0.04965741,
+      "step": 6457
+    },
+    {
+      "epoch": 12.916,
+      "grad_norm": 1.109059453010559,
+      "learning_rate": 2e-05,
+      "loss": 0.0339277,
+      "step": 6458
+    },
+    {
+      "epoch": 12.918,
+      "grad_norm": 1.2943098545074463,
+      "learning_rate": 2e-05,
+      "loss": 0.03731314,
+      "step": 6459
+    },
+    {
+      "epoch": 12.92,
+      "grad_norm": 2.1976876258850098,
+      "learning_rate": 2e-05,
+      "loss": 0.0418474,
+      "step": 6460
+    },
+    {
+      "epoch": 12.922,
+      "grad_norm": 0.9695811867713928,
+      "learning_rate": 2e-05,
+      "loss": 0.02404466,
+      "step": 6461
+    },
+    {
+      "epoch": 12.924,
+      "grad_norm": 1.4028820991516113,
+      "learning_rate": 2e-05,
+      "loss": 0.04700348,
+      "step": 6462
+    },
+    {
+      "epoch": 12.926,
+      "grad_norm": 2.074523448944092,
+      "learning_rate": 2e-05,
+      "loss": 0.04677882,
+      "step": 6463
+    },
+    {
+      "epoch": 12.928,
+      "grad_norm": 0.876520574092865,
+      "learning_rate": 2e-05,
+      "loss": 0.02563859,
+      "step": 6464
+    },
+    {
+      "epoch": 12.93,
+      "grad_norm": 1.8341282606124878,
+      "learning_rate": 2e-05,
+      "loss": 0.04422235,
+      "step": 6465
+    },
+    {
+      "epoch": 12.932,
+      "grad_norm": 1.3097070455551147,
+      "learning_rate": 2e-05,
+      "loss": 0.03325322,
+      "step": 6466
+    },
+    {
+      "epoch": 12.934,
+      "grad_norm": 1.4689263105392456,
+      "learning_rate": 2e-05,
+      "loss": 0.04611597,
+      "step": 6467
+    },
+    {
+      "epoch": 12.936,
+      "grad_norm": 1.5380563735961914,
+      "learning_rate": 2e-05,
+      "loss": 0.05122569,
+      "step": 6468
+    },
+    {
+      "epoch": 12.938,
+      "grad_norm": 0.8768236637115479,
+      "learning_rate": 2e-05,
+      "loss": 0.01557497,
+      "step": 6469
+    },
+    {
+      "epoch": 12.94,
+      "grad_norm": 3.0059239864349365,
+      "learning_rate": 2e-05,
+      "loss": 0.03941962,
+      "step": 6470
+    },
+    {
+      "epoch": 12.942,
+      "grad_norm": 4.760193347930908,
+      "learning_rate": 2e-05,
+      "loss": 0.03318344,
+      "step": 6471
+    },
+    {
+      "epoch": 12.943999999999999,
+      "grad_norm": 2.0453317165374756,
+      "learning_rate": 2e-05,
+      "loss": 0.04006668,
+      "step": 6472
+    },
+    {
+      "epoch": 12.946,
+      "grad_norm": 2.809612512588501,
+      "learning_rate": 2e-05,
+      "loss": 0.04919286,
+      "step": 6473
+    },
+    {
+      "epoch": 12.948,
+      "grad_norm": 1.04146146774292,
+      "learning_rate": 2e-05,
+      "loss": 0.0355149,
+      "step": 6474
+    },
+    {
+      "epoch": 12.95,
+      "grad_norm": 1.4496123790740967,
+      "learning_rate": 2e-05,
+      "loss": 0.04766847,
+      "step": 6475
+    },
+    {
+      "epoch": 12.952,
+      "grad_norm": 1.7231749296188354,
+      "learning_rate": 2e-05,
+      "loss": 0.03629901,
+      "step": 6476
+    },
+    {
+      "epoch": 12.954,
+      "grad_norm": 1.4813939332962036,
+      "learning_rate": 2e-05,
+      "loss": 0.05657486,
+      "step": 6477
+    },
+    {
+      "epoch": 12.956,
+      "grad_norm": 1.8923465013504028,
+      "learning_rate": 2e-05,
+      "loss": 0.03107157,
+      "step": 6478
+    },
+    {
+      "epoch": 12.958,
+      "grad_norm": 1.1124436855316162,
+      "learning_rate": 2e-05,
+      "loss": 0.03300238,
+      "step": 6479
+    },
+    {
+      "epoch": 12.96,
+      "grad_norm": 1.609631061553955,
+      "learning_rate": 2e-05,
+      "loss": 0.05744763,
+      "step": 6480
+    },
+    {
+      "epoch": 12.962,
+      "grad_norm": 1.092146396636963,
+      "learning_rate": 2e-05,
+      "loss": 0.0354538,
+      "step": 6481
+    },
+    {
+      "epoch": 12.964,
+      "grad_norm": 1.0714373588562012,
+      "learning_rate": 2e-05,
+      "loss": 0.03718103,
+      "step": 6482
+    },
+    {
+      "epoch": 12.966,
+      "grad_norm": 1.5786356925964355,
+      "learning_rate": 2e-05,
+      "loss": 0.04812677,
+      "step": 6483
+    },
+    {
+      "epoch": 12.968,
+      "grad_norm": 2.6054508686065674,
+      "learning_rate": 2e-05,
+      "loss": 0.0359851,
+      "step": 6484
+    },
+    {
+      "epoch": 12.97,
+      "grad_norm": 0.9874547123908997,
+      "learning_rate": 2e-05,
+      "loss": 0.03105699,
+      "step": 6485
+    },
+    {
+      "epoch": 12.972,
+      "grad_norm": 2.4184865951538086,
+      "learning_rate": 2e-05,
+      "loss": 0.05367502,
+      "step": 6486
+    },
+    {
+      "epoch": 12.974,
+      "grad_norm": 3.007775068283081,
+      "learning_rate": 2e-05,
+      "loss": 0.0452963,
+      "step": 6487
+    },
+    {
+      "epoch": 12.975999999999999,
+      "grad_norm": 1.4173858165740967,
+      "learning_rate": 2e-05,
+      "loss": 0.05711803,
+      "step": 6488
+    },
+    {
+      "epoch": 12.978,
+      "grad_norm": 1.6504642963409424,
+      "learning_rate": 2e-05,
+      "loss": 0.04591904,
+      "step": 6489
+    },
+    {
+      "epoch": 12.98,
+      "grad_norm": 1.0074936151504517,
+      "learning_rate": 2e-05,
+      "loss": 0.02326766,
+      "step": 6490
+    },
+    {
+      "epoch": 12.982,
+      "grad_norm": 1.0605522394180298,
+      "learning_rate": 2e-05,
+      "loss": 0.02780467,
+      "step": 6491
+    },
+    {
+      "epoch": 12.984,
+      "grad_norm": 2.823035717010498,
+      "learning_rate": 2e-05,
+      "loss": 0.0511883,
+      "step": 6492
+    },
+    {
+      "epoch": 12.986,
+      "grad_norm": 1.2774420976638794,
+      "learning_rate": 2e-05,
+      "loss": 0.0274408,
+      "step": 6493
+    },
+    {
+      "epoch": 12.988,
+      "grad_norm": 0.9702324867248535,
+      "learning_rate": 2e-05,
+      "loss": 0.02415428,
+      "step": 6494
+    },
+    {
+      "epoch": 12.99,
+      "grad_norm": 1.212349534034729,
+      "learning_rate": 2e-05,
+      "loss": 0.04195923,
+      "step": 6495
+    },
+    {
+      "epoch": 12.992,
+      "grad_norm": 1.8738391399383545,
+      "learning_rate": 2e-05,
+      "loss": 0.03428925,
+      "step": 6496
+    },
+    {
+      "epoch": 12.994,
+      "grad_norm": 1.183615803718567,
+      "learning_rate": 2e-05,
+      "loss": 0.03930584,
+      "step": 6497
+    },
+    {
+      "epoch": 12.996,
+      "grad_norm": 3.319544553756714,
+      "learning_rate": 2e-05,
+      "loss": 0.0459759,
+      "step": 6498
+    },
+    {
+      "epoch": 12.998,
+      "grad_norm": 1.1456108093261719,
+      "learning_rate": 2e-05,
+      "loss": 0.03367121,
+      "step": 6499
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 1.9047518968582153,
+      "learning_rate": 2e-05,
+      "loss": 0.04684749,
+      "step": 6500
+    },
+    {
+      "epoch": 13.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.996,
+        "AngleClassification_2": 0.996,
+        "AngleClassification_3": 0.9580838323353293,
+        "Equal_1": 0.996,
+        "Equal_2": 0.9700598802395209,
+        "Equal_3": 0.8882235528942116,
+        "LineComparison_1": 0.998,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9860279441117764,
+        "Parallel_1": 0.9939879759519038,
+        "Parallel_2": 0.9979959919839679,
+        "Parallel_3": 0.992,
+        "Perpendicular_1": 0.984,
+        "Perpendicular_2": 0.966,
+        "Perpendicular_3": 0.7144288577154309,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9916666666666667,
+        "PointLiesOnCircle_3": 0.9896666666666667,
+        "PointLiesOnLine_1": 0.9819639278557114,
+        "PointLiesOnLine_2": 0.9919839679358717,
+        "PointLiesOnLine_3": 0.9680638722554891
+      },
+      "eval_runtime": 319.563,
+      "eval_samples_per_second": 32.857,
+      "eval_steps_per_second": 0.657,
+      "step": 6500
+    },
+    {
+      "epoch": 13.002,
+      "grad_norm": 2.025676727294922,
+      "learning_rate": 2e-05,
+      "loss": 0.05510117,
+      "step": 6501
+    },
+    {
+      "epoch": 13.004,
+      "grad_norm": 1.67191481590271,
+      "learning_rate": 2e-05,
+      "loss": 0.05629362,
+      "step": 6502
+    },
+    {
+      "epoch": 13.006,
+      "grad_norm": 1.4045917987823486,
+      "learning_rate": 2e-05,
+      "loss": 0.0437433,
+      "step": 6503
+    },
+    {
+      "epoch": 13.008,
+      "grad_norm": 2.3563649654388428,
+      "learning_rate": 2e-05,
+      "loss": 0.04698532,
+      "step": 6504
+    },
+    {
+      "epoch": 13.01,
+      "grad_norm": 1.7476520538330078,
+      "learning_rate": 2e-05,
+      "loss": 0.05161941,
+      "step": 6505
+    },
+    {
+      "epoch": 13.012,
+      "grad_norm": 1.3390698432922363,
+      "learning_rate": 2e-05,
+      "loss": 0.03999062,
+      "step": 6506
+    },
+    {
+      "epoch": 13.014,
+      "grad_norm": 1.905305027961731,
+      "learning_rate": 2e-05,
+      "loss": 0.04302887,
+      "step": 6507
+    },
+    {
+      "epoch": 13.016,
+      "grad_norm": 2.7333922386169434,
+      "learning_rate": 2e-05,
+      "loss": 0.04309512,
+      "step": 6508
+    },
+    {
+      "epoch": 13.018,
+      "grad_norm": 1.2732406854629517,
+      "learning_rate": 2e-05,
+      "loss": 0.04579434,
+      "step": 6509
+    },
+    {
+      "epoch": 13.02,
+      "grad_norm": 2.010542631149292,
+      "learning_rate": 2e-05,
+      "loss": 0.04244982,
+      "step": 6510
+    },
+    {
+      "epoch": 13.022,
+      "grad_norm": 1.6808205842971802,
+      "learning_rate": 2e-05,
+      "loss": 0.04541232,
+      "step": 6511
+    },
+    {
+      "epoch": 13.024,
+      "grad_norm": 1.3999242782592773,
+      "learning_rate": 2e-05,
+      "loss": 0.05283143,
+      "step": 6512
+    },
+    {
+      "epoch": 13.026,
+      "grad_norm": 1.0839227437973022,
+      "learning_rate": 2e-05,
+      "loss": 0.03332079,
+      "step": 6513
+    },
+    {
+      "epoch": 13.028,
+      "grad_norm": 1.7581884860992432,
+      "learning_rate": 2e-05,
+      "loss": 0.03697137,
+      "step": 6514
+    },
+    {
+      "epoch": 13.03,
+      "grad_norm": 2.3362925052642822,
+      "learning_rate": 2e-05,
+      "loss": 0.04086022,
+      "step": 6515
+    },
+    {
+      "epoch": 13.032,
+      "grad_norm": 1.2541543245315552,
+      "learning_rate": 2e-05,
+      "loss": 0.04349049,
+      "step": 6516
+    },
+    {
+      "epoch": 13.034,
+      "grad_norm": 3.33878231048584,
+      "learning_rate": 2e-05,
+      "loss": 0.04347111,
+      "step": 6517
+    },
+    {
+      "epoch": 13.036,
+      "grad_norm": 1.553660273551941,
+      "learning_rate": 2e-05,
+      "loss": 0.03859802,
+      "step": 6518
+    },
+    {
+      "epoch": 13.038,
+      "grad_norm": 1.1606173515319824,
+      "learning_rate": 2e-05,
+      "loss": 0.04000216,
+      "step": 6519
+    },
+    {
+      "epoch": 13.04,
+      "grad_norm": 1.0634113550186157,
+      "learning_rate": 2e-05,
+      "loss": 0.02386756,
+      "step": 6520
+    },
+    {
+      "epoch": 13.042,
+      "grad_norm": 1.356263518333435,
+      "learning_rate": 2e-05,
+      "loss": 0.03155567,
+      "step": 6521
+    },
+    {
+      "epoch": 13.044,
+      "grad_norm": 2.3392069339752197,
+      "learning_rate": 2e-05,
+      "loss": 0.06134815,
+      "step": 6522
+    },
+    {
+      "epoch": 13.046,
+      "grad_norm": 4.142246723175049,
+      "learning_rate": 2e-05,
+      "loss": 0.06078862,
+      "step": 6523
+    },
+    {
+      "epoch": 13.048,
+      "grad_norm": 2.212329387664795,
+      "learning_rate": 2e-05,
+      "loss": 0.04393127,
+      "step": 6524
+    },
+    {
+      "epoch": 13.05,
+      "grad_norm": 1.69950270652771,
+      "learning_rate": 2e-05,
+      "loss": 0.04409553,
+      "step": 6525
+    },
+    {
+      "epoch": 13.052,
+      "grad_norm": 1.850864052772522,
+      "learning_rate": 2e-05,
+      "loss": 0.03946983,
+      "step": 6526
+    },
+    {
+      "epoch": 13.054,
+      "grad_norm": 1.1603494882583618,
+      "learning_rate": 2e-05,
+      "loss": 0.02313782,
+      "step": 6527
+    },
+    {
+      "epoch": 13.056,
+      "grad_norm": 2.081894636154175,
+      "learning_rate": 2e-05,
+      "loss": 0.02774069,
+      "step": 6528
+    },
+    {
+      "epoch": 13.058,
+      "grad_norm": 1.8797224760055542,
+      "learning_rate": 2e-05,
+      "loss": 0.05306084,
+      "step": 6529
+    },
+    {
+      "epoch": 13.06,
+      "grad_norm": 1.903994083404541,
+      "learning_rate": 2e-05,
+      "loss": 0.05133477,
+      "step": 6530
+    },
+    {
+      "epoch": 13.062,
+      "grad_norm": 1.7151387929916382,
+      "learning_rate": 2e-05,
+      "loss": 0.04850292,
+      "step": 6531
+    },
+    {
+      "epoch": 13.064,
+      "grad_norm": 2.1803505420684814,
+      "learning_rate": 2e-05,
+      "loss": 0.03637154,
+      "step": 6532
+    },
+    {
+      "epoch": 13.066,
+      "grad_norm": 1.204275369644165,
+      "learning_rate": 2e-05,
+      "loss": 0.04634327,
+      "step": 6533
+    },
+    {
+      "epoch": 13.068,
+      "grad_norm": 1.7479034662246704,
+      "learning_rate": 2e-05,
+      "loss": 0.04111974,
+      "step": 6534
+    },
+    {
+      "epoch": 13.07,
+      "grad_norm": 1.4405919313430786,
+      "learning_rate": 2e-05,
+      "loss": 0.03936667,
+      "step": 6535
+    },
+    {
+      "epoch": 13.072,
+      "grad_norm": 2.053112030029297,
+      "learning_rate": 2e-05,
+      "loss": 0.05977479,
+      "step": 6536
+    },
+    {
+      "epoch": 13.074,
+      "grad_norm": 3.923091173171997,
+      "learning_rate": 2e-05,
+      "loss": 0.03060072,
+      "step": 6537
+    },
+    {
+      "epoch": 13.076,
+      "grad_norm": 1.1463172435760498,
+      "learning_rate": 2e-05,
+      "loss": 0.03833826,
+      "step": 6538
+    },
+    {
+      "epoch": 13.078,
+      "grad_norm": 1.4521420001983643,
+      "learning_rate": 2e-05,
+      "loss": 0.02565883,
+      "step": 6539
+    },
+    {
+      "epoch": 13.08,
+      "grad_norm": 1.5646151304244995,
+      "learning_rate": 2e-05,
+      "loss": 0.04049197,
+      "step": 6540
+    },
+    {
+      "epoch": 13.082,
+      "grad_norm": 1.1040815114974976,
+      "learning_rate": 2e-05,
+      "loss": 0.04397867,
+      "step": 6541
+    },
+    {
+      "epoch": 13.084,
+      "grad_norm": 0.9532331228256226,
+      "learning_rate": 2e-05,
+      "loss": 0.02361744,
+      "step": 6542
+    },
+    {
+      "epoch": 13.086,
+      "grad_norm": 3.3594508171081543,
+      "learning_rate": 2e-05,
+      "loss": 0.06016827,
+      "step": 6543
+    },
+    {
+      "epoch": 13.088,
+      "grad_norm": 1.7429322004318237,
+      "learning_rate": 2e-05,
+      "loss": 0.04198964,
+      "step": 6544
+    },
+    {
+      "epoch": 13.09,
+      "grad_norm": 1.6168110370635986,
+      "learning_rate": 2e-05,
+      "loss": 0.02668218,
+      "step": 6545
+    },
+    {
+      "epoch": 13.092,
+      "grad_norm": 1.9579505920410156,
+      "learning_rate": 2e-05,
+      "loss": 0.0469409,
+      "step": 6546
+    },
+    {
+      "epoch": 13.094,
+      "grad_norm": 1.6008516550064087,
+      "learning_rate": 2e-05,
+      "loss": 0.04403362,
+      "step": 6547
+    },
+    {
+      "epoch": 13.096,
+      "grad_norm": 2.0377395153045654,
+      "learning_rate": 2e-05,
+      "loss": 0.04835338,
+      "step": 6548
+    },
+    {
+      "epoch": 13.098,
+      "grad_norm": 2.0079846382141113,
+      "learning_rate": 2e-05,
+      "loss": 0.04270446,
+      "step": 6549
+    },
+    {
+      "epoch": 13.1,
+      "grad_norm": 1.2587766647338867,
+      "learning_rate": 2e-05,
+      "loss": 0.04245213,
+      "step": 6550
+    },
+    {
+      "epoch": 13.102,
+      "grad_norm": 1.3419069051742554,
+      "learning_rate": 2e-05,
+      "loss": 0.03273055,
+      "step": 6551
+    },
+    {
+      "epoch": 13.104,
+      "grad_norm": 1.0695240497589111,
+      "learning_rate": 2e-05,
+      "loss": 0.02955955,
+      "step": 6552
+    },
+    {
+      "epoch": 13.106,
+      "grad_norm": 1.2960338592529297,
+      "learning_rate": 2e-05,
+      "loss": 0.03386617,
+      "step": 6553
+    },
+    {
+      "epoch": 13.108,
+      "grad_norm": 1.3421612977981567,
+      "learning_rate": 2e-05,
+      "loss": 0.05039021,
+      "step": 6554
+    },
+    {
+      "epoch": 13.11,
+      "grad_norm": 2.230473518371582,
+      "learning_rate": 2e-05,
+      "loss": 0.05388996,
+      "step": 6555
+    },
+    {
+      "epoch": 13.112,
+      "grad_norm": 1.5305428504943848,
+      "learning_rate": 2e-05,
+      "loss": 0.03787341,
+      "step": 6556
+    },
+    {
+      "epoch": 13.114,
+      "grad_norm": 1.2190325260162354,
+      "learning_rate": 2e-05,
+      "loss": 0.04556718,
+      "step": 6557
+    },
+    {
+      "epoch": 13.116,
+      "grad_norm": 1.5937390327453613,
+      "learning_rate": 2e-05,
+      "loss": 0.04280414,
+      "step": 6558
+    },
+    {
+      "epoch": 13.118,
+      "grad_norm": 1.9277565479278564,
+      "learning_rate": 2e-05,
+      "loss": 0.04627956,
+      "step": 6559
+    },
+    {
+      "epoch": 13.12,
+      "grad_norm": 1.2560091018676758,
+      "learning_rate": 2e-05,
+      "loss": 0.04250285,
+      "step": 6560
+    },
+    {
+      "epoch": 13.122,
+      "grad_norm": 1.1613867282867432,
+      "learning_rate": 2e-05,
+      "loss": 0.0368138,
+      "step": 6561
+    },
+    {
+      "epoch": 13.124,
+      "grad_norm": 1.8732789754867554,
+      "learning_rate": 2e-05,
+      "loss": 0.04468385,
+      "step": 6562
+    },
+    {
+      "epoch": 13.126,
+      "grad_norm": 1.9750572443008423,
+      "learning_rate": 2e-05,
+      "loss": 0.03603955,
+      "step": 6563
+    },
+    {
+      "epoch": 13.128,
+      "grad_norm": 1.3057758808135986,
+      "learning_rate": 2e-05,
+      "loss": 0.03955033,
+      "step": 6564
+    },
+    {
+      "epoch": 13.13,
+      "grad_norm": 2.027482032775879,
+      "learning_rate": 2e-05,
+      "loss": 0.05527831,
+      "step": 6565
+    },
+    {
+      "epoch": 13.132,
+      "grad_norm": 1.953943133354187,
+      "learning_rate": 2e-05,
+      "loss": 0.03958747,
+      "step": 6566
+    },
+    {
+      "epoch": 13.134,
+      "grad_norm": 1.0840398073196411,
+      "learning_rate": 2e-05,
+      "loss": 0.03809543,
+      "step": 6567
+    },
+    {
+      "epoch": 13.136,
+      "grad_norm": 2.0900790691375732,
+      "learning_rate": 2e-05,
+      "loss": 0.04004861,
+      "step": 6568
+    },
+    {
+      "epoch": 13.138,
+      "grad_norm": 1.2721304893493652,
+      "learning_rate": 2e-05,
+      "loss": 0.0365794,
+      "step": 6569
+    },
+    {
+      "epoch": 13.14,
+      "grad_norm": 1.4358141422271729,
+      "learning_rate": 2e-05,
+      "loss": 0.04888467,
+      "step": 6570
+    },
+    {
+      "epoch": 13.142,
+      "grad_norm": 1.0475200414657593,
+      "learning_rate": 2e-05,
+      "loss": 0.03867028,
+      "step": 6571
+    },
+    {
+      "epoch": 13.144,
+      "grad_norm": 1.4733918905258179,
+      "learning_rate": 2e-05,
+      "loss": 0.03798759,
+      "step": 6572
+    },
+    {
+      "epoch": 13.146,
+      "grad_norm": 1.2983602285385132,
+      "learning_rate": 2e-05,
+      "loss": 0.04571147,
+      "step": 6573
+    },
+    {
+      "epoch": 13.148,
+      "grad_norm": 1.9528359174728394,
+      "learning_rate": 2e-05,
+      "loss": 0.06328858,
+      "step": 6574
+    },
+    {
+      "epoch": 13.15,
+      "grad_norm": 2.274611234664917,
+      "learning_rate": 2e-05,
+      "loss": 0.05201408,
+      "step": 6575
+    },
+    {
+      "epoch": 13.152,
+      "grad_norm": 1.6580023765563965,
+      "learning_rate": 2e-05,
+      "loss": 0.04882085,
+      "step": 6576
+    },
+    {
+      "epoch": 13.154,
+      "grad_norm": 1.701705813407898,
+      "learning_rate": 2e-05,
+      "loss": 0.06361747,
+      "step": 6577
+    },
+    {
+      "epoch": 13.156,
+      "grad_norm": 1.3338755369186401,
+      "learning_rate": 2e-05,
+      "loss": 0.03825006,
+      "step": 6578
+    },
+    {
+      "epoch": 13.158,
+      "grad_norm": 1.2740284204483032,
+      "learning_rate": 2e-05,
+      "loss": 0.04740117,
+      "step": 6579
+    },
+    {
+      "epoch": 13.16,
+      "grad_norm": 0.855344831943512,
+      "learning_rate": 2e-05,
+      "loss": 0.02979115,
+      "step": 6580
+    },
+    {
+      "epoch": 13.162,
+      "grad_norm": 1.0931096076965332,
+      "learning_rate": 2e-05,
+      "loss": 0.03413231,
+      "step": 6581
+    },
+    {
+      "epoch": 13.164,
+      "grad_norm": 1.7539303302764893,
+      "learning_rate": 2e-05,
+      "loss": 0.03713946,
+      "step": 6582
+    },
+    {
+      "epoch": 13.166,
+      "grad_norm": 1.2781082391738892,
+      "learning_rate": 2e-05,
+      "loss": 0.0324952,
+      "step": 6583
+    },
+    {
+      "epoch": 13.168,
+      "grad_norm": 1.100340723991394,
+      "learning_rate": 2e-05,
+      "loss": 0.04220153,
+      "step": 6584
+    },
+    {
+      "epoch": 13.17,
+      "grad_norm": 1.4666935205459595,
+      "learning_rate": 2e-05,
+      "loss": 0.04995844,
+      "step": 6585
+    },
+    {
+      "epoch": 13.172,
+      "grad_norm": 0.8306145668029785,
+      "learning_rate": 2e-05,
+      "loss": 0.0289952,
+      "step": 6586
+    },
+    {
+      "epoch": 13.174,
+      "grad_norm": 1.7973495721817017,
+      "learning_rate": 2e-05,
+      "loss": 0.03114134,
+      "step": 6587
+    },
+    {
+      "epoch": 13.176,
+      "grad_norm": 1.2895548343658447,
+      "learning_rate": 2e-05,
+      "loss": 0.04376809,
+      "step": 6588
+    },
+    {
+      "epoch": 13.178,
+      "grad_norm": 4.245728015899658,
+      "learning_rate": 2e-05,
+      "loss": 0.07614356,
+      "step": 6589
+    },
+    {
+      "epoch": 13.18,
+      "grad_norm": 1.9204843044281006,
+      "learning_rate": 2e-05,
+      "loss": 0.05193898,
+      "step": 6590
+    },
+    {
+      "epoch": 13.182,
+      "grad_norm": 1.8683027029037476,
+      "learning_rate": 2e-05,
+      "loss": 0.02822795,
+      "step": 6591
+    },
+    {
+      "epoch": 13.184,
+      "grad_norm": 0.9471633434295654,
+      "learning_rate": 2e-05,
+      "loss": 0.02922757,
+      "step": 6592
+    },
+    {
+      "epoch": 13.186,
+      "grad_norm": 1.7378665208816528,
+      "learning_rate": 2e-05,
+      "loss": 0.05698539,
+      "step": 6593
+    },
+    {
+      "epoch": 13.188,
+      "grad_norm": 1.6019706726074219,
+      "learning_rate": 2e-05,
+      "loss": 0.05355989,
+      "step": 6594
+    },
+    {
+      "epoch": 13.19,
+      "grad_norm": 1.856338381767273,
+      "learning_rate": 2e-05,
+      "loss": 0.04503391,
+      "step": 6595
+    },
+    {
+      "epoch": 13.192,
+      "grad_norm": 1.4183052778244019,
+      "learning_rate": 2e-05,
+      "loss": 0.05698609,
+      "step": 6596
+    },
+    {
+      "epoch": 13.194,
+      "grad_norm": 2.9402050971984863,
+      "learning_rate": 2e-05,
+      "loss": 0.05300699,
+      "step": 6597
+    },
+    {
+      "epoch": 13.196,
+      "grad_norm": 1.2691911458969116,
+      "learning_rate": 2e-05,
+      "loss": 0.04289351,
+      "step": 6598
+    },
+    {
+      "epoch": 13.198,
+      "grad_norm": 1.9946510791778564,
+      "learning_rate": 2e-05,
+      "loss": 0.03787369,
+      "step": 6599
+    },
+    {
+      "epoch": 13.2,
+      "grad_norm": 1.5389318466186523,
+      "learning_rate": 2e-05,
+      "loss": 0.04896133,
+      "step": 6600
+    },
+    {
+      "epoch": 13.202,
+      "grad_norm": 1.1009681224822998,
+      "learning_rate": 2e-05,
+      "loss": 0.04099545,
+      "step": 6601
+    },
+    {
+      "epoch": 13.204,
+      "grad_norm": 1.298155665397644,
+      "learning_rate": 2e-05,
+      "loss": 0.04521285,
+      "step": 6602
+    },
+    {
+      "epoch": 13.206,
+      "grad_norm": 1.194268822669983,
+      "learning_rate": 2e-05,
+      "loss": 0.02646183,
+      "step": 6603
+    },
+    {
+      "epoch": 13.208,
+      "grad_norm": 1.166735053062439,
+      "learning_rate": 2e-05,
+      "loss": 0.03302169,
+      "step": 6604
+    },
+    {
+      "epoch": 13.21,
+      "grad_norm": 1.1501606702804565,
+      "learning_rate": 2e-05,
+      "loss": 0.03696949,
+      "step": 6605
+    },
+    {
+      "epoch": 13.212,
+      "grad_norm": 1.2692323923110962,
+      "learning_rate": 2e-05,
+      "loss": 0.04398295,
+      "step": 6606
+    },
+    {
+      "epoch": 13.214,
+      "grad_norm": 1.6035974025726318,
+      "learning_rate": 2e-05,
+      "loss": 0.03624269,
+      "step": 6607
+    },
+    {
+      "epoch": 13.216,
+      "grad_norm": 1.2560043334960938,
+      "learning_rate": 2e-05,
+      "loss": 0.03458342,
+      "step": 6608
+    },
+    {
+      "epoch": 13.218,
+      "grad_norm": 2.6903185844421387,
+      "learning_rate": 2e-05,
+      "loss": 0.03339992,
+      "step": 6609
+    },
+    {
+      "epoch": 13.22,
+      "grad_norm": 1.8403470516204834,
+      "learning_rate": 2e-05,
+      "loss": 0.03888453,
+      "step": 6610
+    },
+    {
+      "epoch": 13.222,
+      "grad_norm": 1.1549571752548218,
+      "learning_rate": 2e-05,
+      "loss": 0.03554495,
+      "step": 6611
+    },
+    {
+      "epoch": 13.224,
+      "grad_norm": 2.7406699657440186,
+      "learning_rate": 2e-05,
+      "loss": 0.04288154,
+      "step": 6612
+    },
+    {
+      "epoch": 13.226,
+      "grad_norm": 1.6124176979064941,
+      "learning_rate": 2e-05,
+      "loss": 0.04322715,
+      "step": 6613
+    },
+    {
+      "epoch": 13.228,
+      "grad_norm": 1.271803617477417,
+      "learning_rate": 2e-05,
+      "loss": 0.04564525,
+      "step": 6614
+    },
+    {
+      "epoch": 13.23,
+      "grad_norm": 2.073120594024658,
+      "learning_rate": 2e-05,
+      "loss": 0.04652428,
+      "step": 6615
+    },
+    {
+      "epoch": 13.232,
+      "grad_norm": 1.5669819116592407,
+      "learning_rate": 2e-05,
+      "loss": 0.04134148,
+      "step": 6616
+    },
+    {
+      "epoch": 13.234,
+      "grad_norm": 0.8931474685668945,
+      "learning_rate": 2e-05,
+      "loss": 0.03444675,
+      "step": 6617
+    },
+    {
+      "epoch": 13.236,
+      "grad_norm": 2.4369468688964844,
+      "learning_rate": 2e-05,
+      "loss": 0.04590879,
+      "step": 6618
+    },
+    {
+      "epoch": 13.238,
+      "grad_norm": 2.894359588623047,
+      "learning_rate": 2e-05,
+      "loss": 0.04611806,
+      "step": 6619
+    },
+    {
+      "epoch": 13.24,
+      "grad_norm": 1.5769118070602417,
+      "learning_rate": 2e-05,
+      "loss": 0.03493973,
+      "step": 6620
+    },
+    {
+      "epoch": 13.242,
+      "grad_norm": 1.3417714834213257,
+      "learning_rate": 2e-05,
+      "loss": 0.04988127,
+      "step": 6621
+    },
+    {
+      "epoch": 13.244,
+      "grad_norm": 1.8980623483657837,
+      "learning_rate": 2e-05,
+      "loss": 0.05576314,
+      "step": 6622
+    },
+    {
+      "epoch": 13.246,
+      "grad_norm": 1.0693109035491943,
+      "learning_rate": 2e-05,
+      "loss": 0.03685558,
+      "step": 6623
+    },
+    {
+      "epoch": 13.248,
+      "grad_norm": 1.5265095233917236,
+      "learning_rate": 2e-05,
+      "loss": 0.05537786,
+      "step": 6624
+    },
+    {
+      "epoch": 13.25,
+      "grad_norm": 2.378041982650757,
+      "learning_rate": 2e-05,
+      "loss": 0.03925188,
+      "step": 6625
+    },
+    {
+      "epoch": 13.252,
+      "grad_norm": 1.2043821811676025,
+      "learning_rate": 2e-05,
+      "loss": 0.030987,
+      "step": 6626
+    },
+    {
+      "epoch": 13.254,
+      "grad_norm": 1.6501712799072266,
+      "learning_rate": 2e-05,
+      "loss": 0.04117061,
+      "step": 6627
+    },
+    {
+      "epoch": 13.256,
+      "grad_norm": 2.1596996784210205,
+      "learning_rate": 2e-05,
+      "loss": 0.05108644,
+      "step": 6628
+    },
+    {
+      "epoch": 13.258,
+      "grad_norm": 2.158721446990967,
+      "learning_rate": 2e-05,
+      "loss": 0.04933317,
+      "step": 6629
+    },
+    {
+      "epoch": 13.26,
+      "grad_norm": 1.9402283430099487,
+      "learning_rate": 2e-05,
+      "loss": 0.04095349,
+      "step": 6630
+    },
+    {
+      "epoch": 13.262,
+      "grad_norm": 1.688955307006836,
+      "learning_rate": 2e-05,
+      "loss": 0.03976238,
+      "step": 6631
+    },
+    {
+      "epoch": 13.264,
+      "grad_norm": 1.8904914855957031,
+      "learning_rate": 2e-05,
+      "loss": 0.05259514,
+      "step": 6632
+    },
+    {
+      "epoch": 13.266,
+      "grad_norm": 1.3631291389465332,
+      "learning_rate": 2e-05,
+      "loss": 0.04219753,
+      "step": 6633
+    },
+    {
+      "epoch": 13.268,
+      "grad_norm": 1.141372799873352,
+      "learning_rate": 2e-05,
+      "loss": 0.03875969,
+      "step": 6634
+    },
+    {
+      "epoch": 13.27,
+      "grad_norm": 1.0180898904800415,
+      "learning_rate": 2e-05,
+      "loss": 0.02585149,
+      "step": 6635
+    },
+    {
+      "epoch": 13.272,
+      "grad_norm": 1.6760014295578003,
+      "learning_rate": 2e-05,
+      "loss": 0.04044281,
+      "step": 6636
+    },
+    {
+      "epoch": 13.274000000000001,
+      "grad_norm": 2.0628297328948975,
+      "learning_rate": 2e-05,
+      "loss": 0.05332665,
+      "step": 6637
+    },
+    {
+      "epoch": 13.276,
+      "grad_norm": 1.05376398563385,
+      "learning_rate": 2e-05,
+      "loss": 0.03772529,
+      "step": 6638
+    },
+    {
+      "epoch": 13.278,
+      "grad_norm": 1.2965894937515259,
+      "learning_rate": 2e-05,
+      "loss": 0.04613126,
+      "step": 6639
+    },
+    {
+      "epoch": 13.28,
+      "grad_norm": 1.0610442161560059,
+      "learning_rate": 2e-05,
+      "loss": 0.02312069,
+      "step": 6640
+    },
+    {
+      "epoch": 13.282,
+      "grad_norm": 1.1080660820007324,
+      "learning_rate": 2e-05,
+      "loss": 0.03545141,
+      "step": 6641
+    },
+    {
+      "epoch": 13.284,
+      "grad_norm": 2.201037883758545,
+      "learning_rate": 2e-05,
+      "loss": 0.03849367,
+      "step": 6642
+    },
+    {
+      "epoch": 13.286,
+      "grad_norm": 1.9222381114959717,
+      "learning_rate": 2e-05,
+      "loss": 0.0556365,
+      "step": 6643
+    },
+    {
+      "epoch": 13.288,
+      "grad_norm": 1.5982733964920044,
+      "learning_rate": 2e-05,
+      "loss": 0.03892791,
+      "step": 6644
+    },
+    {
+      "epoch": 13.29,
+      "grad_norm": 1.0715982913970947,
+      "learning_rate": 2e-05,
+      "loss": 0.02838083,
+      "step": 6645
+    },
+    {
+      "epoch": 13.292,
+      "grad_norm": 1.6198405027389526,
+      "learning_rate": 2e-05,
+      "loss": 0.05888985,
+      "step": 6646
+    },
+    {
+      "epoch": 13.294,
+      "grad_norm": 1.8655623197555542,
+      "learning_rate": 2e-05,
+      "loss": 0.0356196,
+      "step": 6647
+    },
+    {
+      "epoch": 13.296,
+      "grad_norm": 1.737970232963562,
+      "learning_rate": 2e-05,
+      "loss": 0.02180437,
+      "step": 6648
+    },
+    {
+      "epoch": 13.298,
+      "grad_norm": 0.9337108135223389,
+      "learning_rate": 2e-05,
+      "loss": 0.02707865,
+      "step": 6649
+    },
+    {
+      "epoch": 13.3,
+      "grad_norm": 0.9037566781044006,
+      "learning_rate": 2e-05,
+      "loss": 0.02494653,
+      "step": 6650
+    },
+    {
+      "epoch": 13.302,
+      "grad_norm": 0.9231191277503967,
+      "learning_rate": 2e-05,
+      "loss": 0.03142556,
+      "step": 6651
+    },
+    {
+      "epoch": 13.304,
+      "grad_norm": 3.1863584518432617,
+      "learning_rate": 2e-05,
+      "loss": 0.05755781,
+      "step": 6652
+    },
+    {
+      "epoch": 13.306,
+      "grad_norm": 0.9532139897346497,
+      "learning_rate": 2e-05,
+      "loss": 0.03102466,
+      "step": 6653
+    },
+    {
+      "epoch": 13.308,
+      "grad_norm": 1.7882435321807861,
+      "learning_rate": 2e-05,
+      "loss": 0.04434324,
+      "step": 6654
+    },
+    {
+      "epoch": 13.31,
+      "grad_norm": 1.4374666213989258,
+      "learning_rate": 2e-05,
+      "loss": 0.04442374,
+      "step": 6655
+    },
+    {
+      "epoch": 13.312,
+      "grad_norm": 1.6483453512191772,
+      "learning_rate": 2e-05,
+      "loss": 0.04483134,
+      "step": 6656
+    },
+    {
+      "epoch": 13.314,
+      "grad_norm": 1.6020697355270386,
+      "learning_rate": 2e-05,
+      "loss": 0.03660455,
+      "step": 6657
+    },
+    {
+      "epoch": 13.316,
+      "grad_norm": 2.8730368614196777,
+      "learning_rate": 2e-05,
+      "loss": 0.04435084,
+      "step": 6658
+    },
+    {
+      "epoch": 13.318,
+      "grad_norm": 1.9684791564941406,
+      "learning_rate": 2e-05,
+      "loss": 0.04246651,
+      "step": 6659
+    },
+    {
+      "epoch": 13.32,
+      "grad_norm": 1.097242832183838,
+      "learning_rate": 2e-05,
+      "loss": 0.03018874,
+      "step": 6660
+    },
+    {
+      "epoch": 13.322,
+      "grad_norm": 0.9127419590950012,
+      "learning_rate": 2e-05,
+      "loss": 0.03180817,
+      "step": 6661
+    },
+    {
+      "epoch": 13.324,
+      "grad_norm": 1.5429625511169434,
+      "learning_rate": 2e-05,
+      "loss": 0.05303571,
+      "step": 6662
+    },
+    {
+      "epoch": 13.326,
+      "grad_norm": 1.4380091428756714,
+      "learning_rate": 2e-05,
+      "loss": 0.05016965,
+      "step": 6663
+    },
+    {
+      "epoch": 13.328,
+      "grad_norm": 1.411498785018921,
+      "learning_rate": 2e-05,
+      "loss": 0.05673229,
+      "step": 6664
+    },
+    {
+      "epoch": 13.33,
+      "grad_norm": 1.6350773572921753,
+      "learning_rate": 2e-05,
+      "loss": 0.04110835,
+      "step": 6665
+    },
+    {
+      "epoch": 13.332,
+      "grad_norm": 1.2463691234588623,
+      "learning_rate": 2e-05,
+      "loss": 0.05402338,
+      "step": 6666
+    },
+    {
+      "epoch": 13.334,
+      "grad_norm": 2.4059319496154785,
+      "learning_rate": 2e-05,
+      "loss": 0.044544,
+      "step": 6667
+    },
+    {
+      "epoch": 13.336,
+      "grad_norm": 1.8783429861068726,
+      "learning_rate": 2e-05,
+      "loss": 0.04299849,
+      "step": 6668
+    },
+    {
+      "epoch": 13.338,
+      "grad_norm": 2.927933692932129,
+      "learning_rate": 2e-05,
+      "loss": 0.04565507,
+      "step": 6669
+    },
+    {
+      "epoch": 13.34,
+      "grad_norm": 1.7480547428131104,
+      "learning_rate": 2e-05,
+      "loss": 0.03766192,
+      "step": 6670
+    },
+    {
+      "epoch": 13.342,
+      "grad_norm": 2.074599266052246,
+      "learning_rate": 2e-05,
+      "loss": 0.04872315,
+      "step": 6671
+    },
+    {
+      "epoch": 13.344,
+      "grad_norm": 2.1086673736572266,
+      "learning_rate": 2e-05,
+      "loss": 0.06374644,
+      "step": 6672
+    },
+    {
+      "epoch": 13.346,
+      "grad_norm": 2.058213949203491,
+      "learning_rate": 2e-05,
+      "loss": 0.03495604,
+      "step": 6673
+    },
+    {
+      "epoch": 13.348,
+      "grad_norm": 1.0207751989364624,
+      "learning_rate": 2e-05,
+      "loss": 0.03142983,
+      "step": 6674
+    },
+    {
+      "epoch": 13.35,
+      "grad_norm": 2.3387722969055176,
+      "learning_rate": 2e-05,
+      "loss": 0.04296496,
+      "step": 6675
+    },
+    {
+      "epoch": 13.352,
+      "grad_norm": 1.1725760698318481,
+      "learning_rate": 2e-05,
+      "loss": 0.03329622,
+      "step": 6676
+    },
+    {
+      "epoch": 13.354,
+      "grad_norm": 1.225332498550415,
+      "learning_rate": 2e-05,
+      "loss": 0.03967672,
+      "step": 6677
+    },
+    {
+      "epoch": 13.356,
+      "grad_norm": 1.1926147937774658,
+      "learning_rate": 2e-05,
+      "loss": 0.03046719,
+      "step": 6678
+    },
+    {
+      "epoch": 13.358,
+      "grad_norm": 1.0180567502975464,
+      "learning_rate": 2e-05,
+      "loss": 0.03923386,
+      "step": 6679
+    },
+    {
+      "epoch": 13.36,
+      "grad_norm": 1.3657896518707275,
+      "learning_rate": 2e-05,
+      "loss": 0.03993597,
+      "step": 6680
+    },
+    {
+      "epoch": 13.362,
+      "grad_norm": 1.367834448814392,
+      "learning_rate": 2e-05,
+      "loss": 0.0454083,
+      "step": 6681
+    },
+    {
+      "epoch": 13.364,
+      "grad_norm": 1.1526918411254883,
+      "learning_rate": 2e-05,
+      "loss": 0.03554086,
+      "step": 6682
+    },
+    {
+      "epoch": 13.366,
+      "grad_norm": 1.4416218996047974,
+      "learning_rate": 2e-05,
+      "loss": 0.04877956,
+      "step": 6683
+    },
+    {
+      "epoch": 13.368,
+      "grad_norm": 1.735988736152649,
+      "learning_rate": 2e-05,
+      "loss": 0.0392142,
+      "step": 6684
+    },
+    {
+      "epoch": 13.37,
+      "grad_norm": 1.4720538854599,
+      "learning_rate": 2e-05,
+      "loss": 0.04314131,
+      "step": 6685
+    },
+    {
+      "epoch": 13.372,
+      "grad_norm": 1.0763517618179321,
+      "learning_rate": 2e-05,
+      "loss": 0.03109873,
+      "step": 6686
+    },
+    {
+      "epoch": 13.374,
+      "grad_norm": 1.1306136846542358,
+      "learning_rate": 2e-05,
+      "loss": 0.04201383,
+      "step": 6687
+    },
+    {
+      "epoch": 13.376,
+      "grad_norm": 1.4741770029067993,
+      "learning_rate": 2e-05,
+      "loss": 0.04738778,
+      "step": 6688
+    },
+    {
+      "epoch": 13.378,
+      "grad_norm": 1.487601399421692,
+      "learning_rate": 2e-05,
+      "loss": 0.03376631,
+      "step": 6689
+    },
+    {
+      "epoch": 13.38,
+      "grad_norm": 1.2999999523162842,
+      "learning_rate": 2e-05,
+      "loss": 0.04101618,
+      "step": 6690
+    },
+    {
+      "epoch": 13.382,
+      "grad_norm": 1.1709288358688354,
+      "learning_rate": 2e-05,
+      "loss": 0.03829162,
+      "step": 6691
+    },
+    {
+      "epoch": 13.384,
+      "grad_norm": 1.452234148979187,
+      "learning_rate": 2e-05,
+      "loss": 0.03547079,
+      "step": 6692
+    },
+    {
+      "epoch": 13.386,
+      "grad_norm": 0.8371902108192444,
+      "learning_rate": 2e-05,
+      "loss": 0.02435498,
+      "step": 6693
+    },
+    {
+      "epoch": 13.388,
+      "grad_norm": 1.7057963609695435,
+      "learning_rate": 2e-05,
+      "loss": 0.0380133,
+      "step": 6694
+    },
+    {
+      "epoch": 13.39,
+      "grad_norm": 1.3966196775436401,
+      "learning_rate": 2e-05,
+      "loss": 0.04224955,
+      "step": 6695
+    },
+    {
+      "epoch": 13.392,
+      "grad_norm": 1.1923191547393799,
+      "learning_rate": 2e-05,
+      "loss": 0.03705193,
+      "step": 6696
+    },
+    {
+      "epoch": 13.394,
+      "grad_norm": 0.9897952079772949,
+      "learning_rate": 2e-05,
+      "loss": 0.03031838,
+      "step": 6697
+    },
+    {
+      "epoch": 13.396,
+      "grad_norm": 1.9399019479751587,
+      "learning_rate": 2e-05,
+      "loss": 0.05891667,
+      "step": 6698
+    },
+    {
+      "epoch": 13.398,
+      "grad_norm": 1.0432417392730713,
+      "learning_rate": 2e-05,
+      "loss": 0.03055418,
+      "step": 6699
+    },
+    {
+      "epoch": 13.4,
+      "grad_norm": 0.9006956219673157,
+      "learning_rate": 2e-05,
+      "loss": 0.02497819,
+      "step": 6700
+    },
+    {
+      "epoch": 13.402,
+      "grad_norm": 1.248022198677063,
+      "learning_rate": 2e-05,
+      "loss": 0.02955998,
+      "step": 6701
+    },
+    {
+      "epoch": 13.404,
+      "grad_norm": 1.5496799945831299,
+      "learning_rate": 2e-05,
+      "loss": 0.06074677,
+      "step": 6702
+    },
+    {
+      "epoch": 13.406,
+      "grad_norm": 1.4711843729019165,
+      "learning_rate": 2e-05,
+      "loss": 0.04381853,
+      "step": 6703
+    },
+    {
+      "epoch": 13.408,
+      "grad_norm": 1.325208067893982,
+      "learning_rate": 2e-05,
+      "loss": 0.04054024,
+      "step": 6704
+    },
+    {
+      "epoch": 13.41,
+      "grad_norm": 1.5213847160339355,
+      "learning_rate": 2e-05,
+      "loss": 0.03479857,
+      "step": 6705
+    },
+    {
+      "epoch": 13.412,
+      "grad_norm": 1.5465024709701538,
+      "learning_rate": 2e-05,
+      "loss": 0.03222397,
+      "step": 6706
+    },
+    {
+      "epoch": 13.414,
+      "grad_norm": 1.4514657258987427,
+      "learning_rate": 2e-05,
+      "loss": 0.03864145,
+      "step": 6707
+    },
+    {
+      "epoch": 13.416,
+      "grad_norm": 1.5519639253616333,
+      "learning_rate": 2e-05,
+      "loss": 0.04974742,
+      "step": 6708
+    },
+    {
+      "epoch": 13.418,
+      "grad_norm": 1.543533205986023,
+      "learning_rate": 2e-05,
+      "loss": 0.04505913,
+      "step": 6709
+    },
+    {
+      "epoch": 13.42,
+      "grad_norm": 1.244042158126831,
+      "learning_rate": 2e-05,
+      "loss": 0.03692468,
+      "step": 6710
+    },
+    {
+      "epoch": 13.422,
+      "grad_norm": 1.8627500534057617,
+      "learning_rate": 2e-05,
+      "loss": 0.04253852,
+      "step": 6711
+    },
+    {
+      "epoch": 13.424,
+      "grad_norm": 1.4597536325454712,
+      "learning_rate": 2e-05,
+      "loss": 0.03926862,
+      "step": 6712
+    },
+    {
+      "epoch": 13.426,
+      "grad_norm": 1.841884732246399,
+      "learning_rate": 2e-05,
+      "loss": 0.0296302,
+      "step": 6713
+    },
+    {
+      "epoch": 13.428,
+      "grad_norm": 2.594111204147339,
+      "learning_rate": 2e-05,
+      "loss": 0.05455346,
+      "step": 6714
+    },
+    {
+      "epoch": 13.43,
+      "grad_norm": 1.466094732284546,
+      "learning_rate": 2e-05,
+      "loss": 0.03086076,
+      "step": 6715
+    },
+    {
+      "epoch": 13.432,
+      "grad_norm": 2.093371868133545,
+      "learning_rate": 2e-05,
+      "loss": 0.05538841,
+      "step": 6716
+    },
+    {
+      "epoch": 13.434,
+      "grad_norm": 1.1023695468902588,
+      "learning_rate": 2e-05,
+      "loss": 0.04121345,
+      "step": 6717
+    },
+    {
+      "epoch": 13.436,
+      "grad_norm": 2.032099485397339,
+      "learning_rate": 2e-05,
+      "loss": 0.05646689,
+      "step": 6718
+    },
+    {
+      "epoch": 13.438,
+      "grad_norm": 1.150732398033142,
+      "learning_rate": 2e-05,
+      "loss": 0.03391449,
+      "step": 6719
+    },
+    {
+      "epoch": 13.44,
+      "grad_norm": 2.225139617919922,
+      "learning_rate": 2e-05,
+      "loss": 0.05393933,
+      "step": 6720
+    },
+    {
+      "epoch": 13.442,
+      "grad_norm": 3.746210813522339,
+      "learning_rate": 2e-05,
+      "loss": 0.04924157,
+      "step": 6721
+    },
+    {
+      "epoch": 13.444,
+      "grad_norm": 1.8030920028686523,
+      "learning_rate": 2e-05,
+      "loss": 0.04365423,
+      "step": 6722
+    },
+    {
+      "epoch": 13.446,
+      "grad_norm": 0.9142201542854309,
+      "learning_rate": 2e-05,
+      "loss": 0.02320842,
+      "step": 6723
+    },
+    {
+      "epoch": 13.448,
+      "grad_norm": 1.0032247304916382,
+      "learning_rate": 2e-05,
+      "loss": 0.03920649,
+      "step": 6724
+    },
+    {
+      "epoch": 13.45,
+      "grad_norm": 1.1446279287338257,
+      "learning_rate": 2e-05,
+      "loss": 0.0443483,
+      "step": 6725
+    },
+    {
+      "epoch": 13.452,
+      "grad_norm": 1.371544599533081,
+      "learning_rate": 2e-05,
+      "loss": 0.05964202,
+      "step": 6726
+    },
+    {
+      "epoch": 13.454,
+      "grad_norm": 1.4129619598388672,
+      "learning_rate": 2e-05,
+      "loss": 0.0317092,
+      "step": 6727
+    },
+    {
+      "epoch": 13.456,
+      "grad_norm": 2.262296438217163,
+      "learning_rate": 2e-05,
+      "loss": 0.04459076,
+      "step": 6728
+    },
+    {
+      "epoch": 13.458,
+      "grad_norm": 1.2286860942840576,
+      "learning_rate": 2e-05,
+      "loss": 0.04396465,
+      "step": 6729
+    },
+    {
+      "epoch": 13.46,
+      "grad_norm": 1.2002842426300049,
+      "learning_rate": 2e-05,
+      "loss": 0.02897931,
+      "step": 6730
+    },
+    {
+      "epoch": 13.462,
+      "grad_norm": 1.1000972986221313,
+      "learning_rate": 2e-05,
+      "loss": 0.04501866,
+      "step": 6731
+    },
+    {
+      "epoch": 13.464,
+      "grad_norm": 1.1960804462432861,
+      "learning_rate": 2e-05,
+      "loss": 0.02948781,
+      "step": 6732
+    },
+    {
+      "epoch": 13.466,
+      "grad_norm": 1.0015963315963745,
+      "learning_rate": 2e-05,
+      "loss": 0.03117497,
+      "step": 6733
+    },
+    {
+      "epoch": 13.468,
+      "grad_norm": 1.2673149108886719,
+      "learning_rate": 2e-05,
+      "loss": 0.03281594,
+      "step": 6734
+    },
+    {
+      "epoch": 13.47,
+      "grad_norm": 1.048366665840149,
+      "learning_rate": 2e-05,
+      "loss": 0.02636184,
+      "step": 6735
+    },
+    {
+      "epoch": 13.472,
+      "grad_norm": 2.1534323692321777,
+      "learning_rate": 2e-05,
+      "loss": 0.06946588,
+      "step": 6736
+    },
+    {
+      "epoch": 13.474,
+      "grad_norm": 1.2543253898620605,
+      "learning_rate": 2e-05,
+      "loss": 0.03551532,
+      "step": 6737
+    },
+    {
+      "epoch": 13.475999999999999,
+      "grad_norm": 2.5918846130371094,
+      "learning_rate": 2e-05,
+      "loss": 0.04691193,
+      "step": 6738
+    },
+    {
+      "epoch": 13.478,
+      "grad_norm": 0.7969970107078552,
+      "learning_rate": 2e-05,
+      "loss": 0.01839239,
+      "step": 6739
+    },
+    {
+      "epoch": 13.48,
+      "grad_norm": 2.0504300594329834,
+      "learning_rate": 2e-05,
+      "loss": 0.06359866,
+      "step": 6740
+    },
+    {
+      "epoch": 13.482,
+      "grad_norm": 0.8780750036239624,
+      "learning_rate": 2e-05,
+      "loss": 0.0225176,
+      "step": 6741
+    },
+    {
+      "epoch": 13.484,
+      "grad_norm": 1.1083356142044067,
+      "learning_rate": 2e-05,
+      "loss": 0.0324852,
+      "step": 6742
+    },
+    {
+      "epoch": 13.486,
+      "grad_norm": 2.280796527862549,
+      "learning_rate": 2e-05,
+      "loss": 0.0467263,
+      "step": 6743
+    },
+    {
+      "epoch": 13.488,
+      "grad_norm": 1.2050200700759888,
+      "learning_rate": 2e-05,
+      "loss": 0.02860709,
+      "step": 6744
+    },
+    {
+      "epoch": 13.49,
+      "grad_norm": 1.8593145608901978,
+      "learning_rate": 2e-05,
+      "loss": 0.03531153,
+      "step": 6745
+    },
+    {
+      "epoch": 13.492,
+      "grad_norm": 1.798421025276184,
+      "learning_rate": 2e-05,
+      "loss": 0.0394549,
+      "step": 6746
+    },
+    {
+      "epoch": 13.494,
+      "grad_norm": 1.5639235973358154,
+      "learning_rate": 2e-05,
+      "loss": 0.05283806,
+      "step": 6747
+    },
+    {
+      "epoch": 13.496,
+      "grad_norm": 1.026511549949646,
+      "learning_rate": 2e-05,
+      "loss": 0.02420056,
+      "step": 6748
+    },
+    {
+      "epoch": 13.498,
+      "grad_norm": 2.5854506492614746,
+      "learning_rate": 2e-05,
+      "loss": 0.06519131,
+      "step": 6749
+    },
+    {
+      "epoch": 13.5,
+      "grad_norm": 1.8973256349563599,
+      "learning_rate": 2e-05,
+      "loss": 0.04225761,
+      "step": 6750
+    },
+    {
+      "epoch": 13.502,
+      "grad_norm": 2.533794403076172,
+      "learning_rate": 2e-05,
+      "loss": 0.02219105,
+      "step": 6751
+    },
+    {
+      "epoch": 13.504,
+      "grad_norm": 1.8227717876434326,
+      "learning_rate": 2e-05,
+      "loss": 0.04808221,
+      "step": 6752
+    },
+    {
+      "epoch": 13.506,
+      "grad_norm": 1.0413390398025513,
+      "learning_rate": 2e-05,
+      "loss": 0.03295096,
+      "step": 6753
+    },
+    {
+      "epoch": 13.508,
+      "grad_norm": 1.9891551733016968,
+      "learning_rate": 2e-05,
+      "loss": 0.03309003,
+      "step": 6754
+    },
+    {
+      "epoch": 13.51,
+      "grad_norm": 1.173137903213501,
+      "learning_rate": 2e-05,
+      "loss": 0.03191693,
+      "step": 6755
+    },
+    {
+      "epoch": 13.512,
+      "grad_norm": 1.9114857912063599,
+      "learning_rate": 2e-05,
+      "loss": 0.05413975,
+      "step": 6756
+    },
+    {
+      "epoch": 13.514,
+      "grad_norm": 1.889272689819336,
+      "learning_rate": 2e-05,
+      "loss": 0.0451472,
+      "step": 6757
+    },
+    {
+      "epoch": 13.516,
+      "grad_norm": 2.4712700843811035,
+      "learning_rate": 2e-05,
+      "loss": 0.07723338,
+      "step": 6758
+    },
+    {
+      "epoch": 13.518,
+      "grad_norm": 2.2369396686553955,
+      "learning_rate": 2e-05,
+      "loss": 0.03549297,
+      "step": 6759
+    },
+    {
+      "epoch": 13.52,
+      "grad_norm": 2.845893144607544,
+      "learning_rate": 2e-05,
+      "loss": 0.05792838,
+      "step": 6760
+    },
+    {
+      "epoch": 13.522,
+      "grad_norm": 1.513317346572876,
+      "learning_rate": 2e-05,
+      "loss": 0.04458487,
+      "step": 6761
+    },
+    {
+      "epoch": 13.524000000000001,
+      "grad_norm": 1.106540560722351,
+      "learning_rate": 2e-05,
+      "loss": 0.03790012,
+      "step": 6762
+    },
+    {
+      "epoch": 13.526,
+      "grad_norm": 1.4870750904083252,
+      "learning_rate": 2e-05,
+      "loss": 0.02564323,
+      "step": 6763
+    },
+    {
+      "epoch": 13.528,
+      "grad_norm": 1.8651138544082642,
+      "learning_rate": 2e-05,
+      "loss": 0.05272502,
+      "step": 6764
+    },
+    {
+      "epoch": 13.53,
+      "grad_norm": 1.0633503198623657,
+      "learning_rate": 2e-05,
+      "loss": 0.04007112,
+      "step": 6765
+    },
+    {
+      "epoch": 13.532,
+      "grad_norm": 1.0299681425094604,
+      "learning_rate": 2e-05,
+      "loss": 0.03321412,
+      "step": 6766
+    },
+    {
+      "epoch": 13.534,
+      "grad_norm": 2.307370185852051,
+      "learning_rate": 2e-05,
+      "loss": 0.04456577,
+      "step": 6767
+    },
+    {
+      "epoch": 13.536,
+      "grad_norm": 0.9137228727340698,
+      "learning_rate": 2e-05,
+      "loss": 0.02563262,
+      "step": 6768
+    },
+    {
+      "epoch": 13.538,
+      "grad_norm": 1.7050751447677612,
+      "learning_rate": 2e-05,
+      "loss": 0.05230839,
+      "step": 6769
+    },
+    {
+      "epoch": 13.54,
+      "grad_norm": 1.070886492729187,
+      "learning_rate": 2e-05,
+      "loss": 0.02357959,
+      "step": 6770
+    },
+    {
+      "epoch": 13.542,
+      "grad_norm": 1.695765495300293,
+      "learning_rate": 2e-05,
+      "loss": 0.04813949,
+      "step": 6771
+    },
+    {
+      "epoch": 13.544,
+      "grad_norm": 1.1727943420410156,
+      "learning_rate": 2e-05,
+      "loss": 0.04054714,
+      "step": 6772
+    },
+    {
+      "epoch": 13.546,
+      "grad_norm": 0.9893969297409058,
+      "learning_rate": 2e-05,
+      "loss": 0.02479752,
+      "step": 6773
+    },
+    {
+      "epoch": 13.548,
+      "grad_norm": 1.2786779403686523,
+      "learning_rate": 2e-05,
+      "loss": 0.04855724,
+      "step": 6774
+    },
+    {
+      "epoch": 13.55,
+      "grad_norm": 2.4117813110351562,
+      "learning_rate": 2e-05,
+      "loss": 0.06347778,
+      "step": 6775
+    },
+    {
+      "epoch": 13.552,
+      "grad_norm": 1.0595612525939941,
+      "learning_rate": 2e-05,
+      "loss": 0.04373883,
+      "step": 6776
+    },
+    {
+      "epoch": 13.554,
+      "grad_norm": 1.1950441598892212,
+      "learning_rate": 2e-05,
+      "loss": 0.04355627,
+      "step": 6777
+    },
+    {
+      "epoch": 13.556000000000001,
+      "grad_norm": 1.147063136100769,
+      "learning_rate": 2e-05,
+      "loss": 0.03419853,
+      "step": 6778
+    },
+    {
+      "epoch": 13.558,
+      "grad_norm": 1.4784750938415527,
+      "learning_rate": 2e-05,
+      "loss": 0.04642799,
+      "step": 6779
+    },
+    {
+      "epoch": 13.56,
+      "grad_norm": 1.2722127437591553,
+      "learning_rate": 2e-05,
+      "loss": 0.03892024,
+      "step": 6780
+    },
+    {
+      "epoch": 13.562,
+      "grad_norm": 1.6152026653289795,
+      "learning_rate": 2e-05,
+      "loss": 0.04028412,
+      "step": 6781
+    },
+    {
+      "epoch": 13.564,
+      "grad_norm": 1.1551095247268677,
+      "learning_rate": 2e-05,
+      "loss": 0.03943279,
+      "step": 6782
+    },
+    {
+      "epoch": 13.566,
+      "grad_norm": 2.295308828353882,
+      "learning_rate": 2e-05,
+      "loss": 0.054483,
+      "step": 6783
+    },
+    {
+      "epoch": 13.568,
+      "grad_norm": 1.0823659896850586,
+      "learning_rate": 2e-05,
+      "loss": 0.0385061,
+      "step": 6784
+    },
+    {
+      "epoch": 13.57,
+      "grad_norm": 1.4659216403961182,
+      "learning_rate": 2e-05,
+      "loss": 0.0353567,
+      "step": 6785
+    },
+    {
+      "epoch": 13.572,
+      "grad_norm": 1.425752878189087,
+      "learning_rate": 2e-05,
+      "loss": 0.03669987,
+      "step": 6786
+    },
+    {
+      "epoch": 13.574,
+      "grad_norm": 2.6012632846832275,
+      "learning_rate": 2e-05,
+      "loss": 0.04362625,
+      "step": 6787
+    },
+    {
+      "epoch": 13.576,
+      "grad_norm": 1.162837266921997,
+      "learning_rate": 2e-05,
+      "loss": 0.03585674,
+      "step": 6788
+    },
+    {
+      "epoch": 13.578,
+      "grad_norm": 1.107102394104004,
+      "learning_rate": 2e-05,
+      "loss": 0.03959781,
+      "step": 6789
+    },
+    {
+      "epoch": 13.58,
+      "grad_norm": 1.6900997161865234,
+      "learning_rate": 2e-05,
+      "loss": 0.06017102,
+      "step": 6790
+    },
+    {
+      "epoch": 13.582,
+      "grad_norm": 1.4423255920410156,
+      "learning_rate": 2e-05,
+      "loss": 0.04416816,
+      "step": 6791
+    },
+    {
+      "epoch": 13.584,
+      "grad_norm": 1.5928566455841064,
+      "learning_rate": 2e-05,
+      "loss": 0.03386574,
+      "step": 6792
+    },
+    {
+      "epoch": 13.586,
+      "grad_norm": 1.8104335069656372,
+      "learning_rate": 2e-05,
+      "loss": 0.0400222,
+      "step": 6793
+    },
+    {
+      "epoch": 13.588,
+      "grad_norm": 1.451529860496521,
+      "learning_rate": 2e-05,
+      "loss": 0.04577955,
+      "step": 6794
+    },
+    {
+      "epoch": 13.59,
+      "grad_norm": 1.7088810205459595,
+      "learning_rate": 2e-05,
+      "loss": 0.03847235,
+      "step": 6795
+    },
+    {
+      "epoch": 13.592,
+      "grad_norm": 1.926962971687317,
+      "learning_rate": 2e-05,
+      "loss": 0.03429279,
+      "step": 6796
+    },
+    {
+      "epoch": 13.594,
+      "grad_norm": 2.025388479232788,
+      "learning_rate": 2e-05,
+      "loss": 0.04720273,
+      "step": 6797
+    },
+    {
+      "epoch": 13.596,
+      "grad_norm": 1.016129732131958,
+      "learning_rate": 2e-05,
+      "loss": 0.03385385,
+      "step": 6798
+    },
+    {
+      "epoch": 13.598,
+      "grad_norm": 1.0439811944961548,
+      "learning_rate": 2e-05,
+      "loss": 0.02241655,
+      "step": 6799
+    },
+    {
+      "epoch": 13.6,
+      "grad_norm": 0.7543059587478638,
+      "learning_rate": 2e-05,
+      "loss": 0.01822535,
+      "step": 6800
+    },
+    {
+      "epoch": 13.602,
+      "grad_norm": 1.9524015188217163,
+      "learning_rate": 2e-05,
+      "loss": 0.04379565,
+      "step": 6801
+    },
+    {
+      "epoch": 13.604,
+      "grad_norm": 1.0359333753585815,
+      "learning_rate": 2e-05,
+      "loss": 0.0270812,
+      "step": 6802
+    },
+    {
+      "epoch": 13.606,
+      "grad_norm": 1.362512469291687,
+      "learning_rate": 2e-05,
+      "loss": 0.03527067,
+      "step": 6803
+    },
+    {
+      "epoch": 13.608,
+      "grad_norm": 2.679394245147705,
+      "learning_rate": 2e-05,
+      "loss": 0.06183048,
+      "step": 6804
+    },
+    {
+      "epoch": 13.61,
+      "grad_norm": 1.175911784172058,
+      "learning_rate": 2e-05,
+      "loss": 0.03599912,
+      "step": 6805
+    },
+    {
+      "epoch": 13.612,
+      "grad_norm": 2.0313289165496826,
+      "learning_rate": 2e-05,
+      "loss": 0.06214502,
+      "step": 6806
+    },
+    {
+      "epoch": 13.614,
+      "grad_norm": 1.8251034021377563,
+      "learning_rate": 2e-05,
+      "loss": 0.04615673,
+      "step": 6807
+    },
+    {
+      "epoch": 13.616,
+      "grad_norm": 1.214990258216858,
+      "learning_rate": 2e-05,
+      "loss": 0.03257399,
+      "step": 6808
+    },
+    {
+      "epoch": 13.618,
+      "grad_norm": 3.3506529331207275,
+      "learning_rate": 2e-05,
+      "loss": 0.04304178,
+      "step": 6809
+    },
+    {
+      "epoch": 13.62,
+      "grad_norm": 1.2393721342086792,
+      "learning_rate": 2e-05,
+      "loss": 0.04934491,
+      "step": 6810
+    },
+    {
+      "epoch": 13.622,
+      "grad_norm": 2.080261468887329,
+      "learning_rate": 2e-05,
+      "loss": 0.04631698,
+      "step": 6811
+    },
+    {
+      "epoch": 13.624,
+      "grad_norm": 1.2759827375411987,
+      "learning_rate": 2e-05,
+      "loss": 0.03095847,
+      "step": 6812
+    },
+    {
+      "epoch": 13.626,
+      "grad_norm": 1.7272143363952637,
+      "learning_rate": 2e-05,
+      "loss": 0.03796447,
+      "step": 6813
+    },
+    {
+      "epoch": 13.628,
+      "grad_norm": 1.4961469173431396,
+      "learning_rate": 2e-05,
+      "loss": 0.04623634,
+      "step": 6814
+    },
+    {
+      "epoch": 13.63,
+      "grad_norm": 1.548699140548706,
+      "learning_rate": 2e-05,
+      "loss": 0.0516596,
+      "step": 6815
+    },
+    {
+      "epoch": 13.632,
+      "grad_norm": 1.285744309425354,
+      "learning_rate": 2e-05,
+      "loss": 0.03057302,
+      "step": 6816
+    },
+    {
+      "epoch": 13.634,
+      "grad_norm": 1.525428295135498,
+      "learning_rate": 2e-05,
+      "loss": 0.05115666,
+      "step": 6817
+    },
+    {
+      "epoch": 13.636,
+      "grad_norm": 1.7536441087722778,
+      "learning_rate": 2e-05,
+      "loss": 0.04631253,
+      "step": 6818
+    },
+    {
+      "epoch": 13.638,
+      "grad_norm": 1.1306259632110596,
+      "learning_rate": 2e-05,
+      "loss": 0.04088272,
+      "step": 6819
+    },
+    {
+      "epoch": 13.64,
+      "grad_norm": 1.1280131340026855,
+      "learning_rate": 2e-05,
+      "loss": 0.04373218,
+      "step": 6820
+    },
+    {
+      "epoch": 13.642,
+      "grad_norm": 3.139237642288208,
+      "learning_rate": 2e-05,
+      "loss": 0.04641888,
+      "step": 6821
+    },
+    {
+      "epoch": 13.644,
+      "grad_norm": 1.6292080879211426,
+      "learning_rate": 2e-05,
+      "loss": 0.04293904,
+      "step": 6822
+    },
+    {
+      "epoch": 13.646,
+      "grad_norm": 1.7208157777786255,
+      "learning_rate": 2e-05,
+      "loss": 0.04404939,
+      "step": 6823
+    },
+    {
+      "epoch": 13.648,
+      "grad_norm": 1.2277165651321411,
+      "learning_rate": 2e-05,
+      "loss": 0.03341801,
+      "step": 6824
+    },
+    {
+      "epoch": 13.65,
+      "grad_norm": 1.2899742126464844,
+      "learning_rate": 2e-05,
+      "loss": 0.03240266,
+      "step": 6825
+    },
+    {
+      "epoch": 13.652,
+      "grad_norm": 1.4660921096801758,
+      "learning_rate": 2e-05,
+      "loss": 0.03453447,
+      "step": 6826
+    },
+    {
+      "epoch": 13.654,
+      "grad_norm": 1.2889078855514526,
+      "learning_rate": 2e-05,
+      "loss": 0.03558958,
+      "step": 6827
+    },
+    {
+      "epoch": 13.656,
+      "grad_norm": 1.38223397731781,
+      "learning_rate": 2e-05,
+      "loss": 0.05393769,
+      "step": 6828
+    },
+    {
+      "epoch": 13.658,
+      "grad_norm": 1.3276984691619873,
+      "learning_rate": 2e-05,
+      "loss": 0.03997407,
+      "step": 6829
+    },
+    {
+      "epoch": 13.66,
+      "grad_norm": 1.929145097732544,
+      "learning_rate": 2e-05,
+      "loss": 0.0430852,
+      "step": 6830
+    },
+    {
+      "epoch": 13.662,
+      "grad_norm": 1.0870046615600586,
+      "learning_rate": 2e-05,
+      "loss": 0.03382558,
+      "step": 6831
+    },
+    {
+      "epoch": 13.664,
+      "grad_norm": 1.1914732456207275,
+      "learning_rate": 2e-05,
+      "loss": 0.0535154,
+      "step": 6832
+    },
+    {
+      "epoch": 13.666,
+      "grad_norm": 2.2056117057800293,
+      "learning_rate": 2e-05,
+      "loss": 0.05700465,
+      "step": 6833
+    },
+    {
+      "epoch": 13.668,
+      "grad_norm": 1.1149606704711914,
+      "learning_rate": 2e-05,
+      "loss": 0.03322265,
+      "step": 6834
+    },
+    {
+      "epoch": 13.67,
+      "grad_norm": 2.3128271102905273,
+      "learning_rate": 2e-05,
+      "loss": 0.07779928,
+      "step": 6835
+    },
+    {
+      "epoch": 13.672,
+      "grad_norm": 1.0935770273208618,
+      "learning_rate": 2e-05,
+      "loss": 0.03507974,
+      "step": 6836
+    },
+    {
+      "epoch": 13.674,
+      "grad_norm": 1.0744456052780151,
+      "learning_rate": 2e-05,
+      "loss": 0.04397395,
+      "step": 6837
+    },
+    {
+      "epoch": 13.676,
+      "grad_norm": 1.3249156475067139,
+      "learning_rate": 2e-05,
+      "loss": 0.04180006,
+      "step": 6838
+    },
+    {
+      "epoch": 13.678,
+      "grad_norm": 1.1021473407745361,
+      "learning_rate": 2e-05,
+      "loss": 0.03645816,
+      "step": 6839
+    },
+    {
+      "epoch": 13.68,
+      "grad_norm": 1.929029941558838,
+      "learning_rate": 2e-05,
+      "loss": 0.02640364,
+      "step": 6840
+    },
+    {
+      "epoch": 13.682,
+      "grad_norm": 0.9546433687210083,
+      "learning_rate": 2e-05,
+      "loss": 0.02705884,
+      "step": 6841
+    },
+    {
+      "epoch": 13.684,
+      "grad_norm": 1.073384165763855,
+      "learning_rate": 2e-05,
+      "loss": 0.04102583,
+      "step": 6842
+    },
+    {
+      "epoch": 13.686,
+      "grad_norm": 1.257896900177002,
+      "learning_rate": 2e-05,
+      "loss": 0.04304846,
+      "step": 6843
+    },
+    {
+      "epoch": 13.688,
+      "grad_norm": 1.318122148513794,
+      "learning_rate": 2e-05,
+      "loss": 0.04488578,
+      "step": 6844
+    },
+    {
+      "epoch": 13.69,
+      "grad_norm": 0.9590141773223877,
+      "learning_rate": 2e-05,
+      "loss": 0.02748418,
+      "step": 6845
+    },
+    {
+      "epoch": 13.692,
+      "grad_norm": 1.7860455513000488,
+      "learning_rate": 2e-05,
+      "loss": 0.05183525,
+      "step": 6846
+    },
+    {
+      "epoch": 13.693999999999999,
+      "grad_norm": 0.9630289673805237,
+      "learning_rate": 2e-05,
+      "loss": 0.03061388,
+      "step": 6847
+    },
+    {
+      "epoch": 13.696,
+      "grad_norm": 1.2916054725646973,
+      "learning_rate": 2e-05,
+      "loss": 0.03421018,
+      "step": 6848
+    },
+    {
+      "epoch": 13.698,
+      "grad_norm": 2.0008859634399414,
+      "learning_rate": 2e-05,
+      "loss": 0.04936755,
+      "step": 6849
+    },
+    {
+      "epoch": 13.7,
+      "grad_norm": 2.6435530185699463,
+      "learning_rate": 2e-05,
+      "loss": 0.06780395,
+      "step": 6850
+    },
+    {
+      "epoch": 13.702,
+      "grad_norm": 0.9989877939224243,
+      "learning_rate": 2e-05,
+      "loss": 0.03243184,
+      "step": 6851
+    },
+    {
+      "epoch": 13.704,
+      "grad_norm": 1.6317846775054932,
+      "learning_rate": 2e-05,
+      "loss": 0.0594992,
+      "step": 6852
+    },
+    {
+      "epoch": 13.706,
+      "grad_norm": 1.2833094596862793,
+      "learning_rate": 2e-05,
+      "loss": 0.03055147,
+      "step": 6853
+    },
+    {
+      "epoch": 13.708,
+      "grad_norm": 0.9552610516548157,
+      "learning_rate": 2e-05,
+      "loss": 0.0297171,
+      "step": 6854
+    },
+    {
+      "epoch": 13.71,
+      "grad_norm": 1.4244675636291504,
+      "learning_rate": 2e-05,
+      "loss": 0.04444023,
+      "step": 6855
+    },
+    {
+      "epoch": 13.712,
+      "grad_norm": 1.2428985834121704,
+      "learning_rate": 2e-05,
+      "loss": 0.04104583,
+      "step": 6856
+    },
+    {
+      "epoch": 13.714,
+      "grad_norm": 1.4677211046218872,
+      "learning_rate": 2e-05,
+      "loss": 0.05508178,
+      "step": 6857
+    },
+    {
+      "epoch": 13.716,
+      "grad_norm": 1.8807919025421143,
+      "learning_rate": 2e-05,
+      "loss": 0.0591247,
+      "step": 6858
+    },
+    {
+      "epoch": 13.718,
+      "grad_norm": 1.9581363201141357,
+      "learning_rate": 2e-05,
+      "loss": 0.04779023,
+      "step": 6859
+    },
+    {
+      "epoch": 13.72,
+      "grad_norm": 2.0055696964263916,
+      "learning_rate": 2e-05,
+      "loss": 0.0408286,
+      "step": 6860
+    },
+    {
+      "epoch": 13.722,
+      "grad_norm": 1.830039143562317,
+      "learning_rate": 2e-05,
+      "loss": 0.04613936,
+      "step": 6861
+    },
+    {
+      "epoch": 13.724,
+      "grad_norm": 2.8258206844329834,
+      "learning_rate": 2e-05,
+      "loss": 0.04065733,
+      "step": 6862
+    },
+    {
+      "epoch": 13.725999999999999,
+      "grad_norm": 1.871401309967041,
+      "learning_rate": 2e-05,
+      "loss": 0.04201019,
+      "step": 6863
+    },
+    {
+      "epoch": 13.728,
+      "grad_norm": 1.449409008026123,
+      "learning_rate": 2e-05,
+      "loss": 0.03992969,
+      "step": 6864
+    },
+    {
+      "epoch": 13.73,
+      "grad_norm": 0.9711210131645203,
+      "learning_rate": 2e-05,
+      "loss": 0.0266035,
+      "step": 6865
+    },
+    {
+      "epoch": 13.732,
+      "grad_norm": 3.511932849884033,
+      "learning_rate": 2e-05,
+      "loss": 0.04639168,
+      "step": 6866
+    },
+    {
+      "epoch": 13.734,
+      "grad_norm": 1.1094722747802734,
+      "learning_rate": 2e-05,
+      "loss": 0.0350898,
+      "step": 6867
+    },
+    {
+      "epoch": 13.736,
+      "grad_norm": 1.4972585439682007,
+      "learning_rate": 2e-05,
+      "loss": 0.04061377,
+      "step": 6868
+    },
+    {
+      "epoch": 13.738,
+      "grad_norm": 1.1617933511734009,
+      "learning_rate": 2e-05,
+      "loss": 0.0347375,
+      "step": 6869
+    },
+    {
+      "epoch": 13.74,
+      "grad_norm": 1.5653934478759766,
+      "learning_rate": 2e-05,
+      "loss": 0.04277583,
+      "step": 6870
+    },
+    {
+      "epoch": 13.742,
+      "grad_norm": 1.7562743425369263,
+      "learning_rate": 2e-05,
+      "loss": 0.04624029,
+      "step": 6871
+    },
+    {
+      "epoch": 13.744,
+      "grad_norm": 1.6248326301574707,
+      "learning_rate": 2e-05,
+      "loss": 0.05260355,
+      "step": 6872
+    },
+    {
+      "epoch": 13.746,
+      "grad_norm": 1.7137155532836914,
+      "learning_rate": 2e-05,
+      "loss": 0.05001482,
+      "step": 6873
+    },
+    {
+      "epoch": 13.748,
+      "grad_norm": 1.3740524053573608,
+      "learning_rate": 2e-05,
+      "loss": 0.04304221,
+      "step": 6874
+    },
+    {
+      "epoch": 13.75,
+      "grad_norm": 1.070404291152954,
+      "learning_rate": 2e-05,
+      "loss": 0.02320381,
+      "step": 6875
+    },
+    {
+      "epoch": 13.752,
+      "grad_norm": 2.6879231929779053,
+      "learning_rate": 2e-05,
+      "loss": 0.04556799,
+      "step": 6876
+    },
+    {
+      "epoch": 13.754,
+      "grad_norm": 2.933372974395752,
+      "learning_rate": 2e-05,
+      "loss": 0.04154959,
+      "step": 6877
+    },
+    {
+      "epoch": 13.756,
+      "grad_norm": 1.3636802434921265,
+      "learning_rate": 2e-05,
+      "loss": 0.03800491,
+      "step": 6878
+    },
+    {
+      "epoch": 13.758,
+      "grad_norm": 1.157073736190796,
+      "learning_rate": 2e-05,
+      "loss": 0.03626024,
+      "step": 6879
+    },
+    {
+      "epoch": 13.76,
+      "grad_norm": 1.7253233194351196,
+      "learning_rate": 2e-05,
+      "loss": 0.04370718,
+      "step": 6880
+    },
+    {
+      "epoch": 13.762,
+      "grad_norm": 1.3973212242126465,
+      "learning_rate": 2e-05,
+      "loss": 0.03621276,
+      "step": 6881
+    },
+    {
+      "epoch": 13.764,
+      "grad_norm": 1.7019752264022827,
+      "learning_rate": 2e-05,
+      "loss": 0.03458362,
+      "step": 6882
+    },
+    {
+      "epoch": 13.766,
+      "grad_norm": 1.227394700050354,
+      "learning_rate": 2e-05,
+      "loss": 0.04752816,
+      "step": 6883
+    },
+    {
+      "epoch": 13.768,
+      "grad_norm": 1.493351697921753,
+      "learning_rate": 2e-05,
+      "loss": 0.03362212,
+      "step": 6884
+    },
+    {
+      "epoch": 13.77,
+      "grad_norm": 1.694974422454834,
+      "learning_rate": 2e-05,
+      "loss": 0.06347952,
+      "step": 6885
+    },
+    {
+      "epoch": 13.772,
+      "grad_norm": 1.4916173219680786,
+      "learning_rate": 2e-05,
+      "loss": 0.04393812,
+      "step": 6886
+    },
+    {
+      "epoch": 13.774000000000001,
+      "grad_norm": 1.1646591424942017,
+      "learning_rate": 2e-05,
+      "loss": 0.03405294,
+      "step": 6887
+    },
+    {
+      "epoch": 13.776,
+      "grad_norm": 1.1886578798294067,
+      "learning_rate": 2e-05,
+      "loss": 0.02821852,
+      "step": 6888
+    },
+    {
+      "epoch": 13.778,
+      "grad_norm": 1.2024027109146118,
+      "learning_rate": 2e-05,
+      "loss": 0.0455913,
+      "step": 6889
+    },
+    {
+      "epoch": 13.78,
+      "grad_norm": 1.684903621673584,
+      "learning_rate": 2e-05,
+      "loss": 0.04784643,
+      "step": 6890
+    },
+    {
+      "epoch": 13.782,
+      "grad_norm": 0.9186260104179382,
+      "learning_rate": 2e-05,
+      "loss": 0.02574374,
+      "step": 6891
+    },
+    {
+      "epoch": 13.784,
+      "grad_norm": 1.2807894945144653,
+      "learning_rate": 2e-05,
+      "loss": 0.03100124,
+      "step": 6892
+    },
+    {
+      "epoch": 13.786,
+      "grad_norm": 2.1691181659698486,
+      "learning_rate": 2e-05,
+      "loss": 0.02056332,
+      "step": 6893
+    },
+    {
+      "epoch": 13.788,
+      "grad_norm": 1.8777505159378052,
+      "learning_rate": 2e-05,
+      "loss": 0.04661623,
+      "step": 6894
+    },
+    {
+      "epoch": 13.79,
+      "grad_norm": 1.4597963094711304,
+      "learning_rate": 2e-05,
+      "loss": 0.05256153,
+      "step": 6895
+    },
+    {
+      "epoch": 13.792,
+      "grad_norm": 1.8397506475448608,
+      "learning_rate": 2e-05,
+      "loss": 0.05605321,
+      "step": 6896
+    },
+    {
+      "epoch": 13.794,
+      "grad_norm": 1.1106833219528198,
+      "learning_rate": 2e-05,
+      "loss": 0.03157957,
+      "step": 6897
+    },
+    {
+      "epoch": 13.796,
+      "grad_norm": 1.118843674659729,
+      "learning_rate": 2e-05,
+      "loss": 0.0361166,
+      "step": 6898
+    },
+    {
+      "epoch": 13.798,
+      "grad_norm": 1.4842638969421387,
+      "learning_rate": 2e-05,
+      "loss": 0.04007118,
+      "step": 6899
+    },
+    {
+      "epoch": 13.8,
+      "grad_norm": 1.4503836631774902,
+      "learning_rate": 2e-05,
+      "loss": 0.04612395,
+      "step": 6900
+    },
+    {
+      "epoch": 13.802,
+      "grad_norm": 1.2816123962402344,
+      "learning_rate": 2e-05,
+      "loss": 0.04101133,
+      "step": 6901
+    },
+    {
+      "epoch": 13.804,
+      "grad_norm": 1.273935317993164,
+      "learning_rate": 2e-05,
+      "loss": 0.03615731,
+      "step": 6902
+    },
+    {
+      "epoch": 13.806000000000001,
+      "grad_norm": 1.308467984199524,
+      "learning_rate": 2e-05,
+      "loss": 0.04421499,
+      "step": 6903
+    },
+    {
+      "epoch": 13.808,
+      "grad_norm": 1.472285509109497,
+      "learning_rate": 2e-05,
+      "loss": 0.04577631,
+      "step": 6904
+    },
+    {
+      "epoch": 13.81,
+      "grad_norm": 1.998026728630066,
+      "learning_rate": 2e-05,
+      "loss": 0.0379182,
+      "step": 6905
+    },
+    {
+      "epoch": 13.812,
+      "grad_norm": 1.1190993785858154,
+      "learning_rate": 2e-05,
+      "loss": 0.03913017,
+      "step": 6906
+    },
+    {
+      "epoch": 13.814,
+      "grad_norm": 3.0139517784118652,
+      "learning_rate": 2e-05,
+      "loss": 0.0456089,
+      "step": 6907
+    },
+    {
+      "epoch": 13.816,
+      "grad_norm": 1.1194919347763062,
+      "learning_rate": 2e-05,
+      "loss": 0.04423294,
+      "step": 6908
+    },
+    {
+      "epoch": 13.818,
+      "grad_norm": 1.2570995092391968,
+      "learning_rate": 2e-05,
+      "loss": 0.03813727,
+      "step": 6909
+    },
+    {
+      "epoch": 13.82,
+      "grad_norm": 1.317740797996521,
+      "learning_rate": 2e-05,
+      "loss": 0.04127891,
+      "step": 6910
+    },
+    {
+      "epoch": 13.822,
+      "grad_norm": 1.1890889406204224,
+      "learning_rate": 2e-05,
+      "loss": 0.0370289,
+      "step": 6911
+    },
+    {
+      "epoch": 13.824,
+      "grad_norm": 3.053663730621338,
+      "learning_rate": 2e-05,
+      "loss": 0.06167036,
+      "step": 6912
+    },
+    {
+      "epoch": 13.826,
+      "grad_norm": 1.2635729312896729,
+      "learning_rate": 2e-05,
+      "loss": 0.05142916,
+      "step": 6913
+    },
+    {
+      "epoch": 13.828,
+      "grad_norm": 2.6170880794525146,
+      "learning_rate": 2e-05,
+      "loss": 0.0427838,
+      "step": 6914
+    },
+    {
+      "epoch": 13.83,
+      "grad_norm": 1.7016270160675049,
+      "learning_rate": 2e-05,
+      "loss": 0.02756583,
+      "step": 6915
+    },
+    {
+      "epoch": 13.832,
+      "grad_norm": 1.1118059158325195,
+      "learning_rate": 2e-05,
+      "loss": 0.04283454,
+      "step": 6916
+    },
+    {
+      "epoch": 13.834,
+      "grad_norm": 1.7809730768203735,
+      "learning_rate": 2e-05,
+      "loss": 0.04107636,
+      "step": 6917
+    },
+    {
+      "epoch": 13.836,
+      "grad_norm": 1.776057243347168,
+      "learning_rate": 2e-05,
+      "loss": 0.06522802,
+      "step": 6918
+    },
+    {
+      "epoch": 13.838,
+      "grad_norm": 2.162116765975952,
+      "learning_rate": 2e-05,
+      "loss": 0.0393892,
+      "step": 6919
+    },
+    {
+      "epoch": 13.84,
+      "grad_norm": 1.280896544456482,
+      "learning_rate": 2e-05,
+      "loss": 0.05017384,
+      "step": 6920
+    },
+    {
+      "epoch": 13.842,
+      "grad_norm": 1.4492366313934326,
+      "learning_rate": 2e-05,
+      "loss": 0.04526861,
+      "step": 6921
+    },
+    {
+      "epoch": 13.844,
+      "grad_norm": 1.8244743347167969,
+      "learning_rate": 2e-05,
+      "loss": 0.06141206,
+      "step": 6922
+    },
+    {
+      "epoch": 13.846,
+      "grad_norm": 1.2145180702209473,
+      "learning_rate": 2e-05,
+      "loss": 0.03708153,
+      "step": 6923
+    },
+    {
+      "epoch": 13.848,
+      "grad_norm": 1.1694141626358032,
+      "learning_rate": 2e-05,
+      "loss": 0.03655352,
+      "step": 6924
+    },
+    {
+      "epoch": 13.85,
+      "grad_norm": 2.419154644012451,
+      "learning_rate": 2e-05,
+      "loss": 0.06464878,
+      "step": 6925
+    },
+    {
+      "epoch": 13.852,
+      "grad_norm": 1.4023338556289673,
+      "learning_rate": 2e-05,
+      "loss": 0.03684681,
+      "step": 6926
+    },
+    {
+      "epoch": 13.854,
+      "grad_norm": 1.3352280855178833,
+      "learning_rate": 2e-05,
+      "loss": 0.03953333,
+      "step": 6927
+    },
+    {
+      "epoch": 13.856,
+      "grad_norm": 0.8786143660545349,
+      "learning_rate": 2e-05,
+      "loss": 0.0229125,
+      "step": 6928
+    },
+    {
+      "epoch": 13.858,
+      "grad_norm": 1.3116097450256348,
+      "learning_rate": 2e-05,
+      "loss": 0.04031592,
+      "step": 6929
+    },
+    {
+      "epoch": 13.86,
+      "grad_norm": 1.9184274673461914,
+      "learning_rate": 2e-05,
+      "loss": 0.0427547,
+      "step": 6930
+    },
+    {
+      "epoch": 13.862,
+      "grad_norm": 1.721928596496582,
+      "learning_rate": 2e-05,
+      "loss": 0.06148085,
+      "step": 6931
+    },
+    {
+      "epoch": 13.864,
+      "grad_norm": 1.0398845672607422,
+      "learning_rate": 2e-05,
+      "loss": 0.0353225,
+      "step": 6932
+    },
+    {
+      "epoch": 13.866,
+      "grad_norm": 0.929514467716217,
+      "learning_rate": 2e-05,
+      "loss": 0.02477574,
+      "step": 6933
+    },
+    {
+      "epoch": 13.868,
+      "grad_norm": 1.3540958166122437,
+      "learning_rate": 2e-05,
+      "loss": 0.03898114,
+      "step": 6934
+    },
+    {
+      "epoch": 13.87,
+      "grad_norm": 1.7967790365219116,
+      "learning_rate": 2e-05,
+      "loss": 0.04478693,
+      "step": 6935
+    },
+    {
+      "epoch": 13.872,
+      "grad_norm": 1.5192323923110962,
+      "learning_rate": 2e-05,
+      "loss": 0.04696514,
+      "step": 6936
+    },
+    {
+      "epoch": 13.874,
+      "grad_norm": 1.3892743587493896,
+      "learning_rate": 2e-05,
+      "loss": 0.04166014,
+      "step": 6937
+    },
+    {
+      "epoch": 13.876,
+      "grad_norm": 1.4207967519760132,
+      "learning_rate": 2e-05,
+      "loss": 0.03712838,
+      "step": 6938
+    },
+    {
+      "epoch": 13.878,
+      "grad_norm": 1.5932114124298096,
+      "learning_rate": 2e-05,
+      "loss": 0.03432607,
+      "step": 6939
+    },
+    {
+      "epoch": 13.88,
+      "grad_norm": 1.9691510200500488,
+      "learning_rate": 2e-05,
+      "loss": 0.03547703,
+      "step": 6940
+    },
+    {
+      "epoch": 13.882,
+      "grad_norm": 1.0378472805023193,
+      "learning_rate": 2e-05,
+      "loss": 0.02458699,
+      "step": 6941
+    },
+    {
+      "epoch": 13.884,
+      "grad_norm": 2.503615140914917,
+      "learning_rate": 2e-05,
+      "loss": 0.04526497,
+      "step": 6942
+    },
+    {
+      "epoch": 13.886,
+      "grad_norm": 1.3788117170333862,
+      "learning_rate": 2e-05,
+      "loss": 0.04195864,
+      "step": 6943
+    },
+    {
+      "epoch": 13.888,
+      "grad_norm": 1.6336097717285156,
+      "learning_rate": 2e-05,
+      "loss": 0.03892026,
+      "step": 6944
+    },
+    {
+      "epoch": 13.89,
+      "grad_norm": 1.337511420249939,
+      "learning_rate": 2e-05,
+      "loss": 0.03862485,
+      "step": 6945
+    },
+    {
+      "epoch": 13.892,
+      "grad_norm": 1.2629504203796387,
+      "learning_rate": 2e-05,
+      "loss": 0.03670635,
+      "step": 6946
+    },
+    {
+      "epoch": 13.894,
+      "grad_norm": 1.307521104812622,
+      "learning_rate": 2e-05,
+      "loss": 0.04121678,
+      "step": 6947
+    },
+    {
+      "epoch": 13.896,
+      "grad_norm": 1.2313232421875,
+      "learning_rate": 2e-05,
+      "loss": 0.02896846,
+      "step": 6948
+    },
+    {
+      "epoch": 13.898,
+      "grad_norm": 1.50730299949646,
+      "learning_rate": 2e-05,
+      "loss": 0.04983806,
+      "step": 6949
+    },
+    {
+      "epoch": 13.9,
+      "grad_norm": 1.2840025424957275,
+      "learning_rate": 2e-05,
+      "loss": 0.03766822,
+      "step": 6950
+    },
+    {
+      "epoch": 13.902,
+      "grad_norm": 1.5729337930679321,
+      "learning_rate": 2e-05,
+      "loss": 0.04883525,
+      "step": 6951
+    },
+    {
+      "epoch": 13.904,
+      "grad_norm": 1.3678689002990723,
+      "learning_rate": 2e-05,
+      "loss": 0.02985792,
+      "step": 6952
+    },
+    {
+      "epoch": 13.906,
+      "grad_norm": 1.845995545387268,
+      "learning_rate": 2e-05,
+      "loss": 0.03586386,
+      "step": 6953
+    },
+    {
+      "epoch": 13.908,
+      "grad_norm": 1.7878332138061523,
+      "learning_rate": 2e-05,
+      "loss": 0.03654428,
+      "step": 6954
+    },
+    {
+      "epoch": 13.91,
+      "grad_norm": 3.085512161254883,
+      "learning_rate": 2e-05,
+      "loss": 0.05120496,
+      "step": 6955
+    },
+    {
+      "epoch": 13.912,
+      "grad_norm": 3.0082015991210938,
+      "learning_rate": 2e-05,
+      "loss": 0.05480274,
+      "step": 6956
+    },
+    {
+      "epoch": 13.914,
+      "grad_norm": 1.2339752912521362,
+      "learning_rate": 2e-05,
+      "loss": 0.03606438,
+      "step": 6957
+    },
+    {
+      "epoch": 13.916,
+      "grad_norm": 1.9148188829421997,
+      "learning_rate": 2e-05,
+      "loss": 0.04195002,
+      "step": 6958
+    },
+    {
+      "epoch": 13.918,
+      "grad_norm": 1.1235597133636475,
+      "learning_rate": 2e-05,
+      "loss": 0.0217904,
+      "step": 6959
+    },
+    {
+      "epoch": 13.92,
+      "grad_norm": 2.466542959213257,
+      "learning_rate": 2e-05,
+      "loss": 0.06078681,
+      "step": 6960
+    },
+    {
+      "epoch": 13.922,
+      "grad_norm": 1.385965347290039,
+      "learning_rate": 2e-05,
+      "loss": 0.02699665,
+      "step": 6961
+    },
+    {
+      "epoch": 13.924,
+      "grad_norm": 1.464664340019226,
+      "learning_rate": 2e-05,
+      "loss": 0.03695335,
+      "step": 6962
+    },
+    {
+      "epoch": 13.926,
+      "grad_norm": 1.523497462272644,
+      "learning_rate": 2e-05,
+      "loss": 0.04739627,
+      "step": 6963
+    },
+    {
+      "epoch": 13.928,
+      "grad_norm": 2.7052083015441895,
+      "learning_rate": 2e-05,
+      "loss": 0.05714166,
+      "step": 6964
+    },
+    {
+      "epoch": 13.93,
+      "grad_norm": 1.4660812616348267,
+      "learning_rate": 2e-05,
+      "loss": 0.03287406,
+      "step": 6965
+    },
+    {
+      "epoch": 13.932,
+      "grad_norm": 1.166961669921875,
+      "learning_rate": 2e-05,
+      "loss": 0.03073209,
+      "step": 6966
+    },
+    {
+      "epoch": 13.934,
+      "grad_norm": 2.519378423690796,
+      "learning_rate": 2e-05,
+      "loss": 0.06768637,
+      "step": 6967
+    },
+    {
+      "epoch": 13.936,
+      "grad_norm": 1.424590826034546,
+      "learning_rate": 2e-05,
+      "loss": 0.03274088,
+      "step": 6968
+    },
+    {
+      "epoch": 13.938,
+      "grad_norm": 1.7873649597167969,
+      "learning_rate": 2e-05,
+      "loss": 0.0431348,
+      "step": 6969
+    },
+    {
+      "epoch": 13.94,
+      "grad_norm": 1.6397364139556885,
+      "learning_rate": 2e-05,
+      "loss": 0.05216152,
+      "step": 6970
+    },
+    {
+      "epoch": 13.942,
+      "grad_norm": 1.9869099855422974,
+      "learning_rate": 2e-05,
+      "loss": 0.042454,
+      "step": 6971
+    },
+    {
+      "epoch": 13.943999999999999,
+      "grad_norm": 1.582697868347168,
+      "learning_rate": 2e-05,
+      "loss": 0.04447088,
+      "step": 6972
+    },
+    {
+      "epoch": 13.946,
+      "grad_norm": 1.1062511205673218,
+      "learning_rate": 2e-05,
+      "loss": 0.03642577,
+      "step": 6973
+    },
+    {
+      "epoch": 13.948,
+      "grad_norm": 1.2900327444076538,
+      "learning_rate": 2e-05,
+      "loss": 0.04951657,
+      "step": 6974
+    },
+    {
+      "epoch": 13.95,
+      "grad_norm": 1.1730434894561768,
+      "learning_rate": 2e-05,
+      "loss": 0.03196269,
+      "step": 6975
+    },
+    {
+      "epoch": 13.952,
+      "grad_norm": 1.5335745811462402,
+      "learning_rate": 2e-05,
+      "loss": 0.05588776,
+      "step": 6976
+    },
+    {
+      "epoch": 13.954,
+      "grad_norm": 2.73171329498291,
+      "learning_rate": 2e-05,
+      "loss": 0.0453019,
+      "step": 6977
+    },
+    {
+      "epoch": 13.956,
+      "grad_norm": 1.3756558895111084,
+      "learning_rate": 2e-05,
+      "loss": 0.04042321,
+      "step": 6978
+    },
+    {
+      "epoch": 13.958,
+      "grad_norm": 1.025495171546936,
+      "learning_rate": 2e-05,
+      "loss": 0.03254474,
+      "step": 6979
+    },
+    {
+      "epoch": 13.96,
+      "grad_norm": 0.8832976818084717,
+      "learning_rate": 2e-05,
+      "loss": 0.0262373,
+      "step": 6980
+    },
+    {
+      "epoch": 13.962,
+      "grad_norm": 1.3270471096038818,
+      "learning_rate": 2e-05,
+      "loss": 0.03066928,
+      "step": 6981
+    },
+    {
+      "epoch": 13.964,
+      "grad_norm": 1.8798532485961914,
+      "learning_rate": 2e-05,
+      "loss": 0.03866819,
+      "step": 6982
+    },
+    {
+      "epoch": 13.966,
+      "grad_norm": 1.69099760055542,
+      "learning_rate": 2e-05,
+      "loss": 0.05059724,
+      "step": 6983
+    },
+    {
+      "epoch": 13.968,
+      "grad_norm": 1.9903326034545898,
+      "learning_rate": 2e-05,
+      "loss": 0.05271498,
+      "step": 6984
+    },
+    {
+      "epoch": 13.97,
+      "grad_norm": 1.4454317092895508,
+      "learning_rate": 2e-05,
+      "loss": 0.03909637,
+      "step": 6985
+    },
+    {
+      "epoch": 13.972,
+      "grad_norm": 1.4461326599121094,
+      "learning_rate": 2e-05,
+      "loss": 0.04804809,
+      "step": 6986
+    },
+    {
+      "epoch": 13.974,
+      "grad_norm": 1.59605872631073,
+      "learning_rate": 2e-05,
+      "loss": 0.04185,
+      "step": 6987
+    },
+    {
+      "epoch": 13.975999999999999,
+      "grad_norm": 1.0969065427780151,
+      "learning_rate": 2e-05,
+      "loss": 0.03130201,
+      "step": 6988
+    },
+    {
+      "epoch": 13.978,
+      "grad_norm": 1.317733645439148,
+      "learning_rate": 2e-05,
+      "loss": 0.03786597,
+      "step": 6989
+    },
+    {
+      "epoch": 13.98,
+      "grad_norm": 1.1903131008148193,
+      "learning_rate": 2e-05,
+      "loss": 0.03440776,
+      "step": 6990
+    },
+    {
+      "epoch": 13.982,
+      "grad_norm": 1.5716062784194946,
+      "learning_rate": 2e-05,
+      "loss": 0.03851765,
+      "step": 6991
+    },
+    {
+      "epoch": 13.984,
+      "grad_norm": 1.9664435386657715,
+      "learning_rate": 2e-05,
+      "loss": 0.05616831,
+      "step": 6992
+    },
+    {
+      "epoch": 13.986,
+      "grad_norm": 2.0219123363494873,
+      "learning_rate": 2e-05,
+      "loss": 0.04097972,
+      "step": 6993
+    },
+    {
+      "epoch": 13.988,
+      "grad_norm": 1.4964755773544312,
+      "learning_rate": 2e-05,
+      "loss": 0.05085189,
+      "step": 6994
+    },
+    {
+      "epoch": 13.99,
+      "grad_norm": 1.520508885383606,
+      "learning_rate": 2e-05,
+      "loss": 0.03831444,
+      "step": 6995
+    },
+    {
+      "epoch": 13.992,
+      "grad_norm": 1.3518579006195068,
+      "learning_rate": 2e-05,
+      "loss": 0.0365081,
+      "step": 6996
+    },
+    {
+      "epoch": 13.994,
+      "grad_norm": 2.745910882949829,
+      "learning_rate": 2e-05,
+      "loss": 0.03853264,
+      "step": 6997
+    },
+    {
+      "epoch": 13.996,
+      "grad_norm": 1.2294520139694214,
+      "learning_rate": 2e-05,
+      "loss": 0.0548241,
+      "step": 6998
+    },
+    {
+      "epoch": 13.998,
+      "grad_norm": 1.276281714439392,
+      "learning_rate": 2e-05,
+      "loss": 0.04351184,
+      "step": 6999
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 1.2956624031066895,
+      "learning_rate": 2e-05,
+      "loss": 0.04201851,
+      "step": 7000
+    },
+    {
+      "epoch": 14.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.99,
+        "AngleClassification_2": 0.998,
+        "AngleClassification_3": 0.9600798403193613,
+        "Equal_1": 0.996,
+        "Equal_2": 0.9720558882235529,
+        "Equal_3": 0.9001996007984032,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9900199600798403,
+        "Parallel_1": 0.9939879759519038,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 0.994,
+        "Perpendicular_2": 0.976,
+        "Perpendicular_3": 0.7324649298597194,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 0.9976666666666667,
+        "PointLiesOnCircle_3": 0.9916,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9859719438877755,
+        "PointLiesOnLine_3": 0.9720558882235529
+      },
+      "eval_runtime": 318.9857,
+      "eval_samples_per_second": 32.917,
+      "eval_steps_per_second": 0.658,
+      "step": 7000
+    },
+    {
+      "epoch": 14.002,
+      "grad_norm": 1.2843459844589233,
+      "learning_rate": 2e-05,
+      "loss": 0.03288523,
+      "step": 7001
+    },
+    {
+      "epoch": 14.004,
+      "grad_norm": 1.4108099937438965,
+      "learning_rate": 2e-05,
+      "loss": 0.04203776,
+      "step": 7002
+    },
+    {
+      "epoch": 14.006,
+      "grad_norm": 1.3905504941940308,
+      "learning_rate": 2e-05,
+      "loss": 0.03418431,
+      "step": 7003
+    },
+    {
+      "epoch": 14.008,
+      "grad_norm": 1.3126283884048462,
+      "learning_rate": 2e-05,
+      "loss": 0.03478189,
+      "step": 7004
+    },
+    {
+      "epoch": 14.01,
+      "grad_norm": 1.513098120689392,
+      "learning_rate": 2e-05,
+      "loss": 0.02606986,
+      "step": 7005
+    },
+    {
+      "epoch": 14.012,
+      "grad_norm": 0.7904066443443298,
+      "learning_rate": 2e-05,
+      "loss": 0.01787382,
+      "step": 7006
+    },
+    {
+      "epoch": 14.014,
+      "grad_norm": 3.054617166519165,
+      "learning_rate": 2e-05,
+      "loss": 0.04634954,
+      "step": 7007
+    },
+    {
+      "epoch": 14.016,
+      "grad_norm": 1.0060993432998657,
+      "learning_rate": 2e-05,
+      "loss": 0.03051211,
+      "step": 7008
+    },
+    {
+      "epoch": 14.018,
+      "grad_norm": 1.1262900829315186,
+      "learning_rate": 2e-05,
+      "loss": 0.04555785,
+      "step": 7009
+    },
+    {
+      "epoch": 14.02,
+      "grad_norm": 1.0693005323410034,
+      "learning_rate": 2e-05,
+      "loss": 0.02980948,
+      "step": 7010
+    },
+    {
+      "epoch": 14.022,
+      "grad_norm": 2.0782830715179443,
+      "learning_rate": 2e-05,
+      "loss": 0.04862558,
+      "step": 7011
+    },
+    {
+      "epoch": 14.024,
+      "grad_norm": 1.079984188079834,
+      "learning_rate": 2e-05,
+      "loss": 0.03432882,
+      "step": 7012
+    },
+    {
+      "epoch": 14.026,
+      "grad_norm": 1.2360132932662964,
+      "learning_rate": 2e-05,
+      "loss": 0.02996394,
+      "step": 7013
+    },
+    {
+      "epoch": 14.028,
+      "grad_norm": 1.3257410526275635,
+      "learning_rate": 2e-05,
+      "loss": 0.04203,
+      "step": 7014
+    },
+    {
+      "epoch": 14.03,
+      "grad_norm": 1.318507194519043,
+      "learning_rate": 2e-05,
+      "loss": 0.04230007,
+      "step": 7015
+    },
+    {
+      "epoch": 14.032,
+      "grad_norm": 1.3276084661483765,
+      "learning_rate": 2e-05,
+      "loss": 0.03894487,
+      "step": 7016
+    },
+    {
+      "epoch": 14.034,
+      "grad_norm": 3.99560809135437,
+      "learning_rate": 2e-05,
+      "loss": 0.04019061,
+      "step": 7017
+    },
+    {
+      "epoch": 14.036,
+      "grad_norm": 0.97989422082901,
+      "learning_rate": 2e-05,
+      "loss": 0.03579887,
+      "step": 7018
+    },
+    {
+      "epoch": 14.038,
+      "grad_norm": 3.5206971168518066,
+      "learning_rate": 2e-05,
+      "loss": 0.04952036,
+      "step": 7019
+    },
+    {
+      "epoch": 14.04,
+      "grad_norm": 1.5767422914505005,
+      "learning_rate": 2e-05,
+      "loss": 0.04102598,
+      "step": 7020
+    },
+    {
+      "epoch": 14.042,
+      "grad_norm": 2.001638412475586,
+      "learning_rate": 2e-05,
+      "loss": 0.04028576,
+      "step": 7021
+    },
+    {
+      "epoch": 14.044,
+      "grad_norm": 1.0773245096206665,
+      "learning_rate": 2e-05,
+      "loss": 0.04221611,
+      "step": 7022
+    },
+    {
+      "epoch": 14.046,
+      "grad_norm": 1.02744460105896,
+      "learning_rate": 2e-05,
+      "loss": 0.02617998,
+      "step": 7023
+    },
+    {
+      "epoch": 14.048,
+      "grad_norm": 1.0974265336990356,
+      "learning_rate": 2e-05,
+      "loss": 0.02758575,
+      "step": 7024
+    },
+    {
+      "epoch": 14.05,
+      "grad_norm": 1.1479986906051636,
+      "learning_rate": 2e-05,
+      "loss": 0.03531769,
+      "step": 7025
+    },
+    {
+      "epoch": 14.052,
+      "grad_norm": 1.6043879985809326,
+      "learning_rate": 2e-05,
+      "loss": 0.04604035,
+      "step": 7026
+    },
+    {
+      "epoch": 14.054,
+      "grad_norm": 1.2420005798339844,
+      "learning_rate": 2e-05,
+      "loss": 0.04582789,
+      "step": 7027
+    },
+    {
+      "epoch": 14.056,
+      "grad_norm": 2.2049918174743652,
+      "learning_rate": 2e-05,
+      "loss": 0.05222199,
+      "step": 7028
+    },
+    {
+      "epoch": 14.058,
+      "grad_norm": 2.92496395111084,
+      "learning_rate": 2e-05,
+      "loss": 0.04773784,
+      "step": 7029
+    },
+    {
+      "epoch": 14.06,
+      "grad_norm": 1.3038673400878906,
+      "learning_rate": 2e-05,
+      "loss": 0.0424346,
+      "step": 7030
+    },
+    {
+      "epoch": 14.062,
+      "grad_norm": 1.3948389291763306,
+      "learning_rate": 2e-05,
+      "loss": 0.0277782,
+      "step": 7031
+    },
+    {
+      "epoch": 14.064,
+      "grad_norm": 1.42351233959198,
+      "learning_rate": 2e-05,
+      "loss": 0.04358821,
+      "step": 7032
+    },
+    {
+      "epoch": 14.066,
+      "grad_norm": 1.726454257965088,
+      "learning_rate": 2e-05,
+      "loss": 0.050067,
+      "step": 7033
+    },
+    {
+      "epoch": 14.068,
+      "grad_norm": 1.3726768493652344,
+      "learning_rate": 2e-05,
+      "loss": 0.04407897,
+      "step": 7034
+    },
+    {
+      "epoch": 14.07,
+      "grad_norm": 1.9279018640518188,
+      "learning_rate": 2e-05,
+      "loss": 0.0469468,
+      "step": 7035
+    },
+    {
+      "epoch": 14.072,
+      "grad_norm": 1.3610163927078247,
+      "learning_rate": 2e-05,
+      "loss": 0.0371495,
+      "step": 7036
+    },
+    {
+      "epoch": 14.074,
+      "grad_norm": 2.023139238357544,
+      "learning_rate": 2e-05,
+      "loss": 0.05333347,
+      "step": 7037
+    },
+    {
+      "epoch": 14.076,
+      "grad_norm": 1.167094349861145,
+      "learning_rate": 2e-05,
+      "loss": 0.03193357,
+      "step": 7038
+    },
+    {
+      "epoch": 14.078,
+      "grad_norm": 1.3578746318817139,
+      "learning_rate": 2e-05,
+      "loss": 0.04010125,
+      "step": 7039
+    },
+    {
+      "epoch": 14.08,
+      "grad_norm": 2.4089791774749756,
+      "learning_rate": 2e-05,
+      "loss": 0.05072846,
+      "step": 7040
+    },
+    {
+      "epoch": 14.082,
+      "grad_norm": 1.022179365158081,
+      "learning_rate": 2e-05,
+      "loss": 0.03089194,
+      "step": 7041
+    },
+    {
+      "epoch": 14.084,
+      "grad_norm": 1.3668341636657715,
+      "learning_rate": 2e-05,
+      "loss": 0.03569974,
+      "step": 7042
+    },
+    {
+      "epoch": 14.086,
+      "grad_norm": 1.1925535202026367,
+      "learning_rate": 2e-05,
+      "loss": 0.03366784,
+      "step": 7043
+    },
+    {
+      "epoch": 14.088,
+      "grad_norm": 1.3144326210021973,
+      "learning_rate": 2e-05,
+      "loss": 0.03443348,
+      "step": 7044
+    },
+    {
+      "epoch": 14.09,
+      "grad_norm": 1.794022560119629,
+      "learning_rate": 2e-05,
+      "loss": 0.04681649,
+      "step": 7045
+    },
+    {
+      "epoch": 14.092,
+      "grad_norm": 1.2918729782104492,
+      "learning_rate": 2e-05,
+      "loss": 0.04749175,
+      "step": 7046
+    },
+    {
+      "epoch": 14.094,
+      "grad_norm": 1.18765389919281,
+      "learning_rate": 2e-05,
+      "loss": 0.03726712,
+      "step": 7047
+    },
+    {
+      "epoch": 14.096,
+      "grad_norm": 1.0554254055023193,
+      "learning_rate": 2e-05,
+      "loss": 0.03222162,
+      "step": 7048
+    },
+    {
+      "epoch": 14.098,
+      "grad_norm": 0.9862664937973022,
+      "learning_rate": 2e-05,
+      "loss": 0.02374917,
+      "step": 7049
+    },
+    {
+      "epoch": 14.1,
+      "grad_norm": 1.021124005317688,
+      "learning_rate": 2e-05,
+      "loss": 0.02776304,
+      "step": 7050
+    },
+    {
+      "epoch": 14.102,
+      "grad_norm": 1.7811987400054932,
+      "learning_rate": 2e-05,
+      "loss": 0.04543906,
+      "step": 7051
+    },
+    {
+      "epoch": 14.104,
+      "grad_norm": 1.463122844696045,
+      "learning_rate": 2e-05,
+      "loss": 0.02648041,
+      "step": 7052
+    },
+    {
+      "epoch": 14.106,
+      "grad_norm": 0.7668428421020508,
+      "learning_rate": 2e-05,
+      "loss": 0.02454238,
+      "step": 7053
+    },
+    {
+      "epoch": 14.108,
+      "grad_norm": 1.272748589515686,
+      "learning_rate": 2e-05,
+      "loss": 0.0323597,
+      "step": 7054
+    },
+    {
+      "epoch": 14.11,
+      "grad_norm": 1.0245872735977173,
+      "learning_rate": 2e-05,
+      "loss": 0.03127266,
+      "step": 7055
+    },
+    {
+      "epoch": 14.112,
+      "grad_norm": 1.9159295558929443,
+      "learning_rate": 2e-05,
+      "loss": 0.04461393,
+      "step": 7056
+    },
+    {
+      "epoch": 14.114,
+      "grad_norm": 1.5680136680603027,
+      "learning_rate": 2e-05,
+      "loss": 0.04500045,
+      "step": 7057
+    },
+    {
+      "epoch": 14.116,
+      "grad_norm": 1.5329288244247437,
+      "learning_rate": 2e-05,
+      "loss": 0.03792113,
+      "step": 7058
+    },
+    {
+      "epoch": 14.118,
+      "grad_norm": 1.3381918668746948,
+      "learning_rate": 2e-05,
+      "loss": 0.04293994,
+      "step": 7059
+    },
+    {
+      "epoch": 14.12,
+      "grad_norm": 1.2890726327896118,
+      "learning_rate": 2e-05,
+      "loss": 0.04811954,
+      "step": 7060
+    },
+    {
+      "epoch": 14.122,
+      "grad_norm": 1.3316044807434082,
+      "learning_rate": 2e-05,
+      "loss": 0.04357966,
+      "step": 7061
+    },
+    {
+      "epoch": 14.124,
+      "grad_norm": 1.3575929403305054,
+      "learning_rate": 2e-05,
+      "loss": 0.03795036,
+      "step": 7062
+    },
+    {
+      "epoch": 14.126,
+      "grad_norm": 1.5020532608032227,
+      "learning_rate": 2e-05,
+      "loss": 0.04436047,
+      "step": 7063
+    },
+    {
+      "epoch": 14.128,
+      "grad_norm": 1.5422673225402832,
+      "learning_rate": 2e-05,
+      "loss": 0.04654428,
+      "step": 7064
+    },
+    {
+      "epoch": 14.13,
+      "grad_norm": 2.029569149017334,
+      "learning_rate": 2e-05,
+      "loss": 0.04016326,
+      "step": 7065
+    },
+    {
+      "epoch": 14.132,
+      "grad_norm": 1.814108967781067,
+      "learning_rate": 2e-05,
+      "loss": 0.04536037,
+      "step": 7066
+    },
+    {
+      "epoch": 14.134,
+      "grad_norm": 1.3232728242874146,
+      "learning_rate": 2e-05,
+      "loss": 0.02865727,
+      "step": 7067
+    },
+    {
+      "epoch": 14.136,
+      "grad_norm": 3.152207136154175,
+      "learning_rate": 2e-05,
+      "loss": 0.04350835,
+      "step": 7068
+    },
+    {
+      "epoch": 14.138,
+      "grad_norm": 1.4238955974578857,
+      "learning_rate": 2e-05,
+      "loss": 0.03182974,
+      "step": 7069
+    },
+    {
+      "epoch": 14.14,
+      "grad_norm": 1.6935443878173828,
+      "learning_rate": 2e-05,
+      "loss": 0.05719973,
+      "step": 7070
+    },
+    {
+      "epoch": 14.142,
+      "grad_norm": 1.262613296508789,
+      "learning_rate": 2e-05,
+      "loss": 0.04014334,
+      "step": 7071
+    },
+    {
+      "epoch": 14.144,
+      "grad_norm": 1.5854763984680176,
+      "learning_rate": 2e-05,
+      "loss": 0.05335325,
+      "step": 7072
+    },
+    {
+      "epoch": 14.146,
+      "grad_norm": 0.964347779750824,
+      "learning_rate": 2e-05,
+      "loss": 0.02348195,
+      "step": 7073
+    },
+    {
+      "epoch": 14.148,
+      "grad_norm": 1.284382700920105,
+      "learning_rate": 2e-05,
+      "loss": 0.04065283,
+      "step": 7074
+    },
+    {
+      "epoch": 14.15,
+      "grad_norm": 1.5653077363967896,
+      "learning_rate": 2e-05,
+      "loss": 0.06096688,
+      "step": 7075
+    },
+    {
+      "epoch": 14.152,
+      "grad_norm": 1.2976857423782349,
+      "learning_rate": 2e-05,
+      "loss": 0.04557998,
+      "step": 7076
+    },
+    {
+      "epoch": 14.154,
+      "grad_norm": 1.6498374938964844,
+      "learning_rate": 2e-05,
+      "loss": 0.04869898,
+      "step": 7077
+    },
+    {
+      "epoch": 14.156,
+      "grad_norm": 1.2892625331878662,
+      "learning_rate": 2e-05,
+      "loss": 0.03722683,
+      "step": 7078
+    },
+    {
+      "epoch": 14.158,
+      "grad_norm": 1.2497565746307373,
+      "learning_rate": 2e-05,
+      "loss": 0.04251852,
+      "step": 7079
+    },
+    {
+      "epoch": 14.16,
+      "grad_norm": 1.8106011152267456,
+      "learning_rate": 2e-05,
+      "loss": 0.02925546,
+      "step": 7080
+    },
+    {
+      "epoch": 14.162,
+      "grad_norm": 1.6762701272964478,
+      "learning_rate": 2e-05,
+      "loss": 0.06018007,
+      "step": 7081
+    },
+    {
+      "epoch": 14.164,
+      "grad_norm": 1.2910103797912598,
+      "learning_rate": 2e-05,
+      "loss": 0.04167475,
+      "step": 7082
+    },
+    {
+      "epoch": 14.166,
+      "grad_norm": 1.5248066186904907,
+      "learning_rate": 2e-05,
+      "loss": 0.04084751,
+      "step": 7083
+    },
+    {
+      "epoch": 14.168,
+      "grad_norm": 1.5304296016693115,
+      "learning_rate": 2e-05,
+      "loss": 0.04385792,
+      "step": 7084
+    },
+    {
+      "epoch": 14.17,
+      "grad_norm": 1.8973009586334229,
+      "learning_rate": 2e-05,
+      "loss": 0.04529943,
+      "step": 7085
+    },
+    {
+      "epoch": 14.172,
+      "grad_norm": 1.4018219709396362,
+      "learning_rate": 2e-05,
+      "loss": 0.03316261,
+      "step": 7086
+    },
+    {
+      "epoch": 14.174,
+      "grad_norm": 2.388532876968384,
+      "learning_rate": 2e-05,
+      "loss": 0.04198411,
+      "step": 7087
+    },
+    {
+      "epoch": 14.176,
+      "grad_norm": 1.7253050804138184,
+      "learning_rate": 2e-05,
+      "loss": 0.03549905,
+      "step": 7088
+    },
+    {
+      "epoch": 14.178,
+      "grad_norm": 1.5690933465957642,
+      "learning_rate": 2e-05,
+      "loss": 0.043695,
+      "step": 7089
+    },
+    {
+      "epoch": 14.18,
+      "grad_norm": 1.1658984422683716,
+      "learning_rate": 2e-05,
+      "loss": 0.03999247,
+      "step": 7090
+    },
+    {
+      "epoch": 14.182,
+      "grad_norm": 1.1882442235946655,
+      "learning_rate": 2e-05,
+      "loss": 0.04268787,
+      "step": 7091
+    },
+    {
+      "epoch": 14.184,
+      "grad_norm": 1.894171118736267,
+      "learning_rate": 2e-05,
+      "loss": 0.05581329,
+      "step": 7092
+    },
+    {
+      "epoch": 14.186,
+      "grad_norm": 1.7583495378494263,
+      "learning_rate": 2e-05,
+      "loss": 0.05542745,
+      "step": 7093
+    },
+    {
+      "epoch": 14.188,
+      "grad_norm": 1.0371955633163452,
+      "learning_rate": 2e-05,
+      "loss": 0.03195368,
+      "step": 7094
+    },
+    {
+      "epoch": 14.19,
+      "grad_norm": 1.2478240728378296,
+      "learning_rate": 2e-05,
+      "loss": 0.0411337,
+      "step": 7095
+    },
+    {
+      "epoch": 14.192,
+      "grad_norm": 1.3688163757324219,
+      "learning_rate": 2e-05,
+      "loss": 0.04424831,
+      "step": 7096
+    },
+    {
+      "epoch": 14.194,
+      "grad_norm": 1.3419855833053589,
+      "learning_rate": 2e-05,
+      "loss": 0.03607039,
+      "step": 7097
+    },
+    {
+      "epoch": 14.196,
+      "grad_norm": 1.9281373023986816,
+      "learning_rate": 2e-05,
+      "loss": 0.05275191,
+      "step": 7098
+    },
+    {
+      "epoch": 14.198,
+      "grad_norm": 1.3586539030075073,
+      "learning_rate": 2e-05,
+      "loss": 0.03638598,
+      "step": 7099
+    },
+    {
+      "epoch": 14.2,
+      "grad_norm": 1.104002833366394,
+      "learning_rate": 2e-05,
+      "loss": 0.03526532,
+      "step": 7100
+    },
+    {
+      "epoch": 14.202,
+      "grad_norm": 1.1839067935943604,
+      "learning_rate": 2e-05,
+      "loss": 0.03894392,
+      "step": 7101
+    },
+    {
+      "epoch": 14.204,
+      "grad_norm": 1.7897748947143555,
+      "learning_rate": 2e-05,
+      "loss": 0.0426039,
+      "step": 7102
+    },
+    {
+      "epoch": 14.206,
+      "grad_norm": 0.8262537717819214,
+      "learning_rate": 2e-05,
+      "loss": 0.02584026,
+      "step": 7103
+    },
+    {
+      "epoch": 14.208,
+      "grad_norm": 1.3617100715637207,
+      "learning_rate": 2e-05,
+      "loss": 0.03045266,
+      "step": 7104
+    },
+    {
+      "epoch": 14.21,
+      "grad_norm": 1.8911073207855225,
+      "learning_rate": 2e-05,
+      "loss": 0.04614431,
+      "step": 7105
+    },
+    {
+      "epoch": 14.212,
+      "grad_norm": 2.970533847808838,
+      "learning_rate": 2e-05,
+      "loss": 0.05797109,
+      "step": 7106
+    },
+    {
+      "epoch": 14.214,
+      "grad_norm": 1.425166368484497,
+      "learning_rate": 2e-05,
+      "loss": 0.03592297,
+      "step": 7107
+    },
+    {
+      "epoch": 14.216,
+      "grad_norm": 1.0085773468017578,
+      "learning_rate": 2e-05,
+      "loss": 0.03424015,
+      "step": 7108
+    },
+    {
+      "epoch": 14.218,
+      "grad_norm": 1.4457972049713135,
+      "learning_rate": 2e-05,
+      "loss": 0.0365461,
+      "step": 7109
+    },
+    {
+      "epoch": 14.22,
+      "grad_norm": 1.1994272470474243,
+      "learning_rate": 2e-05,
+      "loss": 0.02875265,
+      "step": 7110
+    },
+    {
+      "epoch": 14.222,
+      "grad_norm": 0.986005425453186,
+      "learning_rate": 2e-05,
+      "loss": 0.03896586,
+      "step": 7111
+    },
+    {
+      "epoch": 14.224,
+      "grad_norm": 1.2924270629882812,
+      "learning_rate": 2e-05,
+      "loss": 0.03565193,
+      "step": 7112
+    },
+    {
+      "epoch": 14.226,
+      "grad_norm": 1.6049319505691528,
+      "learning_rate": 2e-05,
+      "loss": 0.0383412,
+      "step": 7113
+    },
+    {
+      "epoch": 14.228,
+      "grad_norm": 1.1894758939743042,
+      "learning_rate": 2e-05,
+      "loss": 0.03838333,
+      "step": 7114
+    },
+    {
+      "epoch": 14.23,
+      "grad_norm": 2.1524245738983154,
+      "learning_rate": 2e-05,
+      "loss": 0.05140578,
+      "step": 7115
+    },
+    {
+      "epoch": 14.232,
+      "grad_norm": 1.6528891324996948,
+      "learning_rate": 2e-05,
+      "loss": 0.05099215,
+      "step": 7116
+    },
+    {
+      "epoch": 14.234,
+      "grad_norm": 1.2502809762954712,
+      "learning_rate": 2e-05,
+      "loss": 0.04991131,
+      "step": 7117
+    },
+    {
+      "epoch": 14.236,
+      "grad_norm": 2.5262179374694824,
+      "learning_rate": 2e-05,
+      "loss": 0.03753746,
+      "step": 7118
+    },
+    {
+      "epoch": 14.238,
+      "grad_norm": 1.5351922512054443,
+      "learning_rate": 2e-05,
+      "loss": 0.02967945,
+      "step": 7119
+    },
+    {
+      "epoch": 14.24,
+      "grad_norm": 0.9251724481582642,
+      "learning_rate": 2e-05,
+      "loss": 0.02450234,
+      "step": 7120
+    },
+    {
+      "epoch": 14.242,
+      "grad_norm": 1.6973142623901367,
+      "learning_rate": 2e-05,
+      "loss": 0.0343626,
+      "step": 7121
+    },
+    {
+      "epoch": 14.244,
+      "grad_norm": 0.960382878780365,
+      "learning_rate": 2e-05,
+      "loss": 0.02384898,
+      "step": 7122
+    },
+    {
+      "epoch": 14.246,
+      "grad_norm": 4.224244117736816,
+      "learning_rate": 2e-05,
+      "loss": 0.03810407,
+      "step": 7123
+    },
+    {
+      "epoch": 14.248,
+      "grad_norm": 1.0069541931152344,
+      "learning_rate": 2e-05,
+      "loss": 0.03257271,
+      "step": 7124
+    },
+    {
+      "epoch": 14.25,
+      "grad_norm": 1.2396907806396484,
+      "learning_rate": 2e-05,
+      "loss": 0.03927642,
+      "step": 7125
+    },
+    {
+      "epoch": 14.252,
+      "grad_norm": 1.199365258216858,
+      "learning_rate": 2e-05,
+      "loss": 0.04600928,
+      "step": 7126
+    },
+    {
+      "epoch": 14.254,
+      "grad_norm": 1.4900243282318115,
+      "learning_rate": 2e-05,
+      "loss": 0.03931838,
+      "step": 7127
+    },
+    {
+      "epoch": 14.256,
+      "grad_norm": 1.528289556503296,
+      "learning_rate": 2e-05,
+      "loss": 0.02803968,
+      "step": 7128
+    },
+    {
+      "epoch": 14.258,
+      "grad_norm": 1.347772240638733,
+      "learning_rate": 2e-05,
+      "loss": 0.03262594,
+      "step": 7129
+    },
+    {
+      "epoch": 14.26,
+      "grad_norm": 2.532207489013672,
+      "learning_rate": 2e-05,
+      "loss": 0.05416807,
+      "step": 7130
+    },
+    {
+      "epoch": 14.262,
+      "grad_norm": 1.2191015481948853,
+      "learning_rate": 2e-05,
+      "loss": 0.03117691,
+      "step": 7131
+    },
+    {
+      "epoch": 14.264,
+      "grad_norm": 2.1217501163482666,
+      "learning_rate": 2e-05,
+      "loss": 0.03810145,
+      "step": 7132
+    },
+    {
+      "epoch": 14.266,
+      "grad_norm": 1.506582498550415,
+      "learning_rate": 2e-05,
+      "loss": 0.0460251,
+      "step": 7133
+    },
+    {
+      "epoch": 14.268,
+      "grad_norm": 1.2872868776321411,
+      "learning_rate": 2e-05,
+      "loss": 0.04550285,
+      "step": 7134
+    },
+    {
+      "epoch": 14.27,
+      "grad_norm": 1.8632010221481323,
+      "learning_rate": 2e-05,
+      "loss": 0.04599176,
+      "step": 7135
+    },
+    {
+      "epoch": 14.272,
+      "grad_norm": 1.4420409202575684,
+      "learning_rate": 2e-05,
+      "loss": 0.04499146,
+      "step": 7136
+    },
+    {
+      "epoch": 14.274000000000001,
+      "grad_norm": 2.0591557025909424,
+      "learning_rate": 2e-05,
+      "loss": 0.04166877,
+      "step": 7137
+    },
+    {
+      "epoch": 14.276,
+      "grad_norm": 2.0912609100341797,
+      "learning_rate": 2e-05,
+      "loss": 0.05226006,
+      "step": 7138
+    },
+    {
+      "epoch": 14.278,
+      "grad_norm": 1.5423698425292969,
+      "learning_rate": 2e-05,
+      "loss": 0.03722187,
+      "step": 7139
+    },
+    {
+      "epoch": 14.28,
+      "grad_norm": 2.2733895778656006,
+      "learning_rate": 2e-05,
+      "loss": 0.04837454,
+      "step": 7140
+    },
+    {
+      "epoch": 14.282,
+      "grad_norm": 1.218680739402771,
+      "learning_rate": 2e-05,
+      "loss": 0.03407823,
+      "step": 7141
+    },
+    {
+      "epoch": 14.284,
+      "grad_norm": 1.3564642667770386,
+      "learning_rate": 2e-05,
+      "loss": 0.04712793,
+      "step": 7142
+    },
+    {
+      "epoch": 14.286,
+      "grad_norm": 1.7670422792434692,
+      "learning_rate": 2e-05,
+      "loss": 0.04199456,
+      "step": 7143
+    },
+    {
+      "epoch": 14.288,
+      "grad_norm": 1.2026097774505615,
+      "learning_rate": 2e-05,
+      "loss": 0.03444853,
+      "step": 7144
+    },
+    {
+      "epoch": 14.29,
+      "grad_norm": 1.443743348121643,
+      "learning_rate": 2e-05,
+      "loss": 0.04106067,
+      "step": 7145
+    },
+    {
+      "epoch": 14.292,
+      "grad_norm": 1.2027446031570435,
+      "learning_rate": 2e-05,
+      "loss": 0.03707718,
+      "step": 7146
+    },
+    {
+      "epoch": 14.294,
+      "grad_norm": 1.0716592073440552,
+      "learning_rate": 2e-05,
+      "loss": 0.04112831,
+      "step": 7147
+    },
+    {
+      "epoch": 14.296,
+      "grad_norm": 1.7344224452972412,
+      "learning_rate": 2e-05,
+      "loss": 0.03029624,
+      "step": 7148
+    },
+    {
+      "epoch": 14.298,
+      "grad_norm": 2.744713068008423,
+      "learning_rate": 2e-05,
+      "loss": 0.03705482,
+      "step": 7149
+    },
+    {
+      "epoch": 14.3,
+      "grad_norm": 1.1636744737625122,
+      "learning_rate": 2e-05,
+      "loss": 0.0361629,
+      "step": 7150
+    },
+    {
+      "epoch": 14.302,
+      "grad_norm": 1.6198397874832153,
+      "learning_rate": 2e-05,
+      "loss": 0.05550705,
+      "step": 7151
+    },
+    {
+      "epoch": 14.304,
+      "grad_norm": 1.4462108612060547,
+      "learning_rate": 2e-05,
+      "loss": 0.03110327,
+      "step": 7152
+    },
+    {
+      "epoch": 14.306,
+      "grad_norm": 2.7012014389038086,
+      "learning_rate": 2e-05,
+      "loss": 0.05303191,
+      "step": 7153
+    },
+    {
+      "epoch": 14.308,
+      "grad_norm": 0.8201614022254944,
+      "learning_rate": 2e-05,
+      "loss": 0.02300132,
+      "step": 7154
+    },
+    {
+      "epoch": 14.31,
+      "grad_norm": 1.4760310649871826,
+      "learning_rate": 2e-05,
+      "loss": 0.04489978,
+      "step": 7155
+    },
+    {
+      "epoch": 14.312,
+      "grad_norm": 1.3183153867721558,
+      "learning_rate": 2e-05,
+      "loss": 0.04012213,
+      "step": 7156
+    },
+    {
+      "epoch": 14.314,
+      "grad_norm": 1.5775775909423828,
+      "learning_rate": 2e-05,
+      "loss": 0.04109515,
+      "step": 7157
+    },
+    {
+      "epoch": 14.316,
+      "grad_norm": 1.6039639711380005,
+      "learning_rate": 2e-05,
+      "loss": 0.04469662,
+      "step": 7158
+    },
+    {
+      "epoch": 14.318,
+      "grad_norm": 1.8237558603286743,
+      "learning_rate": 2e-05,
+      "loss": 0.05006624,
+      "step": 7159
+    },
+    {
+      "epoch": 14.32,
+      "grad_norm": 1.929965615272522,
+      "learning_rate": 2e-05,
+      "loss": 0.05642921,
+      "step": 7160
+    },
+    {
+      "epoch": 14.322,
+      "grad_norm": 1.932225227355957,
+      "learning_rate": 2e-05,
+      "loss": 0.05633236,
+      "step": 7161
+    },
+    {
+      "epoch": 14.324,
+      "grad_norm": 1.1771067380905151,
+      "learning_rate": 2e-05,
+      "loss": 0.03297538,
+      "step": 7162
+    },
+    {
+      "epoch": 14.326,
+      "grad_norm": 1.7193629741668701,
+      "learning_rate": 2e-05,
+      "loss": 0.04758242,
+      "step": 7163
+    },
+    {
+      "epoch": 14.328,
+      "grad_norm": 1.122786521911621,
+      "learning_rate": 2e-05,
+      "loss": 0.04002685,
+      "step": 7164
+    },
+    {
+      "epoch": 14.33,
+      "grad_norm": 1.0502139329910278,
+      "learning_rate": 2e-05,
+      "loss": 0.03614987,
+      "step": 7165
+    },
+    {
+      "epoch": 14.332,
+      "grad_norm": 1.1357371807098389,
+      "learning_rate": 2e-05,
+      "loss": 0.03864967,
+      "step": 7166
+    },
+    {
+      "epoch": 14.334,
+      "grad_norm": 1.074387788772583,
+      "learning_rate": 2e-05,
+      "loss": 0.03561043,
+      "step": 7167
+    },
+    {
+      "epoch": 14.336,
+      "grad_norm": 1.5098060369491577,
+      "learning_rate": 2e-05,
+      "loss": 0.04273804,
+      "step": 7168
+    },
+    {
+      "epoch": 14.338,
+      "grad_norm": 1.3217259645462036,
+      "learning_rate": 2e-05,
+      "loss": 0.05451931,
+      "step": 7169
+    },
+    {
+      "epoch": 14.34,
+      "grad_norm": 1.2757598161697388,
+      "learning_rate": 2e-05,
+      "loss": 0.03788436,
+      "step": 7170
+    },
+    {
+      "epoch": 14.342,
+      "grad_norm": 0.9583383798599243,
+      "learning_rate": 2e-05,
+      "loss": 0.03389092,
+      "step": 7171
+    },
+    {
+      "epoch": 14.344,
+      "grad_norm": 1.8409156799316406,
+      "learning_rate": 2e-05,
+      "loss": 0.05510804,
+      "step": 7172
+    },
+    {
+      "epoch": 14.346,
+      "grad_norm": 1.3510578870773315,
+      "learning_rate": 2e-05,
+      "loss": 0.03992224,
+      "step": 7173
+    },
+    {
+      "epoch": 14.348,
+      "grad_norm": 1.52592933177948,
+      "learning_rate": 2e-05,
+      "loss": 0.04563883,
+      "step": 7174
+    },
+    {
+      "epoch": 14.35,
+      "grad_norm": 1.946403980255127,
+      "learning_rate": 2e-05,
+      "loss": 0.03222995,
+      "step": 7175
+    },
+    {
+      "epoch": 14.352,
+      "grad_norm": 1.9366340637207031,
+      "learning_rate": 2e-05,
+      "loss": 0.05663209,
+      "step": 7176
+    },
+    {
+      "epoch": 14.354,
+      "grad_norm": 0.7667877078056335,
+      "learning_rate": 2e-05,
+      "loss": 0.01848971,
+      "step": 7177
+    },
+    {
+      "epoch": 14.356,
+      "grad_norm": 2.2338192462921143,
+      "learning_rate": 2e-05,
+      "loss": 0.0433307,
+      "step": 7178
+    },
+    {
+      "epoch": 14.358,
+      "grad_norm": 3.6205711364746094,
+      "learning_rate": 2e-05,
+      "loss": 0.0502264,
+      "step": 7179
+    },
+    {
+      "epoch": 14.36,
+      "grad_norm": 1.8776448965072632,
+      "learning_rate": 2e-05,
+      "loss": 0.05654108,
+      "step": 7180
+    },
+    {
+      "epoch": 14.362,
+      "grad_norm": 1.7132173776626587,
+      "learning_rate": 2e-05,
+      "loss": 0.03880975,
+      "step": 7181
+    },
+    {
+      "epoch": 14.364,
+      "grad_norm": 1.5955792665481567,
+      "learning_rate": 2e-05,
+      "loss": 0.04336097,
+      "step": 7182
+    },
+    {
+      "epoch": 14.366,
+      "grad_norm": 1.046798825263977,
+      "learning_rate": 2e-05,
+      "loss": 0.03609883,
+      "step": 7183
+    },
+    {
+      "epoch": 14.368,
+      "grad_norm": 1.6806223392486572,
+      "learning_rate": 2e-05,
+      "loss": 0.04325953,
+      "step": 7184
+    },
+    {
+      "epoch": 14.37,
+      "grad_norm": 1.8983646631240845,
+      "learning_rate": 2e-05,
+      "loss": 0.04521517,
+      "step": 7185
+    },
+    {
+      "epoch": 14.372,
+      "grad_norm": 2.0524489879608154,
+      "learning_rate": 2e-05,
+      "loss": 0.03215559,
+      "step": 7186
+    },
+    {
+      "epoch": 14.374,
+      "grad_norm": 1.1298598051071167,
+      "learning_rate": 2e-05,
+      "loss": 0.03601304,
+      "step": 7187
+    },
+    {
+      "epoch": 14.376,
+      "grad_norm": 1.4891972541809082,
+      "learning_rate": 2e-05,
+      "loss": 0.03756123,
+      "step": 7188
+    },
+    {
+      "epoch": 14.378,
+      "grad_norm": 1.3420482873916626,
+      "learning_rate": 2e-05,
+      "loss": 0.04583802,
+      "step": 7189
+    },
+    {
+      "epoch": 14.38,
+      "grad_norm": 5.2002716064453125,
+      "learning_rate": 2e-05,
+      "loss": 0.04100659,
+      "step": 7190
+    },
+    {
+      "epoch": 14.382,
+      "grad_norm": 1.079121470451355,
+      "learning_rate": 2e-05,
+      "loss": 0.03196656,
+      "step": 7191
+    },
+    {
+      "epoch": 14.384,
+      "grad_norm": 2.1130754947662354,
+      "learning_rate": 2e-05,
+      "loss": 0.05068072,
+      "step": 7192
+    },
+    {
+      "epoch": 14.386,
+      "grad_norm": 1.6315432786941528,
+      "learning_rate": 2e-05,
+      "loss": 0.03901272,
+      "step": 7193
+    },
+    {
+      "epoch": 14.388,
+      "grad_norm": 1.7703206539154053,
+      "learning_rate": 2e-05,
+      "loss": 0.03575943,
+      "step": 7194
+    },
+    {
+      "epoch": 14.39,
+      "grad_norm": 1.4890453815460205,
+      "learning_rate": 2e-05,
+      "loss": 0.03911097,
+      "step": 7195
+    },
+    {
+      "epoch": 14.392,
+      "grad_norm": 1.2721261978149414,
+      "learning_rate": 2e-05,
+      "loss": 0.04248876,
+      "step": 7196
+    },
+    {
+      "epoch": 14.394,
+      "grad_norm": 1.8303420543670654,
+      "learning_rate": 2e-05,
+      "loss": 0.05177449,
+      "step": 7197
+    },
+    {
+      "epoch": 14.396,
+      "grad_norm": 1.322717308998108,
+      "learning_rate": 2e-05,
+      "loss": 0.02881128,
+      "step": 7198
+    },
+    {
+      "epoch": 14.398,
+      "grad_norm": 1.5059839487075806,
+      "learning_rate": 2e-05,
+      "loss": 0.04383879,
+      "step": 7199
+    },
+    {
+      "epoch": 14.4,
+      "grad_norm": 1.64779794216156,
+      "learning_rate": 2e-05,
+      "loss": 0.0525568,
+      "step": 7200
+    },
+    {
+      "epoch": 14.402,
+      "grad_norm": 2.455418586730957,
+      "learning_rate": 2e-05,
+      "loss": 0.04642213,
+      "step": 7201
+    },
+    {
+      "epoch": 14.404,
+      "grad_norm": 1.2725389003753662,
+      "learning_rate": 2e-05,
+      "loss": 0.03739651,
+      "step": 7202
+    },
+    {
+      "epoch": 14.406,
+      "grad_norm": 1.4967567920684814,
+      "learning_rate": 2e-05,
+      "loss": 0.04378046,
+      "step": 7203
+    },
+    {
+      "epoch": 14.408,
+      "grad_norm": 1.1510059833526611,
+      "learning_rate": 2e-05,
+      "loss": 0.03842362,
+      "step": 7204
+    },
+    {
+      "epoch": 14.41,
+      "grad_norm": 1.4995372295379639,
+      "learning_rate": 2e-05,
+      "loss": 0.04412612,
+      "step": 7205
+    },
+    {
+      "epoch": 14.412,
+      "grad_norm": 0.94809490442276,
+      "learning_rate": 2e-05,
+      "loss": 0.03201088,
+      "step": 7206
+    },
+    {
+      "epoch": 14.414,
+      "grad_norm": 1.5775501728057861,
+      "learning_rate": 2e-05,
+      "loss": 0.03421389,
+      "step": 7207
+    },
+    {
+      "epoch": 14.416,
+      "grad_norm": 3.60548996925354,
+      "learning_rate": 2e-05,
+      "loss": 0.0545315,
+      "step": 7208
+    },
+    {
+      "epoch": 14.418,
+      "grad_norm": 1.0393273830413818,
+      "learning_rate": 2e-05,
+      "loss": 0.03813236,
+      "step": 7209
+    },
+    {
+      "epoch": 14.42,
+      "grad_norm": 1.0074549913406372,
+      "learning_rate": 2e-05,
+      "loss": 0.02983594,
+      "step": 7210
+    },
+    {
+      "epoch": 14.422,
+      "grad_norm": 1.203730583190918,
+      "learning_rate": 2e-05,
+      "loss": 0.04196813,
+      "step": 7211
+    },
+    {
+      "epoch": 14.424,
+      "grad_norm": 1.2036176919937134,
+      "learning_rate": 2e-05,
+      "loss": 0.02699343,
+      "step": 7212
+    },
+    {
+      "epoch": 14.426,
+      "grad_norm": 1.3977973461151123,
+      "learning_rate": 2e-05,
+      "loss": 0.02824752,
+      "step": 7213
+    },
+    {
+      "epoch": 14.428,
+      "grad_norm": 1.019425392150879,
+      "learning_rate": 2e-05,
+      "loss": 0.03191173,
+      "step": 7214
+    },
+    {
+      "epoch": 14.43,
+      "grad_norm": 0.8422673940658569,
+      "learning_rate": 2e-05,
+      "loss": 0.02386178,
+      "step": 7215
+    },
+    {
+      "epoch": 14.432,
+      "grad_norm": 1.4867335557937622,
+      "learning_rate": 2e-05,
+      "loss": 0.03482441,
+      "step": 7216
+    },
+    {
+      "epoch": 14.434,
+      "grad_norm": 2.1186928749084473,
+      "learning_rate": 2e-05,
+      "loss": 0.05948957,
+      "step": 7217
+    },
+    {
+      "epoch": 14.436,
+      "grad_norm": 1.7550668716430664,
+      "learning_rate": 2e-05,
+      "loss": 0.04442699,
+      "step": 7218
+    },
+    {
+      "epoch": 14.438,
+      "grad_norm": 1.0397560596466064,
+      "learning_rate": 2e-05,
+      "loss": 0.0290125,
+      "step": 7219
+    },
+    {
+      "epoch": 14.44,
+      "grad_norm": 1.3454571962356567,
+      "learning_rate": 2e-05,
+      "loss": 0.03583448,
+      "step": 7220
+    },
+    {
+      "epoch": 14.442,
+      "grad_norm": 1.2962967157363892,
+      "learning_rate": 2e-05,
+      "loss": 0.0369962,
+      "step": 7221
+    },
+    {
+      "epoch": 14.444,
+      "grad_norm": 1.420045256614685,
+      "learning_rate": 2e-05,
+      "loss": 0.03899933,
+      "step": 7222
+    },
+    {
+      "epoch": 14.446,
+      "grad_norm": 2.546292781829834,
+      "learning_rate": 2e-05,
+      "loss": 0.05628195,
+      "step": 7223
+    },
+    {
+      "epoch": 14.448,
+      "grad_norm": 1.8872132301330566,
+      "learning_rate": 2e-05,
+      "loss": 0.03206664,
+      "step": 7224
+    },
+    {
+      "epoch": 14.45,
+      "grad_norm": 0.9982779622077942,
+      "learning_rate": 2e-05,
+      "loss": 0.01746661,
+      "step": 7225
+    },
+    {
+      "epoch": 14.452,
+      "grad_norm": 2.069234609603882,
+      "learning_rate": 2e-05,
+      "loss": 0.04197174,
+      "step": 7226
+    },
+    {
+      "epoch": 14.454,
+      "grad_norm": 1.5738203525543213,
+      "learning_rate": 2e-05,
+      "loss": 0.04109071,
+      "step": 7227
+    },
+    {
+      "epoch": 14.456,
+      "grad_norm": 1.1500931978225708,
+      "learning_rate": 2e-05,
+      "loss": 0.03633848,
+      "step": 7228
+    },
+    {
+      "epoch": 14.458,
+      "grad_norm": 1.2482191324234009,
+      "learning_rate": 2e-05,
+      "loss": 0.04544804,
+      "step": 7229
+    },
+    {
+      "epoch": 14.46,
+      "grad_norm": 1.5206077098846436,
+      "learning_rate": 2e-05,
+      "loss": 0.03966599,
+      "step": 7230
+    },
+    {
+      "epoch": 14.462,
+      "grad_norm": 1.76610267162323,
+      "learning_rate": 2e-05,
+      "loss": 0.04585,
+      "step": 7231
+    },
+    {
+      "epoch": 14.464,
+      "grad_norm": 1.5877853631973267,
+      "learning_rate": 2e-05,
+      "loss": 0.03790272,
+      "step": 7232
+    },
+    {
+      "epoch": 14.466,
+      "grad_norm": 1.1064239740371704,
+      "learning_rate": 2e-05,
+      "loss": 0.03053769,
+      "step": 7233
+    },
+    {
+      "epoch": 14.468,
+      "grad_norm": 1.0071042776107788,
+      "learning_rate": 2e-05,
+      "loss": 0.03742989,
+      "step": 7234
+    },
+    {
+      "epoch": 14.47,
+      "grad_norm": 2.6944468021392822,
+      "learning_rate": 2e-05,
+      "loss": 0.05254611,
+      "step": 7235
+    },
+    {
+      "epoch": 14.472,
+      "grad_norm": 1.4075301885604858,
+      "learning_rate": 2e-05,
+      "loss": 0.05554926,
+      "step": 7236
+    },
+    {
+      "epoch": 14.474,
+      "grad_norm": 1.048970103263855,
+      "learning_rate": 2e-05,
+      "loss": 0.0344836,
+      "step": 7237
+    },
+    {
+      "epoch": 14.475999999999999,
+      "grad_norm": 0.9587178826332092,
+      "learning_rate": 2e-05,
+      "loss": 0.03106603,
+      "step": 7238
+    },
+    {
+      "epoch": 14.478,
+      "grad_norm": 1.0842715501785278,
+      "learning_rate": 2e-05,
+      "loss": 0.03258051,
+      "step": 7239
+    },
+    {
+      "epoch": 14.48,
+      "grad_norm": 1.1892338991165161,
+      "learning_rate": 2e-05,
+      "loss": 0.04680737,
+      "step": 7240
+    },
+    {
+      "epoch": 14.482,
+      "grad_norm": 1.856372594833374,
+      "learning_rate": 2e-05,
+      "loss": 0.05747771,
+      "step": 7241
+    },
+    {
+      "epoch": 14.484,
+      "grad_norm": 1.1554561853408813,
+      "learning_rate": 2e-05,
+      "loss": 0.02923879,
+      "step": 7242
+    },
+    {
+      "epoch": 14.486,
+      "grad_norm": 1.0455029010772705,
+      "learning_rate": 2e-05,
+      "loss": 0.03191773,
+      "step": 7243
+    },
+    {
+      "epoch": 14.488,
+      "grad_norm": 2.2019166946411133,
+      "learning_rate": 2e-05,
+      "loss": 0.04987948,
+      "step": 7244
+    },
+    {
+      "epoch": 14.49,
+      "grad_norm": 1.404489278793335,
+      "learning_rate": 2e-05,
+      "loss": 0.03941087,
+      "step": 7245
+    },
+    {
+      "epoch": 14.492,
+      "grad_norm": 1.0887120962142944,
+      "learning_rate": 2e-05,
+      "loss": 0.03620033,
+      "step": 7246
+    },
+    {
+      "epoch": 14.494,
+      "grad_norm": 1.727051019668579,
+      "learning_rate": 2e-05,
+      "loss": 0.03951932,
+      "step": 7247
+    },
+    {
+      "epoch": 14.496,
+      "grad_norm": 1.1405435800552368,
+      "learning_rate": 2e-05,
+      "loss": 0.04180471,
+      "step": 7248
+    },
+    {
+      "epoch": 14.498,
+      "grad_norm": 1.45590341091156,
+      "learning_rate": 2e-05,
+      "loss": 0.035485,
+      "step": 7249
+    },
+    {
+      "epoch": 14.5,
+      "grad_norm": 1.9948844909667969,
+      "learning_rate": 2e-05,
+      "loss": 0.04214467,
+      "step": 7250
+    },
+    {
+      "epoch": 14.502,
+      "grad_norm": 1.238334059715271,
+      "learning_rate": 2e-05,
+      "loss": 0.03087533,
+      "step": 7251
+    },
+    {
+      "epoch": 14.504,
+      "grad_norm": 1.136520504951477,
+      "learning_rate": 2e-05,
+      "loss": 0.03111343,
+      "step": 7252
+    },
+    {
+      "epoch": 14.506,
+      "grad_norm": 1.5107890367507935,
+      "learning_rate": 2e-05,
+      "loss": 0.03291022,
+      "step": 7253
+    },
+    {
+      "epoch": 14.508,
+      "grad_norm": 1.3895598649978638,
+      "learning_rate": 2e-05,
+      "loss": 0.04216557,
+      "step": 7254
+    },
+    {
+      "epoch": 14.51,
+      "grad_norm": 2.2805137634277344,
+      "learning_rate": 2e-05,
+      "loss": 0.04727894,
+      "step": 7255
+    },
+    {
+      "epoch": 14.512,
+      "grad_norm": 1.5463753938674927,
+      "learning_rate": 2e-05,
+      "loss": 0.03055894,
+      "step": 7256
+    },
+    {
+      "epoch": 14.514,
+      "grad_norm": 0.9866666793823242,
+      "learning_rate": 2e-05,
+      "loss": 0.02884018,
+      "step": 7257
+    },
+    {
+      "epoch": 14.516,
+      "grad_norm": 1.2003675699234009,
+      "learning_rate": 2e-05,
+      "loss": 0.03621934,
+      "step": 7258
+    },
+    {
+      "epoch": 14.518,
+      "grad_norm": 2.279313802719116,
+      "learning_rate": 2e-05,
+      "loss": 0.08503312,
+      "step": 7259
+    },
+    {
+      "epoch": 14.52,
+      "grad_norm": 2.878683090209961,
+      "learning_rate": 2e-05,
+      "loss": 0.03383691,
+      "step": 7260
+    },
+    {
+      "epoch": 14.522,
+      "grad_norm": 1.457414150238037,
+      "learning_rate": 2e-05,
+      "loss": 0.04102724,
+      "step": 7261
+    },
+    {
+      "epoch": 14.524000000000001,
+      "grad_norm": 1.2874579429626465,
+      "learning_rate": 2e-05,
+      "loss": 0.02973592,
+      "step": 7262
+    },
+    {
+      "epoch": 14.526,
+      "grad_norm": 1.9015241861343384,
+      "learning_rate": 2e-05,
+      "loss": 0.04436665,
+      "step": 7263
+    },
+    {
+      "epoch": 14.528,
+      "grad_norm": 1.2383368015289307,
+      "learning_rate": 2e-05,
+      "loss": 0.03040004,
+      "step": 7264
+    },
+    {
+      "epoch": 14.53,
+      "grad_norm": 1.5358835458755493,
+      "learning_rate": 2e-05,
+      "loss": 0.05725804,
+      "step": 7265
+    },
+    {
+      "epoch": 14.532,
+      "grad_norm": 2.0177321434020996,
+      "learning_rate": 2e-05,
+      "loss": 0.04995024,
+      "step": 7266
+    },
+    {
+      "epoch": 14.534,
+      "grad_norm": 1.6162405014038086,
+      "learning_rate": 2e-05,
+      "loss": 0.03423398,
+      "step": 7267
+    },
+    {
+      "epoch": 14.536,
+      "grad_norm": 1.3776979446411133,
+      "learning_rate": 2e-05,
+      "loss": 0.03925561,
+      "step": 7268
+    },
+    {
+      "epoch": 14.538,
+      "grad_norm": 1.2735576629638672,
+      "learning_rate": 2e-05,
+      "loss": 0.03249297,
+      "step": 7269
+    },
+    {
+      "epoch": 14.54,
+      "grad_norm": 1.3519257307052612,
+      "learning_rate": 2e-05,
+      "loss": 0.03134355,
+      "step": 7270
+    },
+    {
+      "epoch": 14.542,
+      "grad_norm": 1.2011692523956299,
+      "learning_rate": 2e-05,
+      "loss": 0.03818227,
+      "step": 7271
+    },
+    {
+      "epoch": 14.544,
+      "grad_norm": 1.6174978017807007,
+      "learning_rate": 2e-05,
+      "loss": 0.04904899,
+      "step": 7272
+    },
+    {
+      "epoch": 14.546,
+      "grad_norm": 1.3549933433532715,
+      "learning_rate": 2e-05,
+      "loss": 0.04195243,
+      "step": 7273
+    },
+    {
+      "epoch": 14.548,
+      "grad_norm": 1.4086804389953613,
+      "learning_rate": 2e-05,
+      "loss": 0.03927261,
+      "step": 7274
+    },
+    {
+      "epoch": 14.55,
+      "grad_norm": 1.495468258857727,
+      "learning_rate": 2e-05,
+      "loss": 0.03486994,
+      "step": 7275
+    },
+    {
+      "epoch": 14.552,
+      "grad_norm": 1.4550012350082397,
+      "learning_rate": 2e-05,
+      "loss": 0.04446776,
+      "step": 7276
+    },
+    {
+      "epoch": 14.554,
+      "grad_norm": 1.0507636070251465,
+      "learning_rate": 2e-05,
+      "loss": 0.02539073,
+      "step": 7277
+    },
+    {
+      "epoch": 14.556000000000001,
+      "grad_norm": 6.91660213470459,
+      "learning_rate": 2e-05,
+      "loss": 0.03748736,
+      "step": 7278
+    },
+    {
+      "epoch": 14.558,
+      "grad_norm": 2.021453857421875,
+      "learning_rate": 2e-05,
+      "loss": 0.05216369,
+      "step": 7279
+    },
+    {
+      "epoch": 14.56,
+      "grad_norm": 1.6024717092514038,
+      "learning_rate": 2e-05,
+      "loss": 0.05282374,
+      "step": 7280
+    },
+    {
+      "epoch": 14.562,
+      "grad_norm": 1.2320939302444458,
+      "learning_rate": 2e-05,
+      "loss": 0.03876211,
+      "step": 7281
+    },
+    {
+      "epoch": 14.564,
+      "grad_norm": 1.198306679725647,
+      "learning_rate": 2e-05,
+      "loss": 0.03282074,
+      "step": 7282
+    },
+    {
+      "epoch": 14.566,
+      "grad_norm": 1.5773409605026245,
+      "learning_rate": 2e-05,
+      "loss": 0.03945959,
+      "step": 7283
+    },
+    {
+      "epoch": 14.568,
+      "grad_norm": 1.21599543094635,
+      "learning_rate": 2e-05,
+      "loss": 0.04125897,
+      "step": 7284
+    },
+    {
+      "epoch": 14.57,
+      "grad_norm": 2.8407862186431885,
+      "learning_rate": 2e-05,
+      "loss": 0.04945897,
+      "step": 7285
+    },
+    {
+      "epoch": 14.572,
+      "grad_norm": 1.447121500968933,
+      "learning_rate": 2e-05,
+      "loss": 0.03155572,
+      "step": 7286
+    },
+    {
+      "epoch": 14.574,
+      "grad_norm": 1.1770319938659668,
+      "learning_rate": 2e-05,
+      "loss": 0.03424954,
+      "step": 7287
+    },
+    {
+      "epoch": 14.576,
+      "grad_norm": 1.8243449926376343,
+      "learning_rate": 2e-05,
+      "loss": 0.04884764,
+      "step": 7288
+    },
+    {
+      "epoch": 14.578,
+      "grad_norm": 0.871478259563446,
+      "learning_rate": 2e-05,
+      "loss": 0.02348526,
+      "step": 7289
+    },
+    {
+      "epoch": 14.58,
+      "grad_norm": 1.1215943098068237,
+      "learning_rate": 2e-05,
+      "loss": 0.03072283,
+      "step": 7290
+    },
+    {
+      "epoch": 14.582,
+      "grad_norm": 1.7615373134613037,
+      "learning_rate": 2e-05,
+      "loss": 0.05309131,
+      "step": 7291
+    },
+    {
+      "epoch": 14.584,
+      "grad_norm": 2.3434910774230957,
+      "learning_rate": 2e-05,
+      "loss": 0.03328982,
+      "step": 7292
+    },
+    {
+      "epoch": 14.586,
+      "grad_norm": 2.394543409347534,
+      "learning_rate": 2e-05,
+      "loss": 0.03665608,
+      "step": 7293
+    },
+    {
+      "epoch": 14.588,
+      "grad_norm": 1.7310210466384888,
+      "learning_rate": 2e-05,
+      "loss": 0.04254977,
+      "step": 7294
+    },
+    {
+      "epoch": 14.59,
+      "grad_norm": 1.5410484075546265,
+      "learning_rate": 2e-05,
+      "loss": 0.03172578,
+      "step": 7295
+    },
+    {
+      "epoch": 14.592,
+      "grad_norm": 1.7711609601974487,
+      "learning_rate": 2e-05,
+      "loss": 0.04883514,
+      "step": 7296
+    },
+    {
+      "epoch": 14.594,
+      "grad_norm": 1.606502652168274,
+      "learning_rate": 2e-05,
+      "loss": 0.03369958,
+      "step": 7297
+    },
+    {
+      "epoch": 14.596,
+      "grad_norm": 1.9675401449203491,
+      "learning_rate": 2e-05,
+      "loss": 0.03461675,
+      "step": 7298
+    },
+    {
+      "epoch": 14.598,
+      "grad_norm": 1.6831592321395874,
+      "learning_rate": 2e-05,
+      "loss": 0.04358295,
+      "step": 7299
+    },
+    {
+      "epoch": 14.6,
+      "grad_norm": 1.846706748008728,
+      "learning_rate": 2e-05,
+      "loss": 0.04121499,
+      "step": 7300
+    },
+    {
+      "epoch": 14.602,
+      "grad_norm": 1.6495356559753418,
+      "learning_rate": 2e-05,
+      "loss": 0.04087245,
+      "step": 7301
+    },
+    {
+      "epoch": 14.604,
+      "grad_norm": 1.3375145196914673,
+      "learning_rate": 2e-05,
+      "loss": 0.02481322,
+      "step": 7302
+    },
+    {
+      "epoch": 14.606,
+      "grad_norm": 1.3421443700790405,
+      "learning_rate": 2e-05,
+      "loss": 0.04807069,
+      "step": 7303
+    },
+    {
+      "epoch": 14.608,
+      "grad_norm": 1.5804858207702637,
+      "learning_rate": 2e-05,
+      "loss": 0.03445378,
+      "step": 7304
+    },
+    {
+      "epoch": 14.61,
+      "grad_norm": 3.078519344329834,
+      "learning_rate": 2e-05,
+      "loss": 0.05874301,
+      "step": 7305
+    },
+    {
+      "epoch": 14.612,
+      "grad_norm": 1.269014596939087,
+      "learning_rate": 2e-05,
+      "loss": 0.03813095,
+      "step": 7306
+    },
+    {
+      "epoch": 14.614,
+      "grad_norm": 1.667083978652954,
+      "learning_rate": 2e-05,
+      "loss": 0.05303952,
+      "step": 7307
+    },
+    {
+      "epoch": 14.616,
+      "grad_norm": 1.014502763748169,
+      "learning_rate": 2e-05,
+      "loss": 0.0317478,
+      "step": 7308
+    },
+    {
+      "epoch": 14.618,
+      "grad_norm": 1.066616177558899,
+      "learning_rate": 2e-05,
+      "loss": 0.02247791,
+      "step": 7309
+    },
+    {
+      "epoch": 14.62,
+      "grad_norm": 1.208592176437378,
+      "learning_rate": 2e-05,
+      "loss": 0.02994377,
+      "step": 7310
+    },
+    {
+      "epoch": 14.622,
+      "grad_norm": 0.8381512761116028,
+      "learning_rate": 2e-05,
+      "loss": 0.02272216,
+      "step": 7311
+    },
+    {
+      "epoch": 14.624,
+      "grad_norm": 1.4301813840866089,
+      "learning_rate": 2e-05,
+      "loss": 0.04580442,
+      "step": 7312
+    },
+    {
+      "epoch": 14.626,
+      "grad_norm": 1.006054401397705,
+      "learning_rate": 2e-05,
+      "loss": 0.02406761,
+      "step": 7313
+    },
+    {
+      "epoch": 14.628,
+      "grad_norm": 2.2879605293273926,
+      "learning_rate": 2e-05,
+      "loss": 0.05421574,
+      "step": 7314
+    },
+    {
+      "epoch": 14.63,
+      "grad_norm": 1.347216010093689,
+      "learning_rate": 2e-05,
+      "loss": 0.03223877,
+      "step": 7315
+    },
+    {
+      "epoch": 14.632,
+      "grad_norm": 1.3925851583480835,
+      "learning_rate": 2e-05,
+      "loss": 0.04525027,
+      "step": 7316
+    },
+    {
+      "epoch": 14.634,
+      "grad_norm": 1.5058016777038574,
+      "learning_rate": 2e-05,
+      "loss": 0.05671232,
+      "step": 7317
+    },
+    {
+      "epoch": 14.636,
+      "grad_norm": 1.2585381269454956,
+      "learning_rate": 2e-05,
+      "loss": 0.04648329,
+      "step": 7318
+    },
+    {
+      "epoch": 14.638,
+      "grad_norm": 2.345365047454834,
+      "learning_rate": 2e-05,
+      "loss": 0.02963935,
+      "step": 7319
+    },
+    {
+      "epoch": 14.64,
+      "grad_norm": 1.1654967069625854,
+      "learning_rate": 2e-05,
+      "loss": 0.0446506,
+      "step": 7320
+    },
+    {
+      "epoch": 14.642,
+      "grad_norm": 2.1558237075805664,
+      "learning_rate": 2e-05,
+      "loss": 0.04308884,
+      "step": 7321
+    },
+    {
+      "epoch": 14.644,
+      "grad_norm": 1.271641492843628,
+      "learning_rate": 2e-05,
+      "loss": 0.0442902,
+      "step": 7322
+    },
+    {
+      "epoch": 14.646,
+      "grad_norm": 1.7214982509613037,
+      "learning_rate": 2e-05,
+      "loss": 0.0474602,
+      "step": 7323
+    },
+    {
+      "epoch": 14.648,
+      "grad_norm": 1.1854116916656494,
+      "learning_rate": 2e-05,
+      "loss": 0.04027828,
+      "step": 7324
+    },
+    {
+      "epoch": 14.65,
+      "grad_norm": 1.9680629968643188,
+      "learning_rate": 2e-05,
+      "loss": 0.04050666,
+      "step": 7325
+    },
+    {
+      "epoch": 14.652,
+      "grad_norm": 1.021872878074646,
+      "learning_rate": 2e-05,
+      "loss": 0.03418166,
+      "step": 7326
+    },
+    {
+      "epoch": 14.654,
+      "grad_norm": 1.3861567974090576,
+      "learning_rate": 2e-05,
+      "loss": 0.04764885,
+      "step": 7327
+    },
+    {
+      "epoch": 14.656,
+      "grad_norm": 0.8934645056724548,
+      "learning_rate": 2e-05,
+      "loss": 0.02765589,
+      "step": 7328
+    },
+    {
+      "epoch": 14.658,
+      "grad_norm": 1.2780135869979858,
+      "learning_rate": 2e-05,
+      "loss": 0.03148212,
+      "step": 7329
+    },
+    {
+      "epoch": 14.66,
+      "grad_norm": 1.3805195093154907,
+      "learning_rate": 2e-05,
+      "loss": 0.04608523,
+      "step": 7330
+    },
+    {
+      "epoch": 14.662,
+      "grad_norm": 1.4747601747512817,
+      "learning_rate": 2e-05,
+      "loss": 0.0399397,
+      "step": 7331
+    },
+    {
+      "epoch": 14.664,
+      "grad_norm": 2.1372389793395996,
+      "learning_rate": 2e-05,
+      "loss": 0.05007102,
+      "step": 7332
+    },
+    {
+      "epoch": 14.666,
+      "grad_norm": 1.1365556716918945,
+      "learning_rate": 2e-05,
+      "loss": 0.0315646,
+      "step": 7333
+    },
+    {
+      "epoch": 14.668,
+      "grad_norm": 1.1190814971923828,
+      "learning_rate": 2e-05,
+      "loss": 0.03757735,
+      "step": 7334
+    },
+    {
+      "epoch": 14.67,
+      "grad_norm": 1.0525307655334473,
+      "learning_rate": 2e-05,
+      "loss": 0.02962991,
+      "step": 7335
+    },
+    {
+      "epoch": 14.672,
+      "grad_norm": 1.0830663442611694,
+      "learning_rate": 2e-05,
+      "loss": 0.03256768,
+      "step": 7336
+    },
+    {
+      "epoch": 14.674,
+      "grad_norm": 1.400888204574585,
+      "learning_rate": 2e-05,
+      "loss": 0.03843899,
+      "step": 7337
+    },
+    {
+      "epoch": 14.676,
+      "grad_norm": 1.4905816316604614,
+      "learning_rate": 2e-05,
+      "loss": 0.0448906,
+      "step": 7338
+    },
+    {
+      "epoch": 14.678,
+      "grad_norm": 1.2425509691238403,
+      "learning_rate": 2e-05,
+      "loss": 0.0399595,
+      "step": 7339
+    },
+    {
+      "epoch": 14.68,
+      "grad_norm": 1.1389684677124023,
+      "learning_rate": 2e-05,
+      "loss": 0.03919307,
+      "step": 7340
+    },
+    {
+      "epoch": 14.682,
+      "grad_norm": 1.3145670890808105,
+      "learning_rate": 2e-05,
+      "loss": 0.03573439,
+      "step": 7341
+    },
+    {
+      "epoch": 14.684,
+      "grad_norm": 1.6601872444152832,
+      "learning_rate": 2e-05,
+      "loss": 0.04692688,
+      "step": 7342
+    },
+    {
+      "epoch": 14.686,
+      "grad_norm": 1.6752896308898926,
+      "learning_rate": 2e-05,
+      "loss": 0.02998948,
+      "step": 7343
+    },
+    {
+      "epoch": 14.688,
+      "grad_norm": 1.4124311208724976,
+      "learning_rate": 2e-05,
+      "loss": 0.0381167,
+      "step": 7344
+    },
+    {
+      "epoch": 14.69,
+      "grad_norm": 1.7222316265106201,
+      "learning_rate": 2e-05,
+      "loss": 0.04596563,
+      "step": 7345
+    },
+    {
+      "epoch": 14.692,
+      "grad_norm": 1.5030279159545898,
+      "learning_rate": 2e-05,
+      "loss": 0.03170253,
+      "step": 7346
+    },
+    {
+      "epoch": 14.693999999999999,
+      "grad_norm": 1.9802442789077759,
+      "learning_rate": 2e-05,
+      "loss": 0.05240085,
+      "step": 7347
+    },
+    {
+      "epoch": 14.696,
+      "grad_norm": 1.9767365455627441,
+      "learning_rate": 2e-05,
+      "loss": 0.0302583,
+      "step": 7348
+    },
+    {
+      "epoch": 14.698,
+      "grad_norm": 1.5814634561538696,
+      "learning_rate": 2e-05,
+      "loss": 0.03863395,
+      "step": 7349
+    },
+    {
+      "epoch": 14.7,
+      "grad_norm": 0.9202812910079956,
+      "learning_rate": 2e-05,
+      "loss": 0.02484436,
+      "step": 7350
+    },
+    {
+      "epoch": 14.702,
+      "grad_norm": 1.5224653482437134,
+      "learning_rate": 2e-05,
+      "loss": 0.04415604,
+      "step": 7351
+    },
+    {
+      "epoch": 14.704,
+      "grad_norm": 1.4382812976837158,
+      "learning_rate": 2e-05,
+      "loss": 0.04574153,
+      "step": 7352
+    },
+    {
+      "epoch": 14.706,
+      "grad_norm": 1.1333105564117432,
+      "learning_rate": 2e-05,
+      "loss": 0.03003317,
+      "step": 7353
+    },
+    {
+      "epoch": 14.708,
+      "grad_norm": 1.1706465482711792,
+      "learning_rate": 2e-05,
+      "loss": 0.03788555,
+      "step": 7354
+    },
+    {
+      "epoch": 14.71,
+      "grad_norm": 2.07646107673645,
+      "learning_rate": 2e-05,
+      "loss": 0.04297365,
+      "step": 7355
+    },
+    {
+      "epoch": 14.712,
+      "grad_norm": 1.1796342134475708,
+      "learning_rate": 2e-05,
+      "loss": 0.02585716,
+      "step": 7356
+    },
+    {
+      "epoch": 14.714,
+      "grad_norm": 1.6022311449050903,
+      "learning_rate": 2e-05,
+      "loss": 0.03057653,
+      "step": 7357
+    },
+    {
+      "epoch": 14.716,
+      "grad_norm": 1.4186314344406128,
+      "learning_rate": 2e-05,
+      "loss": 0.04783071,
+      "step": 7358
+    },
+    {
+      "epoch": 14.718,
+      "grad_norm": 1.163080096244812,
+      "learning_rate": 2e-05,
+      "loss": 0.04012655,
+      "step": 7359
+    },
+    {
+      "epoch": 14.72,
+      "grad_norm": 1.7531554698944092,
+      "learning_rate": 2e-05,
+      "loss": 0.06195234,
+      "step": 7360
+    },
+    {
+      "epoch": 14.722,
+      "grad_norm": 0.8606005311012268,
+      "learning_rate": 2e-05,
+      "loss": 0.02377318,
+      "step": 7361
+    },
+    {
+      "epoch": 14.724,
+      "grad_norm": 2.0389902591705322,
+      "learning_rate": 2e-05,
+      "loss": 0.03832042,
+      "step": 7362
+    },
+    {
+      "epoch": 14.725999999999999,
+      "grad_norm": 2.013206720352173,
+      "learning_rate": 2e-05,
+      "loss": 0.03251911,
+      "step": 7363
+    },
+    {
+      "epoch": 14.728,
+      "grad_norm": 1.606460690498352,
+      "learning_rate": 2e-05,
+      "loss": 0.03653719,
+      "step": 7364
+    },
+    {
+      "epoch": 14.73,
+      "grad_norm": 0.8811898827552795,
+      "learning_rate": 2e-05,
+      "loss": 0.03039603,
+      "step": 7365
+    },
+    {
+      "epoch": 14.732,
+      "grad_norm": 1.03880774974823,
+      "learning_rate": 2e-05,
+      "loss": 0.0278089,
+      "step": 7366
+    },
+    {
+      "epoch": 14.734,
+      "grad_norm": 1.3265225887298584,
+      "learning_rate": 2e-05,
+      "loss": 0.03921036,
+      "step": 7367
+    },
+    {
+      "epoch": 14.736,
+      "grad_norm": 1.2806612253189087,
+      "learning_rate": 2e-05,
+      "loss": 0.04968328,
+      "step": 7368
+    },
+    {
+      "epoch": 14.738,
+      "grad_norm": 1.5343149900436401,
+      "learning_rate": 2e-05,
+      "loss": 0.04367102,
+      "step": 7369
+    },
+    {
+      "epoch": 14.74,
+      "grad_norm": 1.47340726852417,
+      "learning_rate": 2e-05,
+      "loss": 0.03140695,
+      "step": 7370
+    },
+    {
+      "epoch": 14.742,
+      "grad_norm": 1.4524576663970947,
+      "learning_rate": 2e-05,
+      "loss": 0.03347445,
+      "step": 7371
+    },
+    {
+      "epoch": 14.744,
+      "grad_norm": 2.033576250076294,
+      "learning_rate": 2e-05,
+      "loss": 0.04897091,
+      "step": 7372
+    },
+    {
+      "epoch": 14.746,
+      "grad_norm": 0.9620274305343628,
+      "learning_rate": 2e-05,
+      "loss": 0.02743071,
+      "step": 7373
+    },
+    {
+      "epoch": 14.748,
+      "grad_norm": 1.0374791622161865,
+      "learning_rate": 2e-05,
+      "loss": 0.04549268,
+      "step": 7374
+    },
+    {
+      "epoch": 14.75,
+      "grad_norm": 1.0183595418930054,
+      "learning_rate": 2e-05,
+      "loss": 0.02789947,
+      "step": 7375
+    },
+    {
+      "epoch": 14.752,
+      "grad_norm": 0.9441385269165039,
+      "learning_rate": 2e-05,
+      "loss": 0.03046933,
+      "step": 7376
+    },
+    {
+      "epoch": 14.754,
+      "grad_norm": 1.2050881385803223,
+      "learning_rate": 2e-05,
+      "loss": 0.04037102,
+      "step": 7377
+    },
+    {
+      "epoch": 14.756,
+      "grad_norm": 1.4334502220153809,
+      "learning_rate": 2e-05,
+      "loss": 0.04295122,
+      "step": 7378
+    },
+    {
+      "epoch": 14.758,
+      "grad_norm": 1.652840495109558,
+      "learning_rate": 2e-05,
+      "loss": 0.03737336,
+      "step": 7379
+    },
+    {
+      "epoch": 14.76,
+      "grad_norm": 1.331262469291687,
+      "learning_rate": 2e-05,
+      "loss": 0.04912826,
+      "step": 7380
+    },
+    {
+      "epoch": 14.762,
+      "grad_norm": 1.5204346179962158,
+      "learning_rate": 2e-05,
+      "loss": 0.04136185,
+      "step": 7381
+    },
+    {
+      "epoch": 14.764,
+      "grad_norm": 1.3094161748886108,
+      "learning_rate": 2e-05,
+      "loss": 0.04010442,
+      "step": 7382
+    },
+    {
+      "epoch": 14.766,
+      "grad_norm": 1.716664433479309,
+      "learning_rate": 2e-05,
+      "loss": 0.04056903,
+      "step": 7383
+    },
+    {
+      "epoch": 14.768,
+      "grad_norm": 2.7664754390716553,
+      "learning_rate": 2e-05,
+      "loss": 0.04453249,
+      "step": 7384
+    },
+    {
+      "epoch": 14.77,
+      "grad_norm": 1.1787066459655762,
+      "learning_rate": 2e-05,
+      "loss": 0.04342939,
+      "step": 7385
+    },
+    {
+      "epoch": 14.772,
+      "grad_norm": 1.3815988302230835,
+      "learning_rate": 2e-05,
+      "loss": 0.04383612,
+      "step": 7386
+    },
+    {
+      "epoch": 14.774000000000001,
+      "grad_norm": 1.0485109090805054,
+      "learning_rate": 2e-05,
+      "loss": 0.03212638,
+      "step": 7387
+    },
+    {
+      "epoch": 14.776,
+      "grad_norm": 1.444206714630127,
+      "learning_rate": 2e-05,
+      "loss": 0.04447151,
+      "step": 7388
+    },
+    {
+      "epoch": 14.778,
+      "grad_norm": 0.9816635847091675,
+      "learning_rate": 2e-05,
+      "loss": 0.03149556,
+      "step": 7389
+    },
+    {
+      "epoch": 14.78,
+      "grad_norm": 1.0881997346878052,
+      "learning_rate": 2e-05,
+      "loss": 0.03420866,
+      "step": 7390
+    },
+    {
+      "epoch": 14.782,
+      "grad_norm": 1.6856642961502075,
+      "learning_rate": 2e-05,
+      "loss": 0.04961979,
+      "step": 7391
+    },
+    {
+      "epoch": 14.784,
+      "grad_norm": 4.939731597900391,
+      "learning_rate": 2e-05,
+      "loss": 0.05440149,
+      "step": 7392
+    },
+    {
+      "epoch": 14.786,
+      "grad_norm": 1.1305186748504639,
+      "learning_rate": 2e-05,
+      "loss": 0.03616676,
+      "step": 7393
+    },
+    {
+      "epoch": 14.788,
+      "grad_norm": 1.8306798934936523,
+      "learning_rate": 2e-05,
+      "loss": 0.04979054,
+      "step": 7394
+    },
+    {
+      "epoch": 14.79,
+      "grad_norm": 1.601699948310852,
+      "learning_rate": 2e-05,
+      "loss": 0.0460626,
+      "step": 7395
+    },
+    {
+      "epoch": 14.792,
+      "grad_norm": 1.8894065618515015,
+      "learning_rate": 2e-05,
+      "loss": 0.05282786,
+      "step": 7396
+    },
+    {
+      "epoch": 14.794,
+      "grad_norm": 1.5035409927368164,
+      "learning_rate": 2e-05,
+      "loss": 0.03724657,
+      "step": 7397
+    },
+    {
+      "epoch": 14.796,
+      "grad_norm": 2.4653127193450928,
+      "learning_rate": 2e-05,
+      "loss": 0.03604883,
+      "step": 7398
+    },
+    {
+      "epoch": 14.798,
+      "grad_norm": 1.3967053890228271,
+      "learning_rate": 2e-05,
+      "loss": 0.04081305,
+      "step": 7399
+    },
+    {
+      "epoch": 14.8,
+      "grad_norm": 1.280266523361206,
+      "learning_rate": 2e-05,
+      "loss": 0.03488631,
+      "step": 7400
+    },
+    {
+      "epoch": 14.802,
+      "grad_norm": 1.1659793853759766,
+      "learning_rate": 2e-05,
+      "loss": 0.03118264,
+      "step": 7401
+    },
+    {
+      "epoch": 14.804,
+      "grad_norm": 1.2721163034439087,
+      "learning_rate": 2e-05,
+      "loss": 0.0195914,
+      "step": 7402
+    },
+    {
+      "epoch": 14.806000000000001,
+      "grad_norm": 0.9028842449188232,
+      "learning_rate": 2e-05,
+      "loss": 0.02684617,
+      "step": 7403
+    },
+    {
+      "epoch": 14.808,
+      "grad_norm": 1.3119491338729858,
+      "learning_rate": 2e-05,
+      "loss": 0.0373154,
+      "step": 7404
+    },
+    {
+      "epoch": 14.81,
+      "grad_norm": 2.559068202972412,
+      "learning_rate": 2e-05,
+      "loss": 0.03095316,
+      "step": 7405
+    },
+    {
+      "epoch": 14.812,
+      "grad_norm": 1.241502046585083,
+      "learning_rate": 2e-05,
+      "loss": 0.02786666,
+      "step": 7406
+    },
+    {
+      "epoch": 14.814,
+      "grad_norm": 1.2434546947479248,
+      "learning_rate": 2e-05,
+      "loss": 0.03359643,
+      "step": 7407
+    },
+    {
+      "epoch": 14.816,
+      "grad_norm": 1.3161190748214722,
+      "learning_rate": 2e-05,
+      "loss": 0.03395464,
+      "step": 7408
+    },
+    {
+      "epoch": 14.818,
+      "grad_norm": 1.1681431531906128,
+      "learning_rate": 2e-05,
+      "loss": 0.03358738,
+      "step": 7409
+    },
+    {
+      "epoch": 14.82,
+      "grad_norm": 2.086354970932007,
+      "learning_rate": 2e-05,
+      "loss": 0.0546408,
+      "step": 7410
+    },
+    {
+      "epoch": 14.822,
+      "grad_norm": 1.3456135988235474,
+      "learning_rate": 2e-05,
+      "loss": 0.04479208,
+      "step": 7411
+    },
+    {
+      "epoch": 14.824,
+      "grad_norm": 1.053308367729187,
+      "learning_rate": 2e-05,
+      "loss": 0.02650757,
+      "step": 7412
+    },
+    {
+      "epoch": 14.826,
+      "grad_norm": 1.4386178255081177,
+      "learning_rate": 2e-05,
+      "loss": 0.03358426,
+      "step": 7413
+    },
+    {
+      "epoch": 14.828,
+      "grad_norm": 1.846922755241394,
+      "learning_rate": 2e-05,
+      "loss": 0.037901,
+      "step": 7414
+    },
+    {
+      "epoch": 14.83,
+      "grad_norm": 1.393807053565979,
+      "learning_rate": 2e-05,
+      "loss": 0.05257457,
+      "step": 7415
+    },
+    {
+      "epoch": 14.832,
+      "grad_norm": 1.5299506187438965,
+      "learning_rate": 2e-05,
+      "loss": 0.04238684,
+      "step": 7416
+    },
+    {
+      "epoch": 14.834,
+      "grad_norm": 1.1853361129760742,
+      "learning_rate": 2e-05,
+      "loss": 0.04620481,
+      "step": 7417
+    },
+    {
+      "epoch": 14.836,
+      "grad_norm": 1.9609429836273193,
+      "learning_rate": 2e-05,
+      "loss": 0.05831449,
+      "step": 7418
+    },
+    {
+      "epoch": 14.838,
+      "grad_norm": 0.9670793414115906,
+      "learning_rate": 2e-05,
+      "loss": 0.0317945,
+      "step": 7419
+    },
+    {
+      "epoch": 14.84,
+      "grad_norm": 2.3657424449920654,
+      "learning_rate": 2e-05,
+      "loss": 0.06117197,
+      "step": 7420
+    },
+    {
+      "epoch": 14.842,
+      "grad_norm": 1.2410660982131958,
+      "learning_rate": 2e-05,
+      "loss": 0.03587608,
+      "step": 7421
+    },
+    {
+      "epoch": 14.844,
+      "grad_norm": 1.5402605533599854,
+      "learning_rate": 2e-05,
+      "loss": 0.03572353,
+      "step": 7422
+    },
+    {
+      "epoch": 14.846,
+      "grad_norm": 1.4933197498321533,
+      "learning_rate": 2e-05,
+      "loss": 0.03981678,
+      "step": 7423
+    },
+    {
+      "epoch": 14.848,
+      "grad_norm": 1.4539066553115845,
+      "learning_rate": 2e-05,
+      "loss": 0.05267172,
+      "step": 7424
+    },
+    {
+      "epoch": 14.85,
+      "grad_norm": 1.239725947380066,
+      "learning_rate": 2e-05,
+      "loss": 0.04814597,
+      "step": 7425
+    },
+    {
+      "epoch": 14.852,
+      "grad_norm": 1.062360405921936,
+      "learning_rate": 2e-05,
+      "loss": 0.03394141,
+      "step": 7426
+    },
+    {
+      "epoch": 14.854,
+      "grad_norm": 1.3617041110992432,
+      "learning_rate": 2e-05,
+      "loss": 0.04762523,
+      "step": 7427
+    },
+    {
+      "epoch": 14.856,
+      "grad_norm": 1.218767762184143,
+      "learning_rate": 2e-05,
+      "loss": 0.0463365,
+      "step": 7428
+    },
+    {
+      "epoch": 14.858,
+      "grad_norm": 1.1264668703079224,
+      "learning_rate": 2e-05,
+      "loss": 0.02673528,
+      "step": 7429
+    },
+    {
+      "epoch": 14.86,
+      "grad_norm": 1.9460324048995972,
+      "learning_rate": 2e-05,
+      "loss": 0.03690345,
+      "step": 7430
+    },
+    {
+      "epoch": 14.862,
+      "grad_norm": 1.3760199546813965,
+      "learning_rate": 2e-05,
+      "loss": 0.04101796,
+      "step": 7431
+    },
+    {
+      "epoch": 14.864,
+      "grad_norm": 1.8361059427261353,
+      "learning_rate": 2e-05,
+      "loss": 0.05019148,
+      "step": 7432
+    },
+    {
+      "epoch": 14.866,
+      "grad_norm": 2.861609697341919,
+      "learning_rate": 2e-05,
+      "loss": 0.04384901,
+      "step": 7433
+    },
+    {
+      "epoch": 14.868,
+      "grad_norm": 2.696000814437866,
+      "learning_rate": 2e-05,
+      "loss": 0.04344347,
+      "step": 7434
+    },
+    {
+      "epoch": 14.87,
+      "grad_norm": 1.0894790887832642,
+      "learning_rate": 2e-05,
+      "loss": 0.03905128,
+      "step": 7435
+    },
+    {
+      "epoch": 14.872,
+      "grad_norm": 1.1012567281723022,
+      "learning_rate": 2e-05,
+      "loss": 0.0368652,
+      "step": 7436
+    },
+    {
+      "epoch": 14.874,
+      "grad_norm": 1.441910982131958,
+      "learning_rate": 2e-05,
+      "loss": 0.04437996,
+      "step": 7437
+    },
+    {
+      "epoch": 14.876,
+      "grad_norm": 1.305418610572815,
+      "learning_rate": 2e-05,
+      "loss": 0.03403091,
+      "step": 7438
+    },
+    {
+      "epoch": 14.878,
+      "grad_norm": 0.936972975730896,
+      "learning_rate": 2e-05,
+      "loss": 0.03193845,
+      "step": 7439
+    },
+    {
+      "epoch": 14.88,
+      "grad_norm": 1.2693448066711426,
+      "learning_rate": 2e-05,
+      "loss": 0.03593332,
+      "step": 7440
+    },
+    {
+      "epoch": 14.882,
+      "grad_norm": 0.9230713844299316,
+      "learning_rate": 2e-05,
+      "loss": 0.02694279,
+      "step": 7441
+    },
+    {
+      "epoch": 14.884,
+      "grad_norm": 1.4311070442199707,
+      "learning_rate": 2e-05,
+      "loss": 0.03495328,
+      "step": 7442
+    },
+    {
+      "epoch": 14.886,
+      "grad_norm": 2.3351352214813232,
+      "learning_rate": 2e-05,
+      "loss": 0.04533555,
+      "step": 7443
+    },
+    {
+      "epoch": 14.888,
+      "grad_norm": 1.6613550186157227,
+      "learning_rate": 2e-05,
+      "loss": 0.0545169,
+      "step": 7444
+    },
+    {
+      "epoch": 14.89,
+      "grad_norm": 3.134610414505005,
+      "learning_rate": 2e-05,
+      "loss": 0.06269054,
+      "step": 7445
+    },
+    {
+      "epoch": 14.892,
+      "grad_norm": 1.2900604009628296,
+      "learning_rate": 2e-05,
+      "loss": 0.0403505,
+      "step": 7446
+    },
+    {
+      "epoch": 14.894,
+      "grad_norm": 1.0350453853607178,
+      "learning_rate": 2e-05,
+      "loss": 0.02884517,
+      "step": 7447
+    },
+    {
+      "epoch": 14.896,
+      "grad_norm": 2.1216492652893066,
+      "learning_rate": 2e-05,
+      "loss": 0.05400493,
+      "step": 7448
+    },
+    {
+      "epoch": 14.898,
+      "grad_norm": 1.5325112342834473,
+      "learning_rate": 2e-05,
+      "loss": 0.03483828,
+      "step": 7449
+    },
+    {
+      "epoch": 14.9,
+      "grad_norm": 1.1589800119400024,
+      "learning_rate": 2e-05,
+      "loss": 0.03094343,
+      "step": 7450
+    },
+    {
+      "epoch": 14.902,
+      "grad_norm": 1.6227322816848755,
+      "learning_rate": 2e-05,
+      "loss": 0.0469828,
+      "step": 7451
+    },
+    {
+      "epoch": 14.904,
+      "grad_norm": 1.6878936290740967,
+      "learning_rate": 2e-05,
+      "loss": 0.04225905,
+      "step": 7452
+    },
+    {
+      "epoch": 14.906,
+      "grad_norm": 1.1886628866195679,
+      "learning_rate": 2e-05,
+      "loss": 0.03824829,
+      "step": 7453
+    },
+    {
+      "epoch": 14.908,
+      "grad_norm": 1.4336858987808228,
+      "learning_rate": 2e-05,
+      "loss": 0.04599863,
+      "step": 7454
+    },
+    {
+      "epoch": 14.91,
+      "grad_norm": 1.2423670291900635,
+      "learning_rate": 2e-05,
+      "loss": 0.05265063,
+      "step": 7455
+    },
+    {
+      "epoch": 14.912,
+      "grad_norm": 1.2401492595672607,
+      "learning_rate": 2e-05,
+      "loss": 0.03636886,
+      "step": 7456
+    },
+    {
+      "epoch": 14.914,
+      "grad_norm": 0.7528156638145447,
+      "learning_rate": 2e-05,
+      "loss": 0.02554782,
+      "step": 7457
+    },
+    {
+      "epoch": 14.916,
+      "grad_norm": 1.5916630029678345,
+      "learning_rate": 2e-05,
+      "loss": 0.03105753,
+      "step": 7458
+    },
+    {
+      "epoch": 14.918,
+      "grad_norm": 1.8795676231384277,
+      "learning_rate": 2e-05,
+      "loss": 0.03935018,
+      "step": 7459
+    },
+    {
+      "epoch": 14.92,
+      "grad_norm": 1.8167335987091064,
+      "learning_rate": 2e-05,
+      "loss": 0.04380289,
+      "step": 7460
+    },
+    {
+      "epoch": 14.922,
+      "grad_norm": 1.1691914796829224,
+      "learning_rate": 2e-05,
+      "loss": 0.03274843,
+      "step": 7461
+    },
+    {
+      "epoch": 14.924,
+      "grad_norm": 1.3650342226028442,
+      "learning_rate": 2e-05,
+      "loss": 0.04208003,
+      "step": 7462
+    },
+    {
+      "epoch": 14.926,
+      "grad_norm": 1.24591863155365,
+      "learning_rate": 2e-05,
+      "loss": 0.04226547,
+      "step": 7463
+    },
+    {
+      "epoch": 14.928,
+      "grad_norm": 1.3983882665634155,
+      "learning_rate": 2e-05,
+      "loss": 0.05762651,
+      "step": 7464
+    },
+    {
+      "epoch": 14.93,
+      "grad_norm": 1.1452008485794067,
+      "learning_rate": 2e-05,
+      "loss": 0.02998473,
+      "step": 7465
+    },
+    {
+      "epoch": 14.932,
+      "grad_norm": 0.9939238429069519,
+      "learning_rate": 2e-05,
+      "loss": 0.02803001,
+      "step": 7466
+    },
+    {
+      "epoch": 14.934,
+      "grad_norm": 1.5362166166305542,
+      "learning_rate": 2e-05,
+      "loss": 0.05231043,
+      "step": 7467
+    },
+    {
+      "epoch": 14.936,
+      "grad_norm": 1.7269541025161743,
+      "learning_rate": 2e-05,
+      "loss": 0.04780792,
+      "step": 7468
+    },
+    {
+      "epoch": 14.938,
+      "grad_norm": 1.1512809991836548,
+      "learning_rate": 2e-05,
+      "loss": 0.03671274,
+      "step": 7469
+    },
+    {
+      "epoch": 14.94,
+      "grad_norm": 1.2893776893615723,
+      "learning_rate": 2e-05,
+      "loss": 0.03543502,
+      "step": 7470
+    },
+    {
+      "epoch": 14.942,
+      "grad_norm": 2.2255067825317383,
+      "learning_rate": 2e-05,
+      "loss": 0.05238653,
+      "step": 7471
+    },
+    {
+      "epoch": 14.943999999999999,
+      "grad_norm": 0.9970213770866394,
+      "learning_rate": 2e-05,
+      "loss": 0.02268671,
+      "step": 7472
+    },
+    {
+      "epoch": 14.946,
+      "grad_norm": 1.0443769693374634,
+      "learning_rate": 2e-05,
+      "loss": 0.03668858,
+      "step": 7473
+    },
+    {
+      "epoch": 14.948,
+      "grad_norm": 0.9335085153579712,
+      "learning_rate": 2e-05,
+      "loss": 0.03486544,
+      "step": 7474
+    },
+    {
+      "epoch": 14.95,
+      "grad_norm": 2.1096580028533936,
+      "learning_rate": 2e-05,
+      "loss": 0.06607223,
+      "step": 7475
+    },
+    {
+      "epoch": 14.952,
+      "grad_norm": 2.1149537563323975,
+      "learning_rate": 2e-05,
+      "loss": 0.03609411,
+      "step": 7476
+    },
+    {
+      "epoch": 14.954,
+      "grad_norm": 1.5304067134857178,
+      "learning_rate": 2e-05,
+      "loss": 0.03244342,
+      "step": 7477
+    },
+    {
+      "epoch": 14.956,
+      "grad_norm": 1.4157954454421997,
+      "learning_rate": 2e-05,
+      "loss": 0.04437434,
+      "step": 7478
+    },
+    {
+      "epoch": 14.958,
+      "grad_norm": 1.3668354749679565,
+      "learning_rate": 2e-05,
+      "loss": 0.03368328,
+      "step": 7479
+    },
+    {
+      "epoch": 14.96,
+      "grad_norm": 1.1357492208480835,
+      "learning_rate": 2e-05,
+      "loss": 0.03183534,
+      "step": 7480
+    },
+    {
+      "epoch": 14.962,
+      "grad_norm": 1.2593474388122559,
+      "learning_rate": 2e-05,
+      "loss": 0.03054814,
+      "step": 7481
+    },
+    {
+      "epoch": 14.964,
+      "grad_norm": 1.0651172399520874,
+      "learning_rate": 2e-05,
+      "loss": 0.03272149,
+      "step": 7482
+    },
+    {
+      "epoch": 14.966,
+      "grad_norm": 1.3296829462051392,
+      "learning_rate": 2e-05,
+      "loss": 0.02990423,
+      "step": 7483
+    },
+    {
+      "epoch": 14.968,
+      "grad_norm": 1.153295874595642,
+      "learning_rate": 2e-05,
+      "loss": 0.04132929,
+      "step": 7484
+    },
+    {
+      "epoch": 14.97,
+      "grad_norm": 1.8791780471801758,
+      "learning_rate": 2e-05,
+      "loss": 0.0307649,
+      "step": 7485
+    },
+    {
+      "epoch": 14.972,
+      "grad_norm": 1.850902795791626,
+      "learning_rate": 2e-05,
+      "loss": 0.05484937,
+      "step": 7486
+    },
+    {
+      "epoch": 14.974,
+      "grad_norm": 1.170993685722351,
+      "learning_rate": 2e-05,
+      "loss": 0.03880947,
+      "step": 7487
+    },
+    {
+      "epoch": 14.975999999999999,
+      "grad_norm": 1.0112059116363525,
+      "learning_rate": 2e-05,
+      "loss": 0.05152441,
+      "step": 7488
+    },
+    {
+      "epoch": 14.978,
+      "grad_norm": 0.9202816486358643,
+      "learning_rate": 2e-05,
+      "loss": 0.03082973,
+      "step": 7489
+    },
+    {
+      "epoch": 14.98,
+      "grad_norm": 1.5370155572891235,
+      "learning_rate": 2e-05,
+      "loss": 0.03976063,
+      "step": 7490
+    },
+    {
+      "epoch": 14.982,
+      "grad_norm": 1.6214145421981812,
+      "learning_rate": 2e-05,
+      "loss": 0.03335922,
+      "step": 7491
+    },
+    {
+      "epoch": 14.984,
+      "grad_norm": 1.5763835906982422,
+      "learning_rate": 2e-05,
+      "loss": 0.05680116,
+      "step": 7492
+    },
+    {
+      "epoch": 14.986,
+      "grad_norm": 1.2244529724121094,
+      "learning_rate": 2e-05,
+      "loss": 0.03010012,
+      "step": 7493
+    },
+    {
+      "epoch": 14.988,
+      "grad_norm": 1.0402621030807495,
+      "learning_rate": 2e-05,
+      "loss": 0.03202302,
+      "step": 7494
+    },
+    {
+      "epoch": 14.99,
+      "grad_norm": 1.1672792434692383,
+      "learning_rate": 2e-05,
+      "loss": 0.0357578,
+      "step": 7495
+    },
+    {
+      "epoch": 14.992,
+      "grad_norm": 1.5200659036636353,
+      "learning_rate": 2e-05,
+      "loss": 0.04126337,
+      "step": 7496
+    },
+    {
+      "epoch": 14.994,
+      "grad_norm": 3.4498698711395264,
+      "learning_rate": 2e-05,
+      "loss": 0.06399542,
+      "step": 7497
+    },
+    {
+      "epoch": 14.996,
+      "grad_norm": 2.6905369758605957,
+      "learning_rate": 2e-05,
+      "loss": 0.05302805,
+      "step": 7498
+    },
+    {
+      "epoch": 14.998,
+      "grad_norm": 1.3441663980484009,
+      "learning_rate": 2e-05,
+      "loss": 0.03965209,
+      "step": 7499
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 1.0610202550888062,
+      "learning_rate": 2e-05,
+      "loss": 0.03335898,
+      "step": 7500
+    },
+    {
+      "epoch": 15.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.982,
+        "AngleClassification_2": 0.998,
+        "AngleClassification_3": 0.9600798403193613,
+        "Equal_1": 0.994,
+        "Equal_2": 0.9640718562874252,
+        "Equal_3": 0.8982035928143712,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9960079840319361,
+        "Parallel_1": 0.9939879759519038,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.99,
+        "Perpendicular_1": 0.992,
+        "Perpendicular_2": 0.972,
+        "Perpendicular_3": 0.7114228456913828,
+        "PointLiesOnCircle_1": 0.9939879759519038,
+        "PointLiesOnCircle_2": 0.996,
+        "PointLiesOnCircle_3": 0.99,
+        "PointLiesOnLine_1": 0.9959919839679359,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9860279441117764
+      },
+      "eval_runtime": 320.0741,
+      "eval_samples_per_second": 32.805,
+      "eval_steps_per_second": 0.656,
+      "step": 7500
+    },
+    {
+      "epoch": 15.002,
+      "grad_norm": 1.5323970317840576,
+      "learning_rate": 2e-05,
+      "loss": 0.03357163,
+      "step": 7501
+    },
+    {
+      "epoch": 15.004,
+      "grad_norm": 1.106204628944397,
+      "learning_rate": 2e-05,
+      "loss": 0.02693918,
+      "step": 7502
+    },
+    {
+      "epoch": 15.006,
+      "grad_norm": 2.7281370162963867,
+      "learning_rate": 2e-05,
+      "loss": 0.07801263,
+      "step": 7503
+    },
+    {
+      "epoch": 15.008,
+      "grad_norm": 0.9546897411346436,
+      "learning_rate": 2e-05,
+      "loss": 0.0307308,
+      "step": 7504
+    },
+    {
+      "epoch": 15.01,
+      "grad_norm": 1.819274663925171,
+      "learning_rate": 2e-05,
+      "loss": 0.02562927,
+      "step": 7505
+    },
+    {
+      "epoch": 15.012,
+      "grad_norm": 2.0995659828186035,
+      "learning_rate": 2e-05,
+      "loss": 0.05087785,
+      "step": 7506
+    },
+    {
+      "epoch": 15.014,
+      "grad_norm": 1.5401643514633179,
+      "learning_rate": 2e-05,
+      "loss": 0.05163966,
+      "step": 7507
+    },
+    {
+      "epoch": 15.016,
+      "grad_norm": 3.4083080291748047,
+      "learning_rate": 2e-05,
+      "loss": 0.04428158,
+      "step": 7508
+    },
+    {
+      "epoch": 15.018,
+      "grad_norm": 3.521087646484375,
+      "learning_rate": 2e-05,
+      "loss": 0.04372428,
+      "step": 7509
+    },
+    {
+      "epoch": 15.02,
+      "grad_norm": 1.4367711544036865,
+      "learning_rate": 2e-05,
+      "loss": 0.0497531,
+      "step": 7510
+    },
+    {
+      "epoch": 15.022,
+      "grad_norm": 1.270221471786499,
+      "learning_rate": 2e-05,
+      "loss": 0.03810829,
+      "step": 7511
+    },
+    {
+      "epoch": 15.024,
+      "grad_norm": 1.317637324333191,
+      "learning_rate": 2e-05,
+      "loss": 0.0325052,
+      "step": 7512
+    },
+    {
+      "epoch": 15.026,
+      "grad_norm": 1.2471699714660645,
+      "learning_rate": 2e-05,
+      "loss": 0.02949313,
+      "step": 7513
+    },
+    {
+      "epoch": 15.028,
+      "grad_norm": 1.197861671447754,
+      "learning_rate": 2e-05,
+      "loss": 0.03684509,
+      "step": 7514
+    },
+    {
+      "epoch": 15.03,
+      "grad_norm": 1.8183475732803345,
+      "learning_rate": 2e-05,
+      "loss": 0.06737492,
+      "step": 7515
+    },
+    {
+      "epoch": 15.032,
+      "grad_norm": 1.03377366065979,
+      "learning_rate": 2e-05,
+      "loss": 0.03917066,
+      "step": 7516
+    },
+    {
+      "epoch": 15.034,
+      "grad_norm": 2.038936138153076,
+      "learning_rate": 2e-05,
+      "loss": 0.03911228,
+      "step": 7517
+    },
+    {
+      "epoch": 15.036,
+      "grad_norm": 0.945249617099762,
+      "learning_rate": 2e-05,
+      "loss": 0.02501803,
+      "step": 7518
+    },
+    {
+      "epoch": 15.038,
+      "grad_norm": 3.087181568145752,
+      "learning_rate": 2e-05,
+      "loss": 0.0438556,
+      "step": 7519
+    },
+    {
+      "epoch": 15.04,
+      "grad_norm": 1.3081676959991455,
+      "learning_rate": 2e-05,
+      "loss": 0.04287355,
+      "step": 7520
+    },
+    {
+      "epoch": 15.042,
+      "grad_norm": 1.7691004276275635,
+      "learning_rate": 2e-05,
+      "loss": 0.03438368,
+      "step": 7521
+    },
+    {
+      "epoch": 15.044,
+      "grad_norm": 1.5667086839675903,
+      "learning_rate": 2e-05,
+      "loss": 0.02922621,
+      "step": 7522
+    },
+    {
+      "epoch": 15.046,
+      "grad_norm": 2.0105979442596436,
+      "learning_rate": 2e-05,
+      "loss": 0.04000949,
+      "step": 7523
+    },
+    {
+      "epoch": 15.048,
+      "grad_norm": 1.2400439977645874,
+      "learning_rate": 2e-05,
+      "loss": 0.0464882,
+      "step": 7524
+    },
+    {
+      "epoch": 15.05,
+      "grad_norm": 1.4981132745742798,
+      "learning_rate": 2e-05,
+      "loss": 0.0353303,
+      "step": 7525
+    },
+    {
+      "epoch": 15.052,
+      "grad_norm": 1.7976100444793701,
+      "learning_rate": 2e-05,
+      "loss": 0.04217862,
+      "step": 7526
+    },
+    {
+      "epoch": 15.054,
+      "grad_norm": 1.2558151483535767,
+      "learning_rate": 2e-05,
+      "loss": 0.03909191,
+      "step": 7527
+    },
+    {
+      "epoch": 15.056,
+      "grad_norm": 0.9325153231620789,
+      "learning_rate": 2e-05,
+      "loss": 0.03274619,
+      "step": 7528
+    },
+    {
+      "epoch": 15.058,
+      "grad_norm": 1.0450353622436523,
+      "learning_rate": 2e-05,
+      "loss": 0.03170678,
+      "step": 7529
+    },
+    {
+      "epoch": 15.06,
+      "grad_norm": 2.4308180809020996,
+      "learning_rate": 2e-05,
+      "loss": 0.04601308,
+      "step": 7530
+    },
+    {
+      "epoch": 15.062,
+      "grad_norm": 1.7169846296310425,
+      "learning_rate": 2e-05,
+      "loss": 0.03847077,
+      "step": 7531
+    },
+    {
+      "epoch": 15.064,
+      "grad_norm": 1.9664361476898193,
+      "learning_rate": 2e-05,
+      "loss": 0.06695174,
+      "step": 7532
+    },
+    {
+      "epoch": 15.066,
+      "grad_norm": 1.3882360458374023,
+      "learning_rate": 2e-05,
+      "loss": 0.03631517,
+      "step": 7533
+    },
+    {
+      "epoch": 15.068,
+      "grad_norm": 1.6631497144699097,
+      "learning_rate": 2e-05,
+      "loss": 0.03166562,
+      "step": 7534
+    },
+    {
+      "epoch": 15.07,
+      "grad_norm": 1.215599536895752,
+      "learning_rate": 2e-05,
+      "loss": 0.04256611,
+      "step": 7535
+    },
+    {
+      "epoch": 15.072,
+      "grad_norm": 1.148917555809021,
+      "learning_rate": 2e-05,
+      "loss": 0.04094999,
+      "step": 7536
+    },
+    {
+      "epoch": 15.074,
+      "grad_norm": 1.262528419494629,
+      "learning_rate": 2e-05,
+      "loss": 0.02669748,
+      "step": 7537
+    },
+    {
+      "epoch": 15.076,
+      "grad_norm": 1.4280608892440796,
+      "learning_rate": 2e-05,
+      "loss": 0.03824629,
+      "step": 7538
+    },
+    {
+      "epoch": 15.078,
+      "grad_norm": 1.1914440393447876,
+      "learning_rate": 2e-05,
+      "loss": 0.03420604,
+      "step": 7539
+    },
+    {
+      "epoch": 15.08,
+      "grad_norm": 1.5418035984039307,
+      "learning_rate": 2e-05,
+      "loss": 0.04447289,
+      "step": 7540
+    },
+    {
+      "epoch": 15.082,
+      "grad_norm": 1.7850476503372192,
+      "learning_rate": 2e-05,
+      "loss": 0.05204378,
+      "step": 7541
+    },
+    {
+      "epoch": 15.084,
+      "grad_norm": 1.7155487537384033,
+      "learning_rate": 2e-05,
+      "loss": 0.04036555,
+      "step": 7542
+    },
+    {
+      "epoch": 15.086,
+      "grad_norm": 1.1043587923049927,
+      "learning_rate": 2e-05,
+      "loss": 0.03954956,
+      "step": 7543
+    },
+    {
+      "epoch": 15.088,
+      "grad_norm": 1.4283198118209839,
+      "learning_rate": 2e-05,
+      "loss": 0.04688943,
+      "step": 7544
+    },
+    {
+      "epoch": 15.09,
+      "grad_norm": 1.2153453826904297,
+      "learning_rate": 2e-05,
+      "loss": 0.03587115,
+      "step": 7545
+    },
+    {
+      "epoch": 15.092,
+      "grad_norm": 1.3699722290039062,
+      "learning_rate": 2e-05,
+      "loss": 0.04313603,
+      "step": 7546
+    },
+    {
+      "epoch": 15.094,
+      "grad_norm": 1.0918267965316772,
+      "learning_rate": 2e-05,
+      "loss": 0.03959726,
+      "step": 7547
+    },
+    {
+      "epoch": 15.096,
+      "grad_norm": 1.2050516605377197,
+      "learning_rate": 2e-05,
+      "loss": 0.03665368,
+      "step": 7548
+    },
+    {
+      "epoch": 15.098,
+      "grad_norm": 1.611580729484558,
+      "learning_rate": 2e-05,
+      "loss": 0.04823206,
+      "step": 7549
+    },
+    {
+      "epoch": 15.1,
+      "grad_norm": 1.3868486881256104,
+      "learning_rate": 2e-05,
+      "loss": 0.04774088,
+      "step": 7550
+    },
+    {
+      "epoch": 15.102,
+      "grad_norm": 2.077451229095459,
+      "learning_rate": 2e-05,
+      "loss": 0.03221234,
+      "step": 7551
+    },
+    {
+      "epoch": 15.104,
+      "grad_norm": 1.458793044090271,
+      "learning_rate": 2e-05,
+      "loss": 0.02752698,
+      "step": 7552
+    },
+    {
+      "epoch": 15.106,
+      "grad_norm": 1.0230228900909424,
+      "learning_rate": 2e-05,
+      "loss": 0.03414164,
+      "step": 7553
+    },
+    {
+      "epoch": 15.108,
+      "grad_norm": 1.8648970127105713,
+      "learning_rate": 2e-05,
+      "loss": 0.04431803,
+      "step": 7554
+    },
+    {
+      "epoch": 15.11,
+      "grad_norm": 1.213136076927185,
+      "learning_rate": 2e-05,
+      "loss": 0.03528218,
+      "step": 7555
+    },
+    {
+      "epoch": 15.112,
+      "grad_norm": 1.1067938804626465,
+      "learning_rate": 2e-05,
+      "loss": 0.03371948,
+      "step": 7556
+    },
+    {
+      "epoch": 15.114,
+      "grad_norm": 1.099187970161438,
+      "learning_rate": 2e-05,
+      "loss": 0.04380881,
+      "step": 7557
+    },
+    {
+      "epoch": 15.116,
+      "grad_norm": 1.4904146194458008,
+      "learning_rate": 2e-05,
+      "loss": 0.04857348,
+      "step": 7558
+    },
+    {
+      "epoch": 15.118,
+      "grad_norm": 1.9208067655563354,
+      "learning_rate": 2e-05,
+      "loss": 0.02925631,
+      "step": 7559
+    },
+    {
+      "epoch": 15.12,
+      "grad_norm": 1.3808106184005737,
+      "learning_rate": 2e-05,
+      "loss": 0.05164766,
+      "step": 7560
+    },
+    {
+      "epoch": 15.122,
+      "grad_norm": 1.7450789213180542,
+      "learning_rate": 2e-05,
+      "loss": 0.05939767,
+      "step": 7561
+    },
+    {
+      "epoch": 15.124,
+      "grad_norm": 1.3187388181686401,
+      "learning_rate": 2e-05,
+      "loss": 0.04064866,
+      "step": 7562
+    },
+    {
+      "epoch": 15.126,
+      "grad_norm": 2.158761501312256,
+      "learning_rate": 2e-05,
+      "loss": 0.03640309,
+      "step": 7563
+    },
+    {
+      "epoch": 15.128,
+      "grad_norm": 1.668043613433838,
+      "learning_rate": 2e-05,
+      "loss": 0.03924328,
+      "step": 7564
+    },
+    {
+      "epoch": 15.13,
+      "grad_norm": 1.2158105373382568,
+      "learning_rate": 2e-05,
+      "loss": 0.03478661,
+      "step": 7565
+    },
+    {
+      "epoch": 15.132,
+      "grad_norm": 1.681146264076233,
+      "learning_rate": 2e-05,
+      "loss": 0.05600887,
+      "step": 7566
+    },
+    {
+      "epoch": 15.134,
+      "grad_norm": 0.9708520770072937,
+      "learning_rate": 2e-05,
+      "loss": 0.01993345,
+      "step": 7567
+    },
+    {
+      "epoch": 15.136,
+      "grad_norm": 0.926152765750885,
+      "learning_rate": 2e-05,
+      "loss": 0.03074206,
+      "step": 7568
+    },
+    {
+      "epoch": 15.138,
+      "grad_norm": 1.0853475332260132,
+      "learning_rate": 2e-05,
+      "loss": 0.03965416,
+      "step": 7569
+    },
+    {
+      "epoch": 15.14,
+      "grad_norm": 1.4762166738510132,
+      "learning_rate": 2e-05,
+      "loss": 0.0375259,
+      "step": 7570
+    },
+    {
+      "epoch": 15.142,
+      "grad_norm": 1.307599425315857,
+      "learning_rate": 2e-05,
+      "loss": 0.04117063,
+      "step": 7571
+    },
+    {
+      "epoch": 15.144,
+      "grad_norm": 1.364780068397522,
+      "learning_rate": 2e-05,
+      "loss": 0.03955507,
+      "step": 7572
+    },
+    {
+      "epoch": 15.146,
+      "grad_norm": 1.8140677213668823,
+      "learning_rate": 2e-05,
+      "loss": 0.04545042,
+      "step": 7573
+    },
+    {
+      "epoch": 15.148,
+      "grad_norm": 1.234650731086731,
+      "learning_rate": 2e-05,
+      "loss": 0.03562243,
+      "step": 7574
+    },
+    {
+      "epoch": 15.15,
+      "grad_norm": 1.4778765439987183,
+      "learning_rate": 2e-05,
+      "loss": 0.05000856,
+      "step": 7575
+    },
+    {
+      "epoch": 15.152,
+      "grad_norm": 1.323702096939087,
+      "learning_rate": 2e-05,
+      "loss": 0.02919011,
+      "step": 7576
+    },
+    {
+      "epoch": 15.154,
+      "grad_norm": 0.9931415319442749,
+      "learning_rate": 2e-05,
+      "loss": 0.03235832,
+      "step": 7577
+    },
+    {
+      "epoch": 15.156,
+      "grad_norm": 1.086990237236023,
+      "learning_rate": 2e-05,
+      "loss": 0.02611094,
+      "step": 7578
+    },
+    {
+      "epoch": 15.158,
+      "grad_norm": 1.284932255744934,
+      "learning_rate": 2e-05,
+      "loss": 0.04641385,
+      "step": 7579
+    },
+    {
+      "epoch": 15.16,
+      "grad_norm": 0.9622644782066345,
+      "learning_rate": 2e-05,
+      "loss": 0.02520053,
+      "step": 7580
+    },
+    {
+      "epoch": 15.162,
+      "grad_norm": 1.0503311157226562,
+      "learning_rate": 2e-05,
+      "loss": 0.03852223,
+      "step": 7581
+    },
+    {
+      "epoch": 15.164,
+      "grad_norm": 1.5426743030548096,
+      "learning_rate": 2e-05,
+      "loss": 0.04453558,
+      "step": 7582
+    },
+    {
+      "epoch": 15.166,
+      "grad_norm": 1.0924078226089478,
+      "learning_rate": 2e-05,
+      "loss": 0.03900549,
+      "step": 7583
+    },
+    {
+      "epoch": 15.168,
+      "grad_norm": 1.5701236724853516,
+      "learning_rate": 2e-05,
+      "loss": 0.03460903,
+      "step": 7584
+    },
+    {
+      "epoch": 15.17,
+      "grad_norm": 1.7909417152404785,
+      "learning_rate": 2e-05,
+      "loss": 0.02750588,
+      "step": 7585
+    },
+    {
+      "epoch": 15.172,
+      "grad_norm": 3.5445761680603027,
+      "learning_rate": 2e-05,
+      "loss": 0.05145767,
+      "step": 7586
+    },
+    {
+      "epoch": 15.174,
+      "grad_norm": 1.7344753742218018,
+      "learning_rate": 2e-05,
+      "loss": 0.04307804,
+      "step": 7587
+    },
+    {
+      "epoch": 15.176,
+      "grad_norm": 1.1254594326019287,
+      "learning_rate": 2e-05,
+      "loss": 0.03898653,
+      "step": 7588
+    },
+    {
+      "epoch": 15.178,
+      "grad_norm": 1.2491363286972046,
+      "learning_rate": 2e-05,
+      "loss": 0.0345161,
+      "step": 7589
+    },
+    {
+      "epoch": 15.18,
+      "grad_norm": 1.5165410041809082,
+      "learning_rate": 2e-05,
+      "loss": 0.04044965,
+      "step": 7590
+    },
+    {
+      "epoch": 15.182,
+      "grad_norm": 1.306868553161621,
+      "learning_rate": 2e-05,
+      "loss": 0.04725231,
+      "step": 7591
+    },
+    {
+      "epoch": 15.184,
+      "grad_norm": 1.8153162002563477,
+      "learning_rate": 2e-05,
+      "loss": 0.0434998,
+      "step": 7592
+    },
+    {
+      "epoch": 15.186,
+      "grad_norm": 2.0734148025512695,
+      "learning_rate": 2e-05,
+      "loss": 0.04822708,
+      "step": 7593
+    },
+    {
+      "epoch": 15.188,
+      "grad_norm": 1.193581223487854,
+      "learning_rate": 2e-05,
+      "loss": 0.04005368,
+      "step": 7594
+    },
+    {
+      "epoch": 15.19,
+      "grad_norm": 1.9884793758392334,
+      "learning_rate": 2e-05,
+      "loss": 0.03887493,
+      "step": 7595
+    },
+    {
+      "epoch": 15.192,
+      "grad_norm": 2.539429187774658,
+      "learning_rate": 2e-05,
+      "loss": 0.03816453,
+      "step": 7596
+    },
+    {
+      "epoch": 15.194,
+      "grad_norm": 1.7256981134414673,
+      "learning_rate": 2e-05,
+      "loss": 0.02610845,
+      "step": 7597
+    },
+    {
+      "epoch": 15.196,
+      "grad_norm": 0.9873558282852173,
+      "learning_rate": 2e-05,
+      "loss": 0.02836427,
+      "step": 7598
+    },
+    {
+      "epoch": 15.198,
+      "grad_norm": 3.33852481842041,
+      "learning_rate": 2e-05,
+      "loss": 0.06002682,
+      "step": 7599
+    },
+    {
+      "epoch": 15.2,
+      "grad_norm": 2.537189483642578,
+      "learning_rate": 2e-05,
+      "loss": 0.07417429,
+      "step": 7600
+    },
+    {
+      "epoch": 15.202,
+      "grad_norm": 0.9124858379364014,
+      "learning_rate": 2e-05,
+      "loss": 0.03152982,
+      "step": 7601
+    },
+    {
+      "epoch": 15.204,
+      "grad_norm": 1.0165115594863892,
+      "learning_rate": 2e-05,
+      "loss": 0.03425337,
+      "step": 7602
+    },
+    {
+      "epoch": 15.206,
+      "grad_norm": 1.362256407737732,
+      "learning_rate": 2e-05,
+      "loss": 0.04571758,
+      "step": 7603
+    },
+    {
+      "epoch": 15.208,
+      "grad_norm": 1.2816169261932373,
+      "learning_rate": 2e-05,
+      "loss": 0.04699683,
+      "step": 7604
+    },
+    {
+      "epoch": 15.21,
+      "grad_norm": 1.6668884754180908,
+      "learning_rate": 2e-05,
+      "loss": 0.03962279,
+      "step": 7605
+    },
+    {
+      "epoch": 15.212,
+      "grad_norm": 1.2966806888580322,
+      "learning_rate": 2e-05,
+      "loss": 0.04509297,
+      "step": 7606
+    },
+    {
+      "epoch": 15.214,
+      "grad_norm": 1.191419243812561,
+      "learning_rate": 2e-05,
+      "loss": 0.04549539,
+      "step": 7607
+    },
+    {
+      "epoch": 15.216,
+      "grad_norm": 0.941321611404419,
+      "learning_rate": 2e-05,
+      "loss": 0.02918254,
+      "step": 7608
+    },
+    {
+      "epoch": 15.218,
+      "grad_norm": 1.7886515855789185,
+      "learning_rate": 2e-05,
+      "loss": 0.04620847,
+      "step": 7609
+    },
+    {
+      "epoch": 15.22,
+      "grad_norm": 1.8236616849899292,
+      "learning_rate": 2e-05,
+      "loss": 0.04852483,
+      "step": 7610
+    },
+    {
+      "epoch": 15.222,
+      "grad_norm": 1.224780797958374,
+      "learning_rate": 2e-05,
+      "loss": 0.0375507,
+      "step": 7611
+    },
+    {
+      "epoch": 15.224,
+      "grad_norm": 1.335411548614502,
+      "learning_rate": 2e-05,
+      "loss": 0.04444863,
+      "step": 7612
+    },
+    {
+      "epoch": 15.226,
+      "grad_norm": 1.0695840120315552,
+      "learning_rate": 2e-05,
+      "loss": 0.03443304,
+      "step": 7613
+    },
+    {
+      "epoch": 15.228,
+      "grad_norm": 1.1740689277648926,
+      "learning_rate": 2e-05,
+      "loss": 0.03335565,
+      "step": 7614
+    },
+    {
+      "epoch": 15.23,
+      "grad_norm": 2.6707682609558105,
+      "learning_rate": 2e-05,
+      "loss": 0.05265085,
+      "step": 7615
+    },
+    {
+      "epoch": 15.232,
+      "grad_norm": 4.135421276092529,
+      "learning_rate": 2e-05,
+      "loss": 0.0563576,
+      "step": 7616
+    },
+    {
+      "epoch": 15.234,
+      "grad_norm": 1.716046690940857,
+      "learning_rate": 2e-05,
+      "loss": 0.04460949,
+      "step": 7617
+    },
+    {
+      "epoch": 15.236,
+      "grad_norm": 1.721646785736084,
+      "learning_rate": 2e-05,
+      "loss": 0.04767875,
+      "step": 7618
+    },
+    {
+      "epoch": 15.238,
+      "grad_norm": 1.284906029701233,
+      "learning_rate": 2e-05,
+      "loss": 0.01877031,
+      "step": 7619
+    },
+    {
+      "epoch": 15.24,
+      "grad_norm": 1.517042636871338,
+      "learning_rate": 2e-05,
+      "loss": 0.04870673,
+      "step": 7620
+    },
+    {
+      "epoch": 15.242,
+      "grad_norm": 0.9522220492362976,
+      "learning_rate": 2e-05,
+      "loss": 0.02781921,
+      "step": 7621
+    },
+    {
+      "epoch": 15.244,
+      "grad_norm": 1.4575544595718384,
+      "learning_rate": 2e-05,
+      "loss": 0.04900608,
+      "step": 7622
+    },
+    {
+      "epoch": 15.246,
+      "grad_norm": 2.664926052093506,
+      "learning_rate": 2e-05,
+      "loss": 0.05986939,
+      "step": 7623
+    },
+    {
+      "epoch": 15.248,
+      "grad_norm": 1.0192816257476807,
+      "learning_rate": 2e-05,
+      "loss": 0.03001048,
+      "step": 7624
+    },
+    {
+      "epoch": 15.25,
+      "grad_norm": 2.431614637374878,
+      "learning_rate": 2e-05,
+      "loss": 0.05669653,
+      "step": 7625
+    },
+    {
+      "epoch": 15.252,
+      "grad_norm": 1.4480475187301636,
+      "learning_rate": 2e-05,
+      "loss": 0.0410952,
+      "step": 7626
+    },
+    {
+      "epoch": 15.254,
+      "grad_norm": 1.0749616622924805,
+      "learning_rate": 2e-05,
+      "loss": 0.03243171,
+      "step": 7627
+    },
+    {
+      "epoch": 15.256,
+      "grad_norm": 1.4723212718963623,
+      "learning_rate": 2e-05,
+      "loss": 0.04396664,
+      "step": 7628
+    },
+    {
+      "epoch": 15.258,
+      "grad_norm": 1.4612725973129272,
+      "learning_rate": 2e-05,
+      "loss": 0.05846308,
+      "step": 7629
+    },
+    {
+      "epoch": 15.26,
+      "grad_norm": 1.0961060523986816,
+      "learning_rate": 2e-05,
+      "loss": 0.03180789,
+      "step": 7630
+    },
+    {
+      "epoch": 15.262,
+      "grad_norm": 1.0098501443862915,
+      "learning_rate": 2e-05,
+      "loss": 0.02932501,
+      "step": 7631
+    },
+    {
+      "epoch": 15.264,
+      "grad_norm": 1.500989317893982,
+      "learning_rate": 2e-05,
+      "loss": 0.05016597,
+      "step": 7632
+    },
+    {
+      "epoch": 15.266,
+      "grad_norm": 1.462710976600647,
+      "learning_rate": 2e-05,
+      "loss": 0.04754984,
+      "step": 7633
+    },
+    {
+      "epoch": 15.268,
+      "grad_norm": 1.3882156610488892,
+      "learning_rate": 2e-05,
+      "loss": 0.04774838,
+      "step": 7634
+    },
+    {
+      "epoch": 15.27,
+      "grad_norm": 1.2340351343154907,
+      "learning_rate": 2e-05,
+      "loss": 0.05167802,
+      "step": 7635
+    },
+    {
+      "epoch": 15.272,
+      "grad_norm": 1.213295817375183,
+      "learning_rate": 2e-05,
+      "loss": 0.04757159,
+      "step": 7636
+    },
+    {
+      "epoch": 15.274000000000001,
+      "grad_norm": 1.2619282007217407,
+      "learning_rate": 2e-05,
+      "loss": 0.05256531,
+      "step": 7637
+    },
+    {
+      "epoch": 15.276,
+      "grad_norm": 1.552124261856079,
+      "learning_rate": 2e-05,
+      "loss": 0.04303239,
+      "step": 7638
+    },
+    {
+      "epoch": 15.278,
+      "grad_norm": 1.615976333618164,
+      "learning_rate": 2e-05,
+      "loss": 0.05144659,
+      "step": 7639
+    },
+    {
+      "epoch": 15.28,
+      "grad_norm": 1.2104065418243408,
+      "learning_rate": 2e-05,
+      "loss": 0.02734979,
+      "step": 7640
+    },
+    {
+      "epoch": 15.282,
+      "grad_norm": 2.24019455909729,
+      "learning_rate": 2e-05,
+      "loss": 0.04673993,
+      "step": 7641
+    },
+    {
+      "epoch": 15.284,
+      "grad_norm": 1.7466453313827515,
+      "learning_rate": 2e-05,
+      "loss": 0.03623344,
+      "step": 7642
+    },
+    {
+      "epoch": 15.286,
+      "grad_norm": 1.443344235420227,
+      "learning_rate": 2e-05,
+      "loss": 0.06160107,
+      "step": 7643
+    },
+    {
+      "epoch": 15.288,
+      "grad_norm": 0.9850043654441833,
+      "learning_rate": 2e-05,
+      "loss": 0.03402284,
+      "step": 7644
+    },
+    {
+      "epoch": 15.29,
+      "grad_norm": 1.4699482917785645,
+      "learning_rate": 2e-05,
+      "loss": 0.04960431,
+      "step": 7645
+    },
+    {
+      "epoch": 15.292,
+      "grad_norm": 0.8527708053588867,
+      "learning_rate": 2e-05,
+      "loss": 0.02158963,
+      "step": 7646
+    },
+    {
+      "epoch": 15.294,
+      "grad_norm": 0.7715446352958679,
+      "learning_rate": 2e-05,
+      "loss": 0.02164961,
+      "step": 7647
+    },
+    {
+      "epoch": 15.296,
+      "grad_norm": 1.1812517642974854,
+      "learning_rate": 2e-05,
+      "loss": 0.04358544,
+      "step": 7648
+    },
+    {
+      "epoch": 15.298,
+      "grad_norm": 1.6417049169540405,
+      "learning_rate": 2e-05,
+      "loss": 0.05505785,
+      "step": 7649
+    },
+    {
+      "epoch": 15.3,
+      "grad_norm": 0.9022408723831177,
+      "learning_rate": 2e-05,
+      "loss": 0.02869208,
+      "step": 7650
+    },
+    {
+      "epoch": 15.302,
+      "grad_norm": 1.2955352067947388,
+      "learning_rate": 2e-05,
+      "loss": 0.05353476,
+      "step": 7651
+    },
+    {
+      "epoch": 15.304,
+      "grad_norm": 2.33793044090271,
+      "learning_rate": 2e-05,
+      "loss": 0.02959348,
+      "step": 7652
+    },
+    {
+      "epoch": 15.306,
+      "grad_norm": 1.1095025539398193,
+      "learning_rate": 2e-05,
+      "loss": 0.03602049,
+      "step": 7653
+    },
+    {
+      "epoch": 15.308,
+      "grad_norm": 1.6043916940689087,
+      "learning_rate": 2e-05,
+      "loss": 0.03583729,
+      "step": 7654
+    },
+    {
+      "epoch": 15.31,
+      "grad_norm": 1.0492905378341675,
+      "learning_rate": 2e-05,
+      "loss": 0.03110175,
+      "step": 7655
+    },
+    {
+      "epoch": 15.312,
+      "grad_norm": 1.7085450887680054,
+      "learning_rate": 2e-05,
+      "loss": 0.04954446,
+      "step": 7656
+    },
+    {
+      "epoch": 15.314,
+      "grad_norm": 1.9505372047424316,
+      "learning_rate": 2e-05,
+      "loss": 0.03658181,
+      "step": 7657
+    },
+    {
+      "epoch": 15.316,
+      "grad_norm": 0.9936202168464661,
+      "learning_rate": 2e-05,
+      "loss": 0.02535901,
+      "step": 7658
+    },
+    {
+      "epoch": 15.318,
+      "grad_norm": 1.0207105875015259,
+      "learning_rate": 2e-05,
+      "loss": 0.03539135,
+      "step": 7659
+    },
+    {
+      "epoch": 15.32,
+      "grad_norm": 1.5546672344207764,
+      "learning_rate": 2e-05,
+      "loss": 0.04938481,
+      "step": 7660
+    },
+    {
+      "epoch": 15.322,
+      "grad_norm": 1.247316837310791,
+      "learning_rate": 2e-05,
+      "loss": 0.03931482,
+      "step": 7661
+    },
+    {
+      "epoch": 15.324,
+      "grad_norm": 1.2438799142837524,
+      "learning_rate": 2e-05,
+      "loss": 0.02923568,
+      "step": 7662
+    },
+    {
+      "epoch": 15.326,
+      "grad_norm": 1.87889564037323,
+      "learning_rate": 2e-05,
+      "loss": 0.04272472,
+      "step": 7663
+    },
+    {
+      "epoch": 15.328,
+      "grad_norm": 1.0760419368743896,
+      "learning_rate": 2e-05,
+      "loss": 0.03573871,
+      "step": 7664
+    },
+    {
+      "epoch": 15.33,
+      "grad_norm": 2.030569314956665,
+      "learning_rate": 2e-05,
+      "loss": 0.04339175,
+      "step": 7665
+    },
+    {
+      "epoch": 15.332,
+      "grad_norm": 1.7053855657577515,
+      "learning_rate": 2e-05,
+      "loss": 0.04916566,
+      "step": 7666
+    },
+    {
+      "epoch": 15.334,
+      "grad_norm": 1.4660038948059082,
+      "learning_rate": 2e-05,
+      "loss": 0.03246354,
+      "step": 7667
+    },
+    {
+      "epoch": 15.336,
+      "grad_norm": 1.257335901260376,
+      "learning_rate": 2e-05,
+      "loss": 0.04967248,
+      "step": 7668
+    },
+    {
+      "epoch": 15.338,
+      "grad_norm": 1.4573662281036377,
+      "learning_rate": 2e-05,
+      "loss": 0.04021806,
+      "step": 7669
+    },
+    {
+      "epoch": 15.34,
+      "grad_norm": 1.271093487739563,
+      "learning_rate": 2e-05,
+      "loss": 0.04290817,
+      "step": 7670
+    },
+    {
+      "epoch": 15.342,
+      "grad_norm": 0.9368219971656799,
+      "learning_rate": 2e-05,
+      "loss": 0.0271676,
+      "step": 7671
+    },
+    {
+      "epoch": 15.344,
+      "grad_norm": 1.1794713735580444,
+      "learning_rate": 2e-05,
+      "loss": 0.03208122,
+      "step": 7672
+    },
+    {
+      "epoch": 15.346,
+      "grad_norm": 0.9926052689552307,
+      "learning_rate": 2e-05,
+      "loss": 0.02915346,
+      "step": 7673
+    },
+    {
+      "epoch": 15.348,
+      "grad_norm": 1.1521005630493164,
+      "learning_rate": 2e-05,
+      "loss": 0.03530318,
+      "step": 7674
+    },
+    {
+      "epoch": 15.35,
+      "grad_norm": 1.2285836935043335,
+      "learning_rate": 2e-05,
+      "loss": 0.03318744,
+      "step": 7675
+    },
+    {
+      "epoch": 15.352,
+      "grad_norm": 2.876605987548828,
+      "learning_rate": 2e-05,
+      "loss": 0.02903116,
+      "step": 7676
+    },
+    {
+      "epoch": 15.354,
+      "grad_norm": 2.023113489151001,
+      "learning_rate": 2e-05,
+      "loss": 0.04902362,
+      "step": 7677
+    },
+    {
+      "epoch": 15.356,
+      "grad_norm": 1.7338629961013794,
+      "learning_rate": 2e-05,
+      "loss": 0.03628127,
+      "step": 7678
+    },
+    {
+      "epoch": 15.358,
+      "grad_norm": 1.0066721439361572,
+      "learning_rate": 2e-05,
+      "loss": 0.0289756,
+      "step": 7679
+    },
+    {
+      "epoch": 15.36,
+      "grad_norm": 1.7639458179473877,
+      "learning_rate": 2e-05,
+      "loss": 0.04403928,
+      "step": 7680
+    },
+    {
+      "epoch": 15.362,
+      "grad_norm": 1.4575209617614746,
+      "learning_rate": 2e-05,
+      "loss": 0.0353655,
+      "step": 7681
+    },
+    {
+      "epoch": 15.364,
+      "grad_norm": 1.1432892084121704,
+      "learning_rate": 2e-05,
+      "loss": 0.03814424,
+      "step": 7682
+    },
+    {
+      "epoch": 15.366,
+      "grad_norm": 1.755494236946106,
+      "learning_rate": 2e-05,
+      "loss": 0.0562299,
+      "step": 7683
+    },
+    {
+      "epoch": 15.368,
+      "grad_norm": 1.7338649034500122,
+      "learning_rate": 2e-05,
+      "loss": 0.03804263,
+      "step": 7684
+    },
+    {
+      "epoch": 15.37,
+      "grad_norm": 2.1905031204223633,
+      "learning_rate": 2e-05,
+      "loss": 0.03826035,
+      "step": 7685
+    },
+    {
+      "epoch": 15.372,
+      "grad_norm": 3.7067763805389404,
+      "learning_rate": 2e-05,
+      "loss": 0.05445283,
+      "step": 7686
+    },
+    {
+      "epoch": 15.374,
+      "grad_norm": 2.2767043113708496,
+      "learning_rate": 2e-05,
+      "loss": 0.06008461,
+      "step": 7687
+    },
+    {
+      "epoch": 15.376,
+      "grad_norm": 1.2579028606414795,
+      "learning_rate": 2e-05,
+      "loss": 0.03845136,
+      "step": 7688
+    },
+    {
+      "epoch": 15.378,
+      "grad_norm": 1.5080291032791138,
+      "learning_rate": 2e-05,
+      "loss": 0.04014179,
+      "step": 7689
+    },
+    {
+      "epoch": 15.38,
+      "grad_norm": 1.7374647855758667,
+      "learning_rate": 2e-05,
+      "loss": 0.03181672,
+      "step": 7690
+    },
+    {
+      "epoch": 15.382,
+      "grad_norm": 1.6790274381637573,
+      "learning_rate": 2e-05,
+      "loss": 0.03847949,
+      "step": 7691
+    },
+    {
+      "epoch": 15.384,
+      "grad_norm": 2.135026216506958,
+      "learning_rate": 2e-05,
+      "loss": 0.04469639,
+      "step": 7692
+    },
+    {
+      "epoch": 15.386,
+      "grad_norm": 1.4211525917053223,
+      "learning_rate": 2e-05,
+      "loss": 0.04342383,
+      "step": 7693
+    },
+    {
+      "epoch": 15.388,
+      "grad_norm": 1.0718973875045776,
+      "learning_rate": 2e-05,
+      "loss": 0.03107191,
+      "step": 7694
+    },
+    {
+      "epoch": 15.39,
+      "grad_norm": 0.8555018901824951,
+      "learning_rate": 2e-05,
+      "loss": 0.02420723,
+      "step": 7695
+    },
+    {
+      "epoch": 15.392,
+      "grad_norm": 1.2442560195922852,
+      "learning_rate": 2e-05,
+      "loss": 0.04132719,
+      "step": 7696
+    },
+    {
+      "epoch": 15.394,
+      "grad_norm": 1.0570979118347168,
+      "learning_rate": 2e-05,
+      "loss": 0.03066903,
+      "step": 7697
+    },
+    {
+      "epoch": 15.396,
+      "grad_norm": 1.171759009361267,
+      "learning_rate": 2e-05,
+      "loss": 0.03119994,
+      "step": 7698
+    },
+    {
+      "epoch": 15.398,
+      "grad_norm": 1.6701782941818237,
+      "learning_rate": 2e-05,
+      "loss": 0.03852317,
+      "step": 7699
+    },
+    {
+      "epoch": 15.4,
+      "grad_norm": 1.8172271251678467,
+      "learning_rate": 2e-05,
+      "loss": 0.02544076,
+      "step": 7700
+    },
+    {
+      "epoch": 15.402,
+      "grad_norm": 1.2106478214263916,
+      "learning_rate": 2e-05,
+      "loss": 0.03819906,
+      "step": 7701
+    },
+    {
+      "epoch": 15.404,
+      "grad_norm": 2.4221479892730713,
+      "learning_rate": 2e-05,
+      "loss": 0.04959261,
+      "step": 7702
+    },
+    {
+      "epoch": 15.406,
+      "grad_norm": 1.6806089878082275,
+      "learning_rate": 2e-05,
+      "loss": 0.04468667,
+      "step": 7703
+    },
+    {
+      "epoch": 15.408,
+      "grad_norm": 1.3105989694595337,
+      "learning_rate": 2e-05,
+      "loss": 0.03792799,
+      "step": 7704
+    },
+    {
+      "epoch": 15.41,
+      "grad_norm": 1.3647056818008423,
+      "learning_rate": 2e-05,
+      "loss": 0.04311409,
+      "step": 7705
+    },
+    {
+      "epoch": 15.412,
+      "grad_norm": 1.6955467462539673,
+      "learning_rate": 2e-05,
+      "loss": 0.04090759,
+      "step": 7706
+    },
+    {
+      "epoch": 15.414,
+      "grad_norm": 1.6445177793502808,
+      "learning_rate": 2e-05,
+      "loss": 0.050005,
+      "step": 7707
+    },
+    {
+      "epoch": 15.416,
+      "grad_norm": 2.115290403366089,
+      "learning_rate": 2e-05,
+      "loss": 0.04707139,
+      "step": 7708
+    },
+    {
+      "epoch": 15.418,
+      "grad_norm": 1.0016669034957886,
+      "learning_rate": 2e-05,
+      "loss": 0.02786621,
+      "step": 7709
+    },
+    {
+      "epoch": 15.42,
+      "grad_norm": 1.5075523853302002,
+      "learning_rate": 2e-05,
+      "loss": 0.04682674,
+      "step": 7710
+    },
+    {
+      "epoch": 15.422,
+      "grad_norm": 1.5943173170089722,
+      "learning_rate": 2e-05,
+      "loss": 0.0265224,
+      "step": 7711
+    },
+    {
+      "epoch": 15.424,
+      "grad_norm": 1.5050104856491089,
+      "learning_rate": 2e-05,
+      "loss": 0.04416554,
+      "step": 7712
+    },
+    {
+      "epoch": 15.426,
+      "grad_norm": 0.9919576644897461,
+      "learning_rate": 2e-05,
+      "loss": 0.02424739,
+      "step": 7713
+    },
+    {
+      "epoch": 15.428,
+      "grad_norm": 1.0542445182800293,
+      "learning_rate": 2e-05,
+      "loss": 0.03701505,
+      "step": 7714
+    },
+    {
+      "epoch": 15.43,
+      "grad_norm": 0.9411125779151917,
+      "learning_rate": 2e-05,
+      "loss": 0.03089548,
+      "step": 7715
+    },
+    {
+      "epoch": 15.432,
+      "grad_norm": 1.2335981130599976,
+      "learning_rate": 2e-05,
+      "loss": 0.04131325,
+      "step": 7716
+    },
+    {
+      "epoch": 15.434,
+      "grad_norm": 1.460959553718567,
+      "learning_rate": 2e-05,
+      "loss": 0.03338287,
+      "step": 7717
+    },
+    {
+      "epoch": 15.436,
+      "grad_norm": 0.9841430187225342,
+      "learning_rate": 2e-05,
+      "loss": 0.03597587,
+      "step": 7718
+    },
+    {
+      "epoch": 15.438,
+      "grad_norm": 1.1647802591323853,
+      "learning_rate": 2e-05,
+      "loss": 0.03573506,
+      "step": 7719
+    },
+    {
+      "epoch": 15.44,
+      "grad_norm": 1.4322996139526367,
+      "learning_rate": 2e-05,
+      "loss": 0.04496816,
+      "step": 7720
+    },
+    {
+      "epoch": 15.442,
+      "grad_norm": 3.0110630989074707,
+      "learning_rate": 2e-05,
+      "loss": 0.04119972,
+      "step": 7721
+    },
+    {
+      "epoch": 15.444,
+      "grad_norm": 1.2038313150405884,
+      "learning_rate": 2e-05,
+      "loss": 0.03304677,
+      "step": 7722
+    },
+    {
+      "epoch": 15.446,
+      "grad_norm": 1.3562027215957642,
+      "learning_rate": 2e-05,
+      "loss": 0.04422673,
+      "step": 7723
+    },
+    {
+      "epoch": 15.448,
+      "grad_norm": 1.345652461051941,
+      "learning_rate": 2e-05,
+      "loss": 0.04346151,
+      "step": 7724
+    },
+    {
+      "epoch": 15.45,
+      "grad_norm": 0.9705491662025452,
+      "learning_rate": 2e-05,
+      "loss": 0.02673662,
+      "step": 7725
+    },
+    {
+      "epoch": 15.452,
+      "grad_norm": 1.6109822988510132,
+      "learning_rate": 2e-05,
+      "loss": 0.04332538,
+      "step": 7726
+    },
+    {
+      "epoch": 15.454,
+      "grad_norm": 1.4140931367874146,
+      "learning_rate": 2e-05,
+      "loss": 0.04497489,
+      "step": 7727
+    },
+    {
+      "epoch": 15.456,
+      "grad_norm": 1.2737836837768555,
+      "learning_rate": 2e-05,
+      "loss": 0.03585126,
+      "step": 7728
+    },
+    {
+      "epoch": 15.458,
+      "grad_norm": 1.0965650081634521,
+      "learning_rate": 2e-05,
+      "loss": 0.02634549,
+      "step": 7729
+    },
+    {
+      "epoch": 15.46,
+      "grad_norm": 1.3999693393707275,
+      "learning_rate": 2e-05,
+      "loss": 0.04551007,
+      "step": 7730
+    },
+    {
+      "epoch": 15.462,
+      "grad_norm": 1.013411045074463,
+      "learning_rate": 2e-05,
+      "loss": 0.02215614,
+      "step": 7731
+    },
+    {
+      "epoch": 15.464,
+      "grad_norm": 1.3173936605453491,
+      "learning_rate": 2e-05,
+      "loss": 0.04206047,
+      "step": 7732
+    },
+    {
+      "epoch": 15.466,
+      "grad_norm": 1.0382493734359741,
+      "learning_rate": 2e-05,
+      "loss": 0.02827413,
+      "step": 7733
+    },
+    {
+      "epoch": 15.468,
+      "grad_norm": 2.1286673545837402,
+      "learning_rate": 2e-05,
+      "loss": 0.04167566,
+      "step": 7734
+    },
+    {
+      "epoch": 15.47,
+      "grad_norm": 2.1657168865203857,
+      "learning_rate": 2e-05,
+      "loss": 0.04439145,
+      "step": 7735
+    },
+    {
+      "epoch": 15.472,
+      "grad_norm": 1.6205779314041138,
+      "learning_rate": 2e-05,
+      "loss": 0.04393222,
+      "step": 7736
+    },
+    {
+      "epoch": 15.474,
+      "grad_norm": 1.3515499830245972,
+      "learning_rate": 2e-05,
+      "loss": 0.04231175,
+      "step": 7737
+    },
+    {
+      "epoch": 15.475999999999999,
+      "grad_norm": 1.089202880859375,
+      "learning_rate": 2e-05,
+      "loss": 0.02547337,
+      "step": 7738
+    },
+    {
+      "epoch": 15.478,
+      "grad_norm": 1.4004868268966675,
+      "learning_rate": 2e-05,
+      "loss": 0.0451882,
+      "step": 7739
+    },
+    {
+      "epoch": 15.48,
+      "grad_norm": 1.272910714149475,
+      "learning_rate": 2e-05,
+      "loss": 0.01924961,
+      "step": 7740
+    },
+    {
+      "epoch": 15.482,
+      "grad_norm": 1.2643826007843018,
+      "learning_rate": 2e-05,
+      "loss": 0.04131962,
+      "step": 7741
+    },
+    {
+      "epoch": 15.484,
+      "grad_norm": 2.1027934551239014,
+      "learning_rate": 2e-05,
+      "loss": 0.06375703,
+      "step": 7742
+    },
+    {
+      "epoch": 15.486,
+      "grad_norm": 1.2265628576278687,
+      "learning_rate": 2e-05,
+      "loss": 0.02747299,
+      "step": 7743
+    },
+    {
+      "epoch": 15.488,
+      "grad_norm": 2.7004458904266357,
+      "learning_rate": 2e-05,
+      "loss": 0.06089799,
+      "step": 7744
+    },
+    {
+      "epoch": 15.49,
+      "grad_norm": 0.9138990044593811,
+      "learning_rate": 2e-05,
+      "loss": 0.02951707,
+      "step": 7745
+    },
+    {
+      "epoch": 15.492,
+      "grad_norm": 1.3887770175933838,
+      "learning_rate": 2e-05,
+      "loss": 0.03592655,
+      "step": 7746
+    },
+    {
+      "epoch": 15.494,
+      "grad_norm": 1.399113655090332,
+      "learning_rate": 2e-05,
+      "loss": 0.03822273,
+      "step": 7747
+    },
+    {
+      "epoch": 15.496,
+      "grad_norm": 1.7515449523925781,
+      "learning_rate": 2e-05,
+      "loss": 0.04199392,
+      "step": 7748
+    },
+    {
+      "epoch": 15.498,
+      "grad_norm": 1.4153516292572021,
+      "learning_rate": 2e-05,
+      "loss": 0.04716617,
+      "step": 7749
+    },
+    {
+      "epoch": 15.5,
+      "grad_norm": 2.602876901626587,
+      "learning_rate": 2e-05,
+      "loss": 0.04844777,
+      "step": 7750
+    },
+    {
+      "epoch": 15.502,
+      "grad_norm": 1.2918318510055542,
+      "learning_rate": 2e-05,
+      "loss": 0.04199758,
+      "step": 7751
+    },
+    {
+      "epoch": 15.504,
+      "grad_norm": 0.7387645244598389,
+      "learning_rate": 2e-05,
+      "loss": 0.01926706,
+      "step": 7752
+    },
+    {
+      "epoch": 15.506,
+      "grad_norm": 1.028178095817566,
+      "learning_rate": 2e-05,
+      "loss": 0.02961706,
+      "step": 7753
+    },
+    {
+      "epoch": 15.508,
+      "grad_norm": 1.2695460319519043,
+      "learning_rate": 2e-05,
+      "loss": 0.03514853,
+      "step": 7754
+    },
+    {
+      "epoch": 15.51,
+      "grad_norm": 2.6682016849517822,
+      "learning_rate": 2e-05,
+      "loss": 0.06611276,
+      "step": 7755
+    },
+    {
+      "epoch": 15.512,
+      "grad_norm": 1.2907907962799072,
+      "learning_rate": 2e-05,
+      "loss": 0.03901846,
+      "step": 7756
+    },
+    {
+      "epoch": 15.514,
+      "grad_norm": 1.2247368097305298,
+      "learning_rate": 2e-05,
+      "loss": 0.03462955,
+      "step": 7757
+    },
+    {
+      "epoch": 15.516,
+      "grad_norm": 1.1045002937316895,
+      "learning_rate": 2e-05,
+      "loss": 0.02914516,
+      "step": 7758
+    },
+    {
+      "epoch": 15.518,
+      "grad_norm": 1.3329474925994873,
+      "learning_rate": 2e-05,
+      "loss": 0.03299643,
+      "step": 7759
+    },
+    {
+      "epoch": 15.52,
+      "grad_norm": 0.9705315828323364,
+      "learning_rate": 2e-05,
+      "loss": 0.0249468,
+      "step": 7760
+    },
+    {
+      "epoch": 15.522,
+      "grad_norm": 1.3083375692367554,
+      "learning_rate": 2e-05,
+      "loss": 0.04079135,
+      "step": 7761
+    },
+    {
+      "epoch": 15.524000000000001,
+      "grad_norm": 1.0932661294937134,
+      "learning_rate": 2e-05,
+      "loss": 0.03963382,
+      "step": 7762
+    },
+    {
+      "epoch": 15.526,
+      "grad_norm": 1.1822131872177124,
+      "learning_rate": 2e-05,
+      "loss": 0.03383718,
+      "step": 7763
+    },
+    {
+      "epoch": 15.528,
+      "grad_norm": 1.1285481452941895,
+      "learning_rate": 2e-05,
+      "loss": 0.0485891,
+      "step": 7764
+    },
+    {
+      "epoch": 15.53,
+      "grad_norm": 2.6022520065307617,
+      "learning_rate": 2e-05,
+      "loss": 0.04467677,
+      "step": 7765
+    },
+    {
+      "epoch": 15.532,
+      "grad_norm": 2.1171436309814453,
+      "learning_rate": 2e-05,
+      "loss": 0.05000626,
+      "step": 7766
+    },
+    {
+      "epoch": 15.534,
+      "grad_norm": 0.948603093624115,
+      "learning_rate": 2e-05,
+      "loss": 0.02811343,
+      "step": 7767
+    },
+    {
+      "epoch": 15.536,
+      "grad_norm": 3.1038739681243896,
+      "learning_rate": 2e-05,
+      "loss": 0.04482667,
+      "step": 7768
+    },
+    {
+      "epoch": 15.538,
+      "grad_norm": 1.6356004476547241,
+      "learning_rate": 2e-05,
+      "loss": 0.05364137,
+      "step": 7769
+    },
+    {
+      "epoch": 15.54,
+      "grad_norm": 1.1389317512512207,
+      "learning_rate": 2e-05,
+      "loss": 0.03385459,
+      "step": 7770
+    },
+    {
+      "epoch": 15.542,
+      "grad_norm": 1.53994619846344,
+      "learning_rate": 2e-05,
+      "loss": 0.04406751,
+      "step": 7771
+    },
+    {
+      "epoch": 15.544,
+      "grad_norm": 1.5611374378204346,
+      "learning_rate": 2e-05,
+      "loss": 0.0473385,
+      "step": 7772
+    },
+    {
+      "epoch": 15.546,
+      "grad_norm": 1.0507432222366333,
+      "learning_rate": 2e-05,
+      "loss": 0.02707745,
+      "step": 7773
+    },
+    {
+      "epoch": 15.548,
+      "grad_norm": 1.1123173236846924,
+      "learning_rate": 2e-05,
+      "loss": 0.03449173,
+      "step": 7774
+    },
+    {
+      "epoch": 15.55,
+      "grad_norm": 1.3029831647872925,
+      "learning_rate": 2e-05,
+      "loss": 0.03552601,
+      "step": 7775
+    },
+    {
+      "epoch": 15.552,
+      "grad_norm": 1.6921583414077759,
+      "learning_rate": 2e-05,
+      "loss": 0.05067451,
+      "step": 7776
+    },
+    {
+      "epoch": 15.554,
+      "grad_norm": 1.275075078010559,
+      "learning_rate": 2e-05,
+      "loss": 0.02784466,
+      "step": 7777
+    },
+    {
+      "epoch": 15.556000000000001,
+      "grad_norm": 1.4142465591430664,
+      "learning_rate": 2e-05,
+      "loss": 0.04801814,
+      "step": 7778
+    },
+    {
+      "epoch": 15.558,
+      "grad_norm": 1.0732700824737549,
+      "learning_rate": 2e-05,
+      "loss": 0.03766344,
+      "step": 7779
+    },
+    {
+      "epoch": 15.56,
+      "grad_norm": 1.280580997467041,
+      "learning_rate": 2e-05,
+      "loss": 0.03040556,
+      "step": 7780
+    },
+    {
+      "epoch": 15.562,
+      "grad_norm": 1.8353358507156372,
+      "learning_rate": 2e-05,
+      "loss": 0.05565303,
+      "step": 7781
+    },
+    {
+      "epoch": 15.564,
+      "grad_norm": 1.4312429428100586,
+      "learning_rate": 2e-05,
+      "loss": 0.03630222,
+      "step": 7782
+    },
+    {
+      "epoch": 15.566,
+      "grad_norm": 0.9180557131767273,
+      "learning_rate": 2e-05,
+      "loss": 0.03424197,
+      "step": 7783
+    },
+    {
+      "epoch": 15.568,
+      "grad_norm": 1.8518563508987427,
+      "learning_rate": 2e-05,
+      "loss": 0.05018555,
+      "step": 7784
+    },
+    {
+      "epoch": 15.57,
+      "grad_norm": 1.6587504148483276,
+      "learning_rate": 2e-05,
+      "loss": 0.05404566,
+      "step": 7785
+    },
+    {
+      "epoch": 15.572,
+      "grad_norm": 1.2998889684677124,
+      "learning_rate": 2e-05,
+      "loss": 0.03402172,
+      "step": 7786
+    },
+    {
+      "epoch": 15.574,
+      "grad_norm": 1.43665611743927,
+      "learning_rate": 2e-05,
+      "loss": 0.03349872,
+      "step": 7787
+    },
+    {
+      "epoch": 15.576,
+      "grad_norm": 1.3342171907424927,
+      "learning_rate": 2e-05,
+      "loss": 0.04050234,
+      "step": 7788
+    },
+    {
+      "epoch": 15.578,
+      "grad_norm": 1.462536096572876,
+      "learning_rate": 2e-05,
+      "loss": 0.03564973,
+      "step": 7789
+    },
+    {
+      "epoch": 15.58,
+      "grad_norm": 1.421918511390686,
+      "learning_rate": 2e-05,
+      "loss": 0.02840041,
+      "step": 7790
+    },
+    {
+      "epoch": 15.582,
+      "grad_norm": 2.16733980178833,
+      "learning_rate": 2e-05,
+      "loss": 0.04823796,
+      "step": 7791
+    },
+    {
+      "epoch": 15.584,
+      "grad_norm": 1.9504356384277344,
+      "learning_rate": 2e-05,
+      "loss": 0.04157623,
+      "step": 7792
+    },
+    {
+      "epoch": 15.586,
+      "grad_norm": 2.9325337409973145,
+      "learning_rate": 2e-05,
+      "loss": 0.03892434,
+      "step": 7793
+    },
+    {
+      "epoch": 15.588,
+      "grad_norm": 1.5955098867416382,
+      "learning_rate": 2e-05,
+      "loss": 0.04307658,
+      "step": 7794
+    },
+    {
+      "epoch": 15.59,
+      "grad_norm": 1.1761583089828491,
+      "learning_rate": 2e-05,
+      "loss": 0.03487573,
+      "step": 7795
+    },
+    {
+      "epoch": 15.592,
+      "grad_norm": 1.8221702575683594,
+      "learning_rate": 2e-05,
+      "loss": 0.05856226,
+      "step": 7796
+    },
+    {
+      "epoch": 15.594,
+      "grad_norm": 1.3372985124588013,
+      "learning_rate": 2e-05,
+      "loss": 0.04135559,
+      "step": 7797
+    },
+    {
+      "epoch": 15.596,
+      "grad_norm": 1.5400853157043457,
+      "learning_rate": 2e-05,
+      "loss": 0.04417802,
+      "step": 7798
+    },
+    {
+      "epoch": 15.598,
+      "grad_norm": 1.4287244081497192,
+      "learning_rate": 2e-05,
+      "loss": 0.03800631,
+      "step": 7799
+    },
+    {
+      "epoch": 15.6,
+      "grad_norm": 1.5493749380111694,
+      "learning_rate": 2e-05,
+      "loss": 0.03243243,
+      "step": 7800
+    },
+    {
+      "epoch": 15.602,
+      "grad_norm": 1.8770220279693604,
+      "learning_rate": 2e-05,
+      "loss": 0.04838708,
+      "step": 7801
+    },
+    {
+      "epoch": 15.604,
+      "grad_norm": 1.8443878889083862,
+      "learning_rate": 2e-05,
+      "loss": 0.04210856,
+      "step": 7802
+    },
+    {
+      "epoch": 15.606,
+      "grad_norm": 1.1599087715148926,
+      "learning_rate": 2e-05,
+      "loss": 0.03727244,
+      "step": 7803
+    },
+    {
+      "epoch": 15.608,
+      "grad_norm": 0.8143905997276306,
+      "learning_rate": 2e-05,
+      "loss": 0.02723217,
+      "step": 7804
+    },
+    {
+      "epoch": 15.61,
+      "grad_norm": 1.514428734779358,
+      "learning_rate": 2e-05,
+      "loss": 0.04333169,
+      "step": 7805
+    },
+    {
+      "epoch": 15.612,
+      "grad_norm": 1.0530762672424316,
+      "learning_rate": 2e-05,
+      "loss": 0.02669099,
+      "step": 7806
+    },
+    {
+      "epoch": 15.614,
+      "grad_norm": 1.2812447547912598,
+      "learning_rate": 2e-05,
+      "loss": 0.03541653,
+      "step": 7807
+    },
+    {
+      "epoch": 15.616,
+      "grad_norm": 0.8917434215545654,
+      "learning_rate": 2e-05,
+      "loss": 0.02033772,
+      "step": 7808
+    },
+    {
+      "epoch": 15.618,
+      "grad_norm": 1.388106346130371,
+      "learning_rate": 2e-05,
+      "loss": 0.0356602,
+      "step": 7809
+    },
+    {
+      "epoch": 15.62,
+      "grad_norm": 1.629955530166626,
+      "learning_rate": 2e-05,
+      "loss": 0.04595727,
+      "step": 7810
+    },
+    {
+      "epoch": 15.622,
+      "grad_norm": 1.964969277381897,
+      "learning_rate": 2e-05,
+      "loss": 0.04757115,
+      "step": 7811
+    },
+    {
+      "epoch": 15.624,
+      "grad_norm": 1.00034761428833,
+      "learning_rate": 2e-05,
+      "loss": 0.02988995,
+      "step": 7812
+    },
+    {
+      "epoch": 15.626,
+      "grad_norm": 1.4420907497406006,
+      "learning_rate": 2e-05,
+      "loss": 0.03713204,
+      "step": 7813
+    },
+    {
+      "epoch": 15.628,
+      "grad_norm": 1.7254669666290283,
+      "learning_rate": 2e-05,
+      "loss": 0.05012753,
+      "step": 7814
+    },
+    {
+      "epoch": 15.63,
+      "grad_norm": 1.617090106010437,
+      "learning_rate": 2e-05,
+      "loss": 0.04258318,
+      "step": 7815
+    },
+    {
+      "epoch": 15.632,
+      "grad_norm": 1.8159074783325195,
+      "learning_rate": 2e-05,
+      "loss": 0.0533081,
+      "step": 7816
+    },
+    {
+      "epoch": 15.634,
+      "grad_norm": 1.0557678937911987,
+      "learning_rate": 2e-05,
+      "loss": 0.03816307,
+      "step": 7817
+    },
+    {
+      "epoch": 15.636,
+      "grad_norm": 1.1911511421203613,
+      "learning_rate": 2e-05,
+      "loss": 0.02825535,
+      "step": 7818
+    },
+    {
+      "epoch": 15.638,
+      "grad_norm": 1.1290611028671265,
+      "learning_rate": 2e-05,
+      "loss": 0.03365977,
+      "step": 7819
+    },
+    {
+      "epoch": 15.64,
+      "grad_norm": 1.4087905883789062,
+      "learning_rate": 2e-05,
+      "loss": 0.03113106,
+      "step": 7820
+    },
+    {
+      "epoch": 15.642,
+      "grad_norm": 1.3932369947433472,
+      "learning_rate": 2e-05,
+      "loss": 0.03929411,
+      "step": 7821
+    },
+    {
+      "epoch": 15.644,
+      "grad_norm": 1.611405611038208,
+      "learning_rate": 2e-05,
+      "loss": 0.05581418,
+      "step": 7822
+    },
+    {
+      "epoch": 15.646,
+      "grad_norm": 1.394471526145935,
+      "learning_rate": 2e-05,
+      "loss": 0.03460547,
+      "step": 7823
+    },
+    {
+      "epoch": 15.648,
+      "grad_norm": 1.114952564239502,
+      "learning_rate": 2e-05,
+      "loss": 0.03128414,
+      "step": 7824
+    },
+    {
+      "epoch": 15.65,
+      "grad_norm": 1.2153035402297974,
+      "learning_rate": 2e-05,
+      "loss": 0.03026392,
+      "step": 7825
+    },
+    {
+      "epoch": 15.652,
+      "grad_norm": 1.245772361755371,
+      "learning_rate": 2e-05,
+      "loss": 0.04067977,
+      "step": 7826
+    },
+    {
+      "epoch": 15.654,
+      "grad_norm": 1.2498624324798584,
+      "learning_rate": 2e-05,
+      "loss": 0.0399544,
+      "step": 7827
+    },
+    {
+      "epoch": 15.656,
+      "grad_norm": 1.4734687805175781,
+      "learning_rate": 2e-05,
+      "loss": 0.0403645,
+      "step": 7828
+    },
+    {
+      "epoch": 15.658,
+      "grad_norm": 1.184058666229248,
+      "learning_rate": 2e-05,
+      "loss": 0.03515352,
+      "step": 7829
+    },
+    {
+      "epoch": 15.66,
+      "grad_norm": 1.6777468919754028,
+      "learning_rate": 2e-05,
+      "loss": 0.04460485,
+      "step": 7830
+    },
+    {
+      "epoch": 15.662,
+      "grad_norm": 1.869699478149414,
+      "learning_rate": 2e-05,
+      "loss": 0.03454949,
+      "step": 7831
+    },
+    {
+      "epoch": 15.664,
+      "grad_norm": 1.1511545181274414,
+      "learning_rate": 2e-05,
+      "loss": 0.03649542,
+      "step": 7832
+    },
+    {
+      "epoch": 15.666,
+      "grad_norm": 1.5351507663726807,
+      "learning_rate": 2e-05,
+      "loss": 0.03351032,
+      "step": 7833
+    },
+    {
+      "epoch": 15.668,
+      "grad_norm": 1.0590790510177612,
+      "learning_rate": 2e-05,
+      "loss": 0.04073928,
+      "step": 7834
+    },
+    {
+      "epoch": 15.67,
+      "grad_norm": 1.9279193878173828,
+      "learning_rate": 2e-05,
+      "loss": 0.05023941,
+      "step": 7835
+    },
+    {
+      "epoch": 15.672,
+      "grad_norm": 1.389474630355835,
+      "learning_rate": 2e-05,
+      "loss": 0.0375604,
+      "step": 7836
+    },
+    {
+      "epoch": 15.674,
+      "grad_norm": 1.5739963054656982,
+      "learning_rate": 2e-05,
+      "loss": 0.05597922,
+      "step": 7837
+    },
+    {
+      "epoch": 15.676,
+      "grad_norm": 2.872504472732544,
+      "learning_rate": 2e-05,
+      "loss": 0.03245638,
+      "step": 7838
+    },
+    {
+      "epoch": 15.678,
+      "grad_norm": 1.1751714944839478,
+      "learning_rate": 2e-05,
+      "loss": 0.04733478,
+      "step": 7839
+    },
+    {
+      "epoch": 15.68,
+      "grad_norm": 1.265234351158142,
+      "learning_rate": 2e-05,
+      "loss": 0.04258277,
+      "step": 7840
+    },
+    {
+      "epoch": 15.682,
+      "grad_norm": 1.1896003484725952,
+      "learning_rate": 2e-05,
+      "loss": 0.03169346,
+      "step": 7841
+    },
+    {
+      "epoch": 15.684,
+      "grad_norm": 1.7013142108917236,
+      "learning_rate": 2e-05,
+      "loss": 0.05300517,
+      "step": 7842
+    },
+    {
+      "epoch": 15.686,
+      "grad_norm": 1.666005253791809,
+      "learning_rate": 2e-05,
+      "loss": 0.04027229,
+      "step": 7843
+    },
+    {
+      "epoch": 15.688,
+      "grad_norm": 1.2857798337936401,
+      "learning_rate": 2e-05,
+      "loss": 0.02931785,
+      "step": 7844
+    },
+    {
+      "epoch": 15.69,
+      "grad_norm": 1.989986777305603,
+      "learning_rate": 2e-05,
+      "loss": 0.06056058,
+      "step": 7845
+    },
+    {
+      "epoch": 15.692,
+      "grad_norm": 1.0686649084091187,
+      "learning_rate": 2e-05,
+      "loss": 0.03009918,
+      "step": 7846
+    },
+    {
+      "epoch": 15.693999999999999,
+      "grad_norm": 1.2238051891326904,
+      "learning_rate": 2e-05,
+      "loss": 0.03506152,
+      "step": 7847
+    },
+    {
+      "epoch": 15.696,
+      "grad_norm": 1.7078624963760376,
+      "learning_rate": 2e-05,
+      "loss": 0.04438978,
+      "step": 7848
+    },
+    {
+      "epoch": 15.698,
+      "grad_norm": 2.4543190002441406,
+      "learning_rate": 2e-05,
+      "loss": 0.03001017,
+      "step": 7849
+    },
+    {
+      "epoch": 15.7,
+      "grad_norm": 1.3647072315216064,
+      "learning_rate": 2e-05,
+      "loss": 0.04168668,
+      "step": 7850
+    },
+    {
+      "epoch": 15.702,
+      "grad_norm": 1.3066121339797974,
+      "learning_rate": 2e-05,
+      "loss": 0.04153648,
+      "step": 7851
+    },
+    {
+      "epoch": 15.704,
+      "grad_norm": 1.0811482667922974,
+      "learning_rate": 2e-05,
+      "loss": 0.03107862,
+      "step": 7852
+    },
+    {
+      "epoch": 15.706,
+      "grad_norm": 3.0416204929351807,
+      "learning_rate": 2e-05,
+      "loss": 0.05520962,
+      "step": 7853
+    },
+    {
+      "epoch": 15.708,
+      "grad_norm": 2.0119881629943848,
+      "learning_rate": 2e-05,
+      "loss": 0.03293385,
+      "step": 7854
+    },
+    {
+      "epoch": 15.71,
+      "grad_norm": 1.3695378303527832,
+      "learning_rate": 2e-05,
+      "loss": 0.03456626,
+      "step": 7855
+    },
+    {
+      "epoch": 15.712,
+      "grad_norm": 1.410203218460083,
+      "learning_rate": 2e-05,
+      "loss": 0.0467374,
+      "step": 7856
+    },
+    {
+      "epoch": 15.714,
+      "grad_norm": 0.9683268666267395,
+      "learning_rate": 2e-05,
+      "loss": 0.02218588,
+      "step": 7857
+    },
+    {
+      "epoch": 15.716,
+      "grad_norm": 1.1725350618362427,
+      "learning_rate": 2e-05,
+      "loss": 0.0399307,
+      "step": 7858
+    },
+    {
+      "epoch": 15.718,
+      "grad_norm": 1.156903862953186,
+      "learning_rate": 2e-05,
+      "loss": 0.04084864,
+      "step": 7859
+    },
+    {
+      "epoch": 15.72,
+      "grad_norm": 1.099974274635315,
+      "learning_rate": 2e-05,
+      "loss": 0.03805087,
+      "step": 7860
+    },
+    {
+      "epoch": 15.722,
+      "grad_norm": 0.9585137367248535,
+      "learning_rate": 2e-05,
+      "loss": 0.02989551,
+      "step": 7861
+    },
+    {
+      "epoch": 15.724,
+      "grad_norm": 0.9844270348548889,
+      "learning_rate": 2e-05,
+      "loss": 0.02908732,
+      "step": 7862
+    },
+    {
+      "epoch": 15.725999999999999,
+      "grad_norm": 1.854518175125122,
+      "learning_rate": 2e-05,
+      "loss": 0.04460219,
+      "step": 7863
+    },
+    {
+      "epoch": 15.728,
+      "grad_norm": 1.2902519702911377,
+      "learning_rate": 2e-05,
+      "loss": 0.0421297,
+      "step": 7864
+    },
+    {
+      "epoch": 15.73,
+      "grad_norm": 0.9230936169624329,
+      "learning_rate": 2e-05,
+      "loss": 0.02882043,
+      "step": 7865
+    },
+    {
+      "epoch": 15.732,
+      "grad_norm": 1.5599287748336792,
+      "learning_rate": 2e-05,
+      "loss": 0.03928513,
+      "step": 7866
+    },
+    {
+      "epoch": 15.734,
+      "grad_norm": 1.1548335552215576,
+      "learning_rate": 2e-05,
+      "loss": 0.03820485,
+      "step": 7867
+    },
+    {
+      "epoch": 15.736,
+      "grad_norm": 1.2006397247314453,
+      "learning_rate": 2e-05,
+      "loss": 0.02806983,
+      "step": 7868
+    },
+    {
+      "epoch": 15.738,
+      "grad_norm": 2.920689582824707,
+      "learning_rate": 2e-05,
+      "loss": 0.0410319,
+      "step": 7869
+    },
+    {
+      "epoch": 15.74,
+      "grad_norm": 1.2876091003417969,
+      "learning_rate": 2e-05,
+      "loss": 0.03893904,
+      "step": 7870
+    },
+    {
+      "epoch": 15.742,
+      "grad_norm": 1.2860698699951172,
+      "learning_rate": 2e-05,
+      "loss": 0.03412844,
+      "step": 7871
+    },
+    {
+      "epoch": 15.744,
+      "grad_norm": 1.402205467224121,
+      "learning_rate": 2e-05,
+      "loss": 0.04057181,
+      "step": 7872
+    },
+    {
+      "epoch": 15.746,
+      "grad_norm": 1.643515706062317,
+      "learning_rate": 2e-05,
+      "loss": 0.03522312,
+      "step": 7873
+    },
+    {
+      "epoch": 15.748,
+      "grad_norm": 1.7808668613433838,
+      "learning_rate": 2e-05,
+      "loss": 0.051851,
+      "step": 7874
+    },
+    {
+      "epoch": 15.75,
+      "grad_norm": 1.259395956993103,
+      "learning_rate": 2e-05,
+      "loss": 0.03302209,
+      "step": 7875
+    },
+    {
+      "epoch": 15.752,
+      "grad_norm": 1.9360871315002441,
+      "learning_rate": 2e-05,
+      "loss": 0.03717834,
+      "step": 7876
+    },
+    {
+      "epoch": 15.754,
+      "grad_norm": 1.372341513633728,
+      "learning_rate": 2e-05,
+      "loss": 0.04391986,
+      "step": 7877
+    },
+    {
+      "epoch": 15.756,
+      "grad_norm": 2.2007229328155518,
+      "learning_rate": 2e-05,
+      "loss": 0.03659112,
+      "step": 7878
+    },
+    {
+      "epoch": 15.758,
+      "grad_norm": 2.7030446529388428,
+      "learning_rate": 2e-05,
+      "loss": 0.05342212,
+      "step": 7879
+    },
+    {
+      "epoch": 15.76,
+      "grad_norm": 1.5437792539596558,
+      "learning_rate": 2e-05,
+      "loss": 0.03884444,
+      "step": 7880
+    },
+    {
+      "epoch": 15.762,
+      "grad_norm": 0.9608680605888367,
+      "learning_rate": 2e-05,
+      "loss": 0.0303247,
+      "step": 7881
+    },
+    {
+      "epoch": 15.764,
+      "grad_norm": 1.3708651065826416,
+      "learning_rate": 2e-05,
+      "loss": 0.03259217,
+      "step": 7882
+    },
+    {
+      "epoch": 15.766,
+      "grad_norm": 1.9985793828964233,
+      "learning_rate": 2e-05,
+      "loss": 0.04025662,
+      "step": 7883
+    },
+    {
+      "epoch": 15.768,
+      "grad_norm": 1.3265514373779297,
+      "learning_rate": 2e-05,
+      "loss": 0.04500604,
+      "step": 7884
+    },
+    {
+      "epoch": 15.77,
+      "grad_norm": 1.3119704723358154,
+      "learning_rate": 2e-05,
+      "loss": 0.03770127,
+      "step": 7885
+    },
+    {
+      "epoch": 15.772,
+      "grad_norm": 1.0361597537994385,
+      "learning_rate": 2e-05,
+      "loss": 0.02482234,
+      "step": 7886
+    },
+    {
+      "epoch": 15.774000000000001,
+      "grad_norm": 1.0766968727111816,
+      "learning_rate": 2e-05,
+      "loss": 0.03262966,
+      "step": 7887
+    },
+    {
+      "epoch": 15.776,
+      "grad_norm": 1.372813105583191,
+      "learning_rate": 2e-05,
+      "loss": 0.04390799,
+      "step": 7888
+    },
+    {
+      "epoch": 15.778,
+      "grad_norm": 1.996690273284912,
+      "learning_rate": 2e-05,
+      "loss": 0.04705422,
+      "step": 7889
+    },
+    {
+      "epoch": 15.78,
+      "grad_norm": 1.3125027418136597,
+      "learning_rate": 2e-05,
+      "loss": 0.05340045,
+      "step": 7890
+    },
+    {
+      "epoch": 15.782,
+      "grad_norm": 1.0259758234024048,
+      "learning_rate": 2e-05,
+      "loss": 0.03187435,
+      "step": 7891
+    },
+    {
+      "epoch": 15.784,
+      "grad_norm": 1.176693320274353,
+      "learning_rate": 2e-05,
+      "loss": 0.03257523,
+      "step": 7892
+    },
+    {
+      "epoch": 15.786,
+      "grad_norm": 0.9599425196647644,
+      "learning_rate": 2e-05,
+      "loss": 0.02640433,
+      "step": 7893
+    },
+    {
+      "epoch": 15.788,
+      "grad_norm": 1.4598140716552734,
+      "learning_rate": 2e-05,
+      "loss": 0.03303267,
+      "step": 7894
+    },
+    {
+      "epoch": 15.79,
+      "grad_norm": 1.3222160339355469,
+      "learning_rate": 2e-05,
+      "loss": 0.03670667,
+      "step": 7895
+    },
+    {
+      "epoch": 15.792,
+      "grad_norm": 0.9463991522789001,
+      "learning_rate": 2e-05,
+      "loss": 0.02554088,
+      "step": 7896
+    },
+    {
+      "epoch": 15.794,
+      "grad_norm": 1.3123785257339478,
+      "learning_rate": 2e-05,
+      "loss": 0.0461954,
+      "step": 7897
+    },
+    {
+      "epoch": 15.796,
+      "grad_norm": 1.9543800354003906,
+      "learning_rate": 2e-05,
+      "loss": 0.0520162,
+      "step": 7898
+    },
+    {
+      "epoch": 15.798,
+      "grad_norm": 1.3964611291885376,
+      "learning_rate": 2e-05,
+      "loss": 0.03359364,
+      "step": 7899
+    },
+    {
+      "epoch": 15.8,
+      "grad_norm": 1.7428091764450073,
+      "learning_rate": 2e-05,
+      "loss": 0.04074088,
+      "step": 7900
+    },
+    {
+      "epoch": 15.802,
+      "grad_norm": 1.9242103099822998,
+      "learning_rate": 2e-05,
+      "loss": 0.02629791,
+      "step": 7901
+    },
+    {
+      "epoch": 15.804,
+      "grad_norm": 2.1996188163757324,
+      "learning_rate": 2e-05,
+      "loss": 0.04311199,
+      "step": 7902
+    },
+    {
+      "epoch": 15.806000000000001,
+      "grad_norm": 1.1592961549758911,
+      "learning_rate": 2e-05,
+      "loss": 0.0427461,
+      "step": 7903
+    },
+    {
+      "epoch": 15.808,
+      "grad_norm": 1.3940585851669312,
+      "learning_rate": 2e-05,
+      "loss": 0.04456202,
+      "step": 7904
+    },
+    {
+      "epoch": 15.81,
+      "grad_norm": 1.175157070159912,
+      "learning_rate": 2e-05,
+      "loss": 0.02939461,
+      "step": 7905
+    },
+    {
+      "epoch": 15.812,
+      "grad_norm": 1.855078935623169,
+      "learning_rate": 2e-05,
+      "loss": 0.06233601,
+      "step": 7906
+    },
+    {
+      "epoch": 15.814,
+      "grad_norm": 1.4420384168624878,
+      "learning_rate": 2e-05,
+      "loss": 0.04433277,
+      "step": 7907
+    },
+    {
+      "epoch": 15.816,
+      "grad_norm": 1.6261996030807495,
+      "learning_rate": 2e-05,
+      "loss": 0.05150939,
+      "step": 7908
+    },
+    {
+      "epoch": 15.818,
+      "grad_norm": 0.9924352765083313,
+      "learning_rate": 2e-05,
+      "loss": 0.03078244,
+      "step": 7909
+    },
+    {
+      "epoch": 15.82,
+      "grad_norm": 1.1630253791809082,
+      "learning_rate": 2e-05,
+      "loss": 0.02466009,
+      "step": 7910
+    },
+    {
+      "epoch": 15.822,
+      "grad_norm": 1.1106728315353394,
+      "learning_rate": 2e-05,
+      "loss": 0.04763328,
+      "step": 7911
+    },
+    {
+      "epoch": 15.824,
+      "grad_norm": 1.9522145986557007,
+      "learning_rate": 2e-05,
+      "loss": 0.04942587,
+      "step": 7912
+    },
+    {
+      "epoch": 15.826,
+      "grad_norm": 0.9739556312561035,
+      "learning_rate": 2e-05,
+      "loss": 0.03213318,
+      "step": 7913
+    },
+    {
+      "epoch": 15.828,
+      "grad_norm": 2.2910735607147217,
+      "learning_rate": 2e-05,
+      "loss": 0.04557914,
+      "step": 7914
+    },
+    {
+      "epoch": 15.83,
+      "grad_norm": 2.2116222381591797,
+      "learning_rate": 2e-05,
+      "loss": 0.04482364,
+      "step": 7915
+    },
+    {
+      "epoch": 15.832,
+      "grad_norm": 1.637697696685791,
+      "learning_rate": 2e-05,
+      "loss": 0.03125627,
+      "step": 7916
+    },
+    {
+      "epoch": 15.834,
+      "grad_norm": 1.0995755195617676,
+      "learning_rate": 2e-05,
+      "loss": 0.03378412,
+      "step": 7917
+    },
+    {
+      "epoch": 15.836,
+      "grad_norm": 2.1459877490997314,
+      "learning_rate": 2e-05,
+      "loss": 0.03629637,
+      "step": 7918
+    },
+    {
+      "epoch": 15.838,
+      "grad_norm": 1.2889842987060547,
+      "learning_rate": 2e-05,
+      "loss": 0.04306239,
+      "step": 7919
+    },
+    {
+      "epoch": 15.84,
+      "grad_norm": 4.523320198059082,
+      "learning_rate": 2e-05,
+      "loss": 0.04296661,
+      "step": 7920
+    },
+    {
+      "epoch": 15.842,
+      "grad_norm": 1.7894271612167358,
+      "learning_rate": 2e-05,
+      "loss": 0.03566499,
+      "step": 7921
+    },
+    {
+      "epoch": 15.844,
+      "grad_norm": 1.401320219039917,
+      "learning_rate": 2e-05,
+      "loss": 0.02614049,
+      "step": 7922
+    },
+    {
+      "epoch": 15.846,
+      "grad_norm": 1.4759875535964966,
+      "learning_rate": 2e-05,
+      "loss": 0.02819363,
+      "step": 7923
+    },
+    {
+      "epoch": 15.848,
+      "grad_norm": 1.3998092412948608,
+      "learning_rate": 2e-05,
+      "loss": 0.03906362,
+      "step": 7924
+    },
+    {
+      "epoch": 15.85,
+      "grad_norm": 1.1265228986740112,
+      "learning_rate": 2e-05,
+      "loss": 0.03560297,
+      "step": 7925
+    },
+    {
+      "epoch": 15.852,
+      "grad_norm": 2.067176342010498,
+      "learning_rate": 2e-05,
+      "loss": 0.04355159,
+      "step": 7926
+    },
+    {
+      "epoch": 15.854,
+      "grad_norm": 1.7233660221099854,
+      "learning_rate": 2e-05,
+      "loss": 0.029684,
+      "step": 7927
+    },
+    {
+      "epoch": 15.856,
+      "grad_norm": 2.1673214435577393,
+      "learning_rate": 2e-05,
+      "loss": 0.04684749,
+      "step": 7928
+    },
+    {
+      "epoch": 15.858,
+      "grad_norm": 3.047663450241089,
+      "learning_rate": 2e-05,
+      "loss": 0.04030456,
+      "step": 7929
+    },
+    {
+      "epoch": 15.86,
+      "grad_norm": 1.677051305770874,
+      "learning_rate": 2e-05,
+      "loss": 0.03150005,
+      "step": 7930
+    },
+    {
+      "epoch": 15.862,
+      "grad_norm": 1.5520917177200317,
+      "learning_rate": 2e-05,
+      "loss": 0.03935989,
+      "step": 7931
+    },
+    {
+      "epoch": 15.864,
+      "grad_norm": 2.1706817150115967,
+      "learning_rate": 2e-05,
+      "loss": 0.03663088,
+      "step": 7932
+    },
+    {
+      "epoch": 15.866,
+      "grad_norm": 1.029218316078186,
+      "learning_rate": 2e-05,
+      "loss": 0.03442633,
+      "step": 7933
+    },
+    {
+      "epoch": 15.868,
+      "grad_norm": 0.9123724699020386,
+      "learning_rate": 2e-05,
+      "loss": 0.02511028,
+      "step": 7934
+    },
+    {
+      "epoch": 15.87,
+      "grad_norm": 1.1108697652816772,
+      "learning_rate": 2e-05,
+      "loss": 0.04436678,
+      "step": 7935
+    },
+    {
+      "epoch": 15.872,
+      "grad_norm": 1.1441642045974731,
+      "learning_rate": 2e-05,
+      "loss": 0.03152635,
+      "step": 7936
+    },
+    {
+      "epoch": 15.874,
+      "grad_norm": 0.9864327907562256,
+      "learning_rate": 2e-05,
+      "loss": 0.02202858,
+      "step": 7937
+    },
+    {
+      "epoch": 15.876,
+      "grad_norm": 0.9307705760002136,
+      "learning_rate": 2e-05,
+      "loss": 0.02411098,
+      "step": 7938
+    },
+    {
+      "epoch": 15.878,
+      "grad_norm": 1.2462923526763916,
+      "learning_rate": 2e-05,
+      "loss": 0.03807527,
+      "step": 7939
+    },
+    {
+      "epoch": 15.88,
+      "grad_norm": 0.9074599146842957,
+      "learning_rate": 2e-05,
+      "loss": 0.02559845,
+      "step": 7940
+    },
+    {
+      "epoch": 15.882,
+      "grad_norm": 1.3017256259918213,
+      "learning_rate": 2e-05,
+      "loss": 0.04016859,
+      "step": 7941
+    },
+    {
+      "epoch": 15.884,
+      "grad_norm": 2.846317768096924,
+      "learning_rate": 2e-05,
+      "loss": 0.0447731,
+      "step": 7942
+    },
+    {
+      "epoch": 15.886,
+      "grad_norm": 2.893179416656494,
+      "learning_rate": 2e-05,
+      "loss": 0.05213889,
+      "step": 7943
+    },
+    {
+      "epoch": 15.888,
+      "grad_norm": 1.6494814157485962,
+      "learning_rate": 2e-05,
+      "loss": 0.0391501,
+      "step": 7944
+    },
+    {
+      "epoch": 15.89,
+      "grad_norm": 1.257524847984314,
+      "learning_rate": 2e-05,
+      "loss": 0.05023734,
+      "step": 7945
+    },
+    {
+      "epoch": 15.892,
+      "grad_norm": 1.13717782497406,
+      "learning_rate": 2e-05,
+      "loss": 0.0257063,
+      "step": 7946
+    },
+    {
+      "epoch": 15.894,
+      "grad_norm": 1.18727707862854,
+      "learning_rate": 2e-05,
+      "loss": 0.03686336,
+      "step": 7947
+    },
+    {
+      "epoch": 15.896,
+      "grad_norm": 1.3855704069137573,
+      "learning_rate": 2e-05,
+      "loss": 0.03288057,
+      "step": 7948
+    },
+    {
+      "epoch": 15.898,
+      "grad_norm": 1.4399868249893188,
+      "learning_rate": 2e-05,
+      "loss": 0.04080542,
+      "step": 7949
+    },
+    {
+      "epoch": 15.9,
+      "grad_norm": 1.5103329420089722,
+      "learning_rate": 2e-05,
+      "loss": 0.04284345,
+      "step": 7950
+    },
+    {
+      "epoch": 15.902,
+      "grad_norm": 1.8133221864700317,
+      "learning_rate": 2e-05,
+      "loss": 0.04338891,
+      "step": 7951
+    },
+    {
+      "epoch": 15.904,
+      "grad_norm": 1.0757697820663452,
+      "learning_rate": 2e-05,
+      "loss": 0.03184106,
+      "step": 7952
+    },
+    {
+      "epoch": 15.906,
+      "grad_norm": 1.2824019193649292,
+      "learning_rate": 2e-05,
+      "loss": 0.03566642,
+      "step": 7953
+    },
+    {
+      "epoch": 15.908,
+      "grad_norm": 1.087025761604309,
+      "learning_rate": 2e-05,
+      "loss": 0.02856821,
+      "step": 7954
+    },
+    {
+      "epoch": 15.91,
+      "grad_norm": 3.879120349884033,
+      "learning_rate": 2e-05,
+      "loss": 0.0441326,
+      "step": 7955
+    },
+    {
+      "epoch": 15.912,
+      "grad_norm": 1.081153392791748,
+      "learning_rate": 2e-05,
+      "loss": 0.04057573,
+      "step": 7956
+    },
+    {
+      "epoch": 15.914,
+      "grad_norm": 2.346388578414917,
+      "learning_rate": 2e-05,
+      "loss": 0.04326249,
+      "step": 7957
+    },
+    {
+      "epoch": 15.916,
+      "grad_norm": 1.8399007320404053,
+      "learning_rate": 2e-05,
+      "loss": 0.05309239,
+      "step": 7958
+    },
+    {
+      "epoch": 15.918,
+      "grad_norm": 1.1926047801971436,
+      "learning_rate": 2e-05,
+      "loss": 0.02422775,
+      "step": 7959
+    },
+    {
+      "epoch": 15.92,
+      "grad_norm": 2.166907787322998,
+      "learning_rate": 2e-05,
+      "loss": 0.03924531,
+      "step": 7960
+    },
+    {
+      "epoch": 15.922,
+      "grad_norm": 2.182678461074829,
+      "learning_rate": 2e-05,
+      "loss": 0.03237635,
+      "step": 7961
+    },
+    {
+      "epoch": 15.924,
+      "grad_norm": 1.284359335899353,
+      "learning_rate": 2e-05,
+      "loss": 0.03156743,
+      "step": 7962
+    },
+    {
+      "epoch": 15.926,
+      "grad_norm": 2.093848705291748,
+      "learning_rate": 2e-05,
+      "loss": 0.05261733,
+      "step": 7963
+    },
+    {
+      "epoch": 15.928,
+      "grad_norm": 2.00311541557312,
+      "learning_rate": 2e-05,
+      "loss": 0.03726339,
+      "step": 7964
+    },
+    {
+      "epoch": 15.93,
+      "grad_norm": 1.8436921834945679,
+      "learning_rate": 2e-05,
+      "loss": 0.05885259,
+      "step": 7965
+    },
+    {
+      "epoch": 15.932,
+      "grad_norm": 1.5446103811264038,
+      "learning_rate": 2e-05,
+      "loss": 0.0387977,
+      "step": 7966
+    },
+    {
+      "epoch": 15.934,
+      "grad_norm": 1.519590973854065,
+      "learning_rate": 2e-05,
+      "loss": 0.03424424,
+      "step": 7967
+    },
+    {
+      "epoch": 15.936,
+      "grad_norm": 1.6442177295684814,
+      "learning_rate": 2e-05,
+      "loss": 0.05894388,
+      "step": 7968
+    },
+    {
+      "epoch": 15.938,
+      "grad_norm": 2.041292905807495,
+      "learning_rate": 2e-05,
+      "loss": 0.04844762,
+      "step": 7969
+    },
+    {
+      "epoch": 15.94,
+      "grad_norm": 1.1668561697006226,
+      "learning_rate": 2e-05,
+      "loss": 0.03171788,
+      "step": 7970
+    },
+    {
+      "epoch": 15.942,
+      "grad_norm": 2.4591891765594482,
+      "learning_rate": 2e-05,
+      "loss": 0.05597981,
+      "step": 7971
+    },
+    {
+      "epoch": 15.943999999999999,
+      "grad_norm": 1.7161046266555786,
+      "learning_rate": 2e-05,
+      "loss": 0.04831922,
+      "step": 7972
+    },
+    {
+      "epoch": 15.946,
+      "grad_norm": 1.3476585149765015,
+      "learning_rate": 2e-05,
+      "loss": 0.03861203,
+      "step": 7973
+    },
+    {
+      "epoch": 15.948,
+      "grad_norm": 1.5570892095565796,
+      "learning_rate": 2e-05,
+      "loss": 0.04717612,
+      "step": 7974
+    },
+    {
+      "epoch": 15.95,
+      "grad_norm": 2.5573222637176514,
+      "learning_rate": 2e-05,
+      "loss": 0.04429121,
+      "step": 7975
+    },
+    {
+      "epoch": 15.952,
+      "grad_norm": 1.9069948196411133,
+      "learning_rate": 2e-05,
+      "loss": 0.04974307,
+      "step": 7976
+    },
+    {
+      "epoch": 15.954,
+      "grad_norm": 3.160170316696167,
+      "learning_rate": 2e-05,
+      "loss": 0.03945491,
+      "step": 7977
+    },
+    {
+      "epoch": 15.956,
+      "grad_norm": 1.0503807067871094,
+      "learning_rate": 2e-05,
+      "loss": 0.03511816,
+      "step": 7978
+    },
+    {
+      "epoch": 15.958,
+      "grad_norm": 1.0775737762451172,
+      "learning_rate": 2e-05,
+      "loss": 0.03104691,
+      "step": 7979
+    },
+    {
+      "epoch": 15.96,
+      "grad_norm": 1.1653412580490112,
+      "learning_rate": 2e-05,
+      "loss": 0.04440337,
+      "step": 7980
+    },
+    {
+      "epoch": 15.962,
+      "grad_norm": 1.088532567024231,
+      "learning_rate": 2e-05,
+      "loss": 0.04312788,
+      "step": 7981
+    },
+    {
+      "epoch": 15.964,
+      "grad_norm": 1.1584064960479736,
+      "learning_rate": 2e-05,
+      "loss": 0.02283434,
+      "step": 7982
+    },
+    {
+      "epoch": 15.966,
+      "grad_norm": 1.043766975402832,
+      "learning_rate": 2e-05,
+      "loss": 0.02896805,
+      "step": 7983
+    },
+    {
+      "epoch": 15.968,
+      "grad_norm": 0.9442967176437378,
+      "learning_rate": 2e-05,
+      "loss": 0.02886542,
+      "step": 7984
+    },
+    {
+      "epoch": 15.97,
+      "grad_norm": 1.3705523014068604,
+      "learning_rate": 2e-05,
+      "loss": 0.04226013,
+      "step": 7985
+    },
+    {
+      "epoch": 15.972,
+      "grad_norm": 3.2384965419769287,
+      "learning_rate": 2e-05,
+      "loss": 0.05349871,
+      "step": 7986
+    },
+    {
+      "epoch": 15.974,
+      "grad_norm": 1.393133282661438,
+      "learning_rate": 2e-05,
+      "loss": 0.02859312,
+      "step": 7987
+    },
+    {
+      "epoch": 15.975999999999999,
+      "grad_norm": 1.0506641864776611,
+      "learning_rate": 2e-05,
+      "loss": 0.03882669,
+      "step": 7988
+    },
+    {
+      "epoch": 15.978,
+      "grad_norm": 1.3604516983032227,
+      "learning_rate": 2e-05,
+      "loss": 0.04398809,
+      "step": 7989
+    },
+    {
+      "epoch": 15.98,
+      "grad_norm": 1.3526203632354736,
+      "learning_rate": 2e-05,
+      "loss": 0.03669821,
+      "step": 7990
+    },
+    {
+      "epoch": 15.982,
+      "grad_norm": 1.1345425844192505,
+      "learning_rate": 2e-05,
+      "loss": 0.03266115,
+      "step": 7991
+    },
+    {
+      "epoch": 15.984,
+      "grad_norm": 1.23661470413208,
+      "learning_rate": 2e-05,
+      "loss": 0.02663805,
+      "step": 7992
+    },
+    {
+      "epoch": 15.986,
+      "grad_norm": 2.231520652770996,
+      "learning_rate": 2e-05,
+      "loss": 0.034388,
+      "step": 7993
+    },
+    {
+      "epoch": 15.988,
+      "grad_norm": 1.1724952459335327,
+      "learning_rate": 2e-05,
+      "loss": 0.0357694,
+      "step": 7994
+    },
+    {
+      "epoch": 15.99,
+      "grad_norm": 1.698375940322876,
+      "learning_rate": 2e-05,
+      "loss": 0.04616644,
+      "step": 7995
+    },
+    {
+      "epoch": 15.992,
+      "grad_norm": 1.5685293674468994,
+      "learning_rate": 2e-05,
+      "loss": 0.03160156,
+      "step": 7996
+    },
+    {
+      "epoch": 15.994,
+      "grad_norm": 2.4769506454467773,
+      "learning_rate": 2e-05,
+      "loss": 0.04571313,
+      "step": 7997
+    },
+    {
+      "epoch": 15.996,
+      "grad_norm": 1.7058403491973877,
+      "learning_rate": 2e-05,
+      "loss": 0.03262362,
+      "step": 7998
+    },
+    {
+      "epoch": 15.998,
+      "grad_norm": 1.0402840375900269,
+      "learning_rate": 2e-05,
+      "loss": 0.02683425,
+      "step": 7999
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 1.4686156511306763,
+      "learning_rate": 2e-05,
+      "loss": 0.03669425,
+      "step": 8000
+    },
+    {
+      "epoch": 16.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.996,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9481037924151696,
+        "Equal_1": 0.994,
+        "Equal_2": 0.9660678642714571,
+        "Equal_3": 0.8902195608782435,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9940119760479041,
+        "Parallel_1": 0.9879759519038076,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.986,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.982,
+        "Perpendicular_3": 0.7274549098196392,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9996666666666667,
+        "PointLiesOnCircle_3": 0.9912666666666666,
+        "PointLiesOnLine_1": 0.9959919839679359,
+        "PointLiesOnLine_2": 0.9939879759519038,
+        "PointLiesOnLine_3": 0.9780439121756487
+      },
+      "eval_runtime": 319.9457,
+      "eval_samples_per_second": 32.818,
+      "eval_steps_per_second": 0.656,
+      "step": 8000
+    },
+    {
+      "epoch": 16.002,
+      "grad_norm": 1.1055775880813599,
+      "learning_rate": 2e-05,
+      "loss": 0.0335177,
+      "step": 8001
+    },
+    {
+      "epoch": 16.004,
+      "grad_norm": 1.2921119928359985,
+      "learning_rate": 2e-05,
+      "loss": 0.02823064,
+      "step": 8002
+    },
+    {
+      "epoch": 16.006,
+      "grad_norm": 2.824982166290283,
+      "learning_rate": 2e-05,
+      "loss": 0.02834114,
+      "step": 8003
+    },
+    {
+      "epoch": 16.008,
+      "grad_norm": 2.268012523651123,
+      "learning_rate": 2e-05,
+      "loss": 0.03194273,
+      "step": 8004
+    },
+    {
+      "epoch": 16.01,
+      "grad_norm": 1.4545838832855225,
+      "learning_rate": 2e-05,
+      "loss": 0.02953038,
+      "step": 8005
+    },
+    {
+      "epoch": 16.012,
+      "grad_norm": 1.845914363861084,
+      "learning_rate": 2e-05,
+      "loss": 0.03039683,
+      "step": 8006
+    },
+    {
+      "epoch": 16.014,
+      "grad_norm": 1.2593659162521362,
+      "learning_rate": 2e-05,
+      "loss": 0.03985323,
+      "step": 8007
+    },
+    {
+      "epoch": 16.016,
+      "grad_norm": 1.1578806638717651,
+      "learning_rate": 2e-05,
+      "loss": 0.04368738,
+      "step": 8008
+    },
+    {
+      "epoch": 16.018,
+      "grad_norm": 2.670903444290161,
+      "learning_rate": 2e-05,
+      "loss": 0.05551417,
+      "step": 8009
+    },
+    {
+      "epoch": 16.02,
+      "grad_norm": 1.4139609336853027,
+      "learning_rate": 2e-05,
+      "loss": 0.03750867,
+      "step": 8010
+    },
+    {
+      "epoch": 16.022,
+      "grad_norm": 1.3667073249816895,
+      "learning_rate": 2e-05,
+      "loss": 0.05370542,
+      "step": 8011
+    },
+    {
+      "epoch": 16.024,
+      "grad_norm": 2.2714273929595947,
+      "learning_rate": 2e-05,
+      "loss": 0.0414988,
+      "step": 8012
+    },
+    {
+      "epoch": 16.026,
+      "grad_norm": 2.840731382369995,
+      "learning_rate": 2e-05,
+      "loss": 0.03340321,
+      "step": 8013
+    },
+    {
+      "epoch": 16.028,
+      "grad_norm": 2.810915946960449,
+      "learning_rate": 2e-05,
+      "loss": 0.05696411,
+      "step": 8014
+    },
+    {
+      "epoch": 16.03,
+      "grad_norm": 1.45564866065979,
+      "learning_rate": 2e-05,
+      "loss": 0.04313691,
+      "step": 8015
+    },
+    {
+      "epoch": 16.032,
+      "grad_norm": 1.4301260709762573,
+      "learning_rate": 2e-05,
+      "loss": 0.04525199,
+      "step": 8016
+    },
+    {
+      "epoch": 16.034,
+      "grad_norm": 1.1844078302383423,
+      "learning_rate": 2e-05,
+      "loss": 0.0316691,
+      "step": 8017
+    },
+    {
+      "epoch": 16.036,
+      "grad_norm": 1.4068323373794556,
+      "learning_rate": 2e-05,
+      "loss": 0.03633066,
+      "step": 8018
+    },
+    {
+      "epoch": 16.038,
+      "grad_norm": 1.5108731985092163,
+      "learning_rate": 2e-05,
+      "loss": 0.04725716,
+      "step": 8019
+    },
+    {
+      "epoch": 16.04,
+      "grad_norm": 1.654616355895996,
+      "learning_rate": 2e-05,
+      "loss": 0.04446092,
+      "step": 8020
+    },
+    {
+      "epoch": 16.042,
+      "grad_norm": 1.4129464626312256,
+      "learning_rate": 2e-05,
+      "loss": 0.03359355,
+      "step": 8021
+    },
+    {
+      "epoch": 16.044,
+      "grad_norm": 1.0702934265136719,
+      "learning_rate": 2e-05,
+      "loss": 0.02224928,
+      "step": 8022
+    },
+    {
+      "epoch": 16.046,
+      "grad_norm": 1.9897451400756836,
+      "learning_rate": 2e-05,
+      "loss": 0.04445896,
+      "step": 8023
+    },
+    {
+      "epoch": 16.048,
+      "grad_norm": 1.2325469255447388,
+      "learning_rate": 2e-05,
+      "loss": 0.03780691,
+      "step": 8024
+    },
+    {
+      "epoch": 16.05,
+      "grad_norm": 1.2062650918960571,
+      "learning_rate": 2e-05,
+      "loss": 0.03396706,
+      "step": 8025
+    },
+    {
+      "epoch": 16.052,
+      "grad_norm": 1.6414982080459595,
+      "learning_rate": 2e-05,
+      "loss": 0.04513337,
+      "step": 8026
+    },
+    {
+      "epoch": 16.054,
+      "grad_norm": 3.0704195499420166,
+      "learning_rate": 2e-05,
+      "loss": 0.05739868,
+      "step": 8027
+    },
+    {
+      "epoch": 16.056,
+      "grad_norm": 1.746216893196106,
+      "learning_rate": 2e-05,
+      "loss": 0.04020358,
+      "step": 8028
+    },
+    {
+      "epoch": 16.058,
+      "grad_norm": 1.0398463010787964,
+      "learning_rate": 2e-05,
+      "loss": 0.02914593,
+      "step": 8029
+    },
+    {
+      "epoch": 16.06,
+      "grad_norm": 1.3823471069335938,
+      "learning_rate": 2e-05,
+      "loss": 0.0427549,
+      "step": 8030
+    },
+    {
+      "epoch": 16.062,
+      "grad_norm": 0.9713700413703918,
+      "learning_rate": 2e-05,
+      "loss": 0.02133579,
+      "step": 8031
+    },
+    {
+      "epoch": 16.064,
+      "grad_norm": 2.3745758533477783,
+      "learning_rate": 2e-05,
+      "loss": 0.03649534,
+      "step": 8032
+    },
+    {
+      "epoch": 16.066,
+      "grad_norm": 1.386365294456482,
+      "learning_rate": 2e-05,
+      "loss": 0.04500858,
+      "step": 8033
+    },
+    {
+      "epoch": 16.068,
+      "grad_norm": 2.186955690383911,
+      "learning_rate": 2e-05,
+      "loss": 0.05879531,
+      "step": 8034
+    },
+    {
+      "epoch": 16.07,
+      "grad_norm": 2.0882043838500977,
+      "learning_rate": 2e-05,
+      "loss": 0.04632348,
+      "step": 8035
+    },
+    {
+      "epoch": 16.072,
+      "grad_norm": 1.3914663791656494,
+      "learning_rate": 2e-05,
+      "loss": 0.03216641,
+      "step": 8036
+    },
+    {
+      "epoch": 16.074,
+      "grad_norm": 1.4008915424346924,
+      "learning_rate": 2e-05,
+      "loss": 0.03341067,
+      "step": 8037
+    },
+    {
+      "epoch": 16.076,
+      "grad_norm": 1.3791346549987793,
+      "learning_rate": 2e-05,
+      "loss": 0.0505546,
+      "step": 8038
+    },
+    {
+      "epoch": 16.078,
+      "grad_norm": 1.302667498588562,
+      "learning_rate": 2e-05,
+      "loss": 0.04582535,
+      "step": 8039
+    },
+    {
+      "epoch": 16.08,
+      "grad_norm": 1.6824592351913452,
+      "learning_rate": 2e-05,
+      "loss": 0.0420664,
+      "step": 8040
+    },
+    {
+      "epoch": 16.082,
+      "grad_norm": 1.354354977607727,
+      "learning_rate": 2e-05,
+      "loss": 0.02892293,
+      "step": 8041
+    },
+    {
+      "epoch": 16.084,
+      "grad_norm": 1.0839284658432007,
+      "learning_rate": 2e-05,
+      "loss": 0.03057658,
+      "step": 8042
+    },
+    {
+      "epoch": 16.086,
+      "grad_norm": 1.1779735088348389,
+      "learning_rate": 2e-05,
+      "loss": 0.02183773,
+      "step": 8043
+    },
+    {
+      "epoch": 16.088,
+      "grad_norm": 1.1066539287567139,
+      "learning_rate": 2e-05,
+      "loss": 0.03328647,
+      "step": 8044
+    },
+    {
+      "epoch": 16.09,
+      "grad_norm": 1.5935084819793701,
+      "learning_rate": 2e-05,
+      "loss": 0.04813398,
+      "step": 8045
+    },
+    {
+      "epoch": 16.092,
+      "grad_norm": 1.7510586977005005,
+      "learning_rate": 2e-05,
+      "loss": 0.03419684,
+      "step": 8046
+    },
+    {
+      "epoch": 16.094,
+      "grad_norm": 1.2361801862716675,
+      "learning_rate": 2e-05,
+      "loss": 0.03396645,
+      "step": 8047
+    },
+    {
+      "epoch": 16.096,
+      "grad_norm": 1.069996953010559,
+      "learning_rate": 2e-05,
+      "loss": 0.02971686,
+      "step": 8048
+    },
+    {
+      "epoch": 16.098,
+      "grad_norm": 1.4288445711135864,
+      "learning_rate": 2e-05,
+      "loss": 0.03799113,
+      "step": 8049
+    },
+    {
+      "epoch": 16.1,
+      "grad_norm": 0.9517697691917419,
+      "learning_rate": 2e-05,
+      "loss": 0.02629692,
+      "step": 8050
+    },
+    {
+      "epoch": 16.102,
+      "grad_norm": 1.0445160865783691,
+      "learning_rate": 2e-05,
+      "loss": 0.03541003,
+      "step": 8051
+    },
+    {
+      "epoch": 16.104,
+      "grad_norm": 1.2751644849777222,
+      "learning_rate": 2e-05,
+      "loss": 0.03903734,
+      "step": 8052
+    },
+    {
+      "epoch": 16.106,
+      "grad_norm": 1.3192799091339111,
+      "learning_rate": 2e-05,
+      "loss": 0.04088945,
+      "step": 8053
+    },
+    {
+      "epoch": 16.108,
+      "grad_norm": 0.8523271679878235,
+      "learning_rate": 2e-05,
+      "loss": 0.02510538,
+      "step": 8054
+    },
+    {
+      "epoch": 16.11,
+      "grad_norm": 1.4209495782852173,
+      "learning_rate": 2e-05,
+      "loss": 0.03651667,
+      "step": 8055
+    },
+    {
+      "epoch": 16.112,
+      "grad_norm": 2.66572904586792,
+      "learning_rate": 2e-05,
+      "loss": 0.02939776,
+      "step": 8056
+    },
+    {
+      "epoch": 16.114,
+      "grad_norm": 1.7730138301849365,
+      "learning_rate": 2e-05,
+      "loss": 0.05316716,
+      "step": 8057
+    },
+    {
+      "epoch": 16.116,
+      "grad_norm": 1.0424127578735352,
+      "learning_rate": 2e-05,
+      "loss": 0.03889843,
+      "step": 8058
+    },
+    {
+      "epoch": 16.118,
+      "grad_norm": 1.0265769958496094,
+      "learning_rate": 2e-05,
+      "loss": 0.03288617,
+      "step": 8059
+    },
+    {
+      "epoch": 16.12,
+      "grad_norm": 1.0434588193893433,
+      "learning_rate": 2e-05,
+      "loss": 0.04037582,
+      "step": 8060
+    },
+    {
+      "epoch": 16.122,
+      "grad_norm": 1.2070543766021729,
+      "learning_rate": 2e-05,
+      "loss": 0.02952406,
+      "step": 8061
+    },
+    {
+      "epoch": 16.124,
+      "grad_norm": 1.59628427028656,
+      "learning_rate": 2e-05,
+      "loss": 0.04046272,
+      "step": 8062
+    },
+    {
+      "epoch": 16.126,
+      "grad_norm": 1.1951675415039062,
+      "learning_rate": 2e-05,
+      "loss": 0.03478213,
+      "step": 8063
+    },
+    {
+      "epoch": 16.128,
+      "grad_norm": 1.0106141567230225,
+      "learning_rate": 2e-05,
+      "loss": 0.03325225,
+      "step": 8064
+    },
+    {
+      "epoch": 16.13,
+      "grad_norm": 0.993449330329895,
+      "learning_rate": 2e-05,
+      "loss": 0.01964446,
+      "step": 8065
+    },
+    {
+      "epoch": 16.132,
+      "grad_norm": 1.4265450239181519,
+      "learning_rate": 2e-05,
+      "loss": 0.03322139,
+      "step": 8066
+    },
+    {
+      "epoch": 16.134,
+      "grad_norm": 1.0770517587661743,
+      "learning_rate": 2e-05,
+      "loss": 0.03661602,
+      "step": 8067
+    },
+    {
+      "epoch": 16.136,
+      "grad_norm": 1.4938647747039795,
+      "learning_rate": 2e-05,
+      "loss": 0.03427796,
+      "step": 8068
+    },
+    {
+      "epoch": 16.138,
+      "grad_norm": 1.4040374755859375,
+      "learning_rate": 2e-05,
+      "loss": 0.03253736,
+      "step": 8069
+    },
+    {
+      "epoch": 16.14,
+      "grad_norm": 1.381060242652893,
+      "learning_rate": 2e-05,
+      "loss": 0.03683833,
+      "step": 8070
+    },
+    {
+      "epoch": 16.142,
+      "grad_norm": 1.1228476762771606,
+      "learning_rate": 2e-05,
+      "loss": 0.04204592,
+      "step": 8071
+    },
+    {
+      "epoch": 16.144,
+      "grad_norm": 1.1399726867675781,
+      "learning_rate": 2e-05,
+      "loss": 0.0393802,
+      "step": 8072
+    },
+    {
+      "epoch": 16.146,
+      "grad_norm": 1.6664113998413086,
+      "learning_rate": 2e-05,
+      "loss": 0.03525341,
+      "step": 8073
+    },
+    {
+      "epoch": 16.148,
+      "grad_norm": 1.0361666679382324,
+      "learning_rate": 2e-05,
+      "loss": 0.03461679,
+      "step": 8074
+    },
+    {
+      "epoch": 16.15,
+      "grad_norm": 1.3984792232513428,
+      "learning_rate": 2e-05,
+      "loss": 0.04337152,
+      "step": 8075
+    },
+    {
+      "epoch": 16.152,
+      "grad_norm": 1.3185367584228516,
+      "learning_rate": 2e-05,
+      "loss": 0.04654099,
+      "step": 8076
+    },
+    {
+      "epoch": 16.154,
+      "grad_norm": 2.2266247272491455,
+      "learning_rate": 2e-05,
+      "loss": 0.05011526,
+      "step": 8077
+    },
+    {
+      "epoch": 16.156,
+      "grad_norm": 1.3377346992492676,
+      "learning_rate": 2e-05,
+      "loss": 0.04082262,
+      "step": 8078
+    },
+    {
+      "epoch": 16.158,
+      "grad_norm": 3.1692075729370117,
+      "learning_rate": 2e-05,
+      "loss": 0.05622632,
+      "step": 8079
+    },
+    {
+      "epoch": 16.16,
+      "grad_norm": 1.0942643880844116,
+      "learning_rate": 2e-05,
+      "loss": 0.02746512,
+      "step": 8080
+    },
+    {
+      "epoch": 16.162,
+      "grad_norm": 1.217752456665039,
+      "learning_rate": 2e-05,
+      "loss": 0.03617308,
+      "step": 8081
+    },
+    {
+      "epoch": 16.164,
+      "grad_norm": 1.8734045028686523,
+      "learning_rate": 2e-05,
+      "loss": 0.04952653,
+      "step": 8082
+    },
+    {
+      "epoch": 16.166,
+      "grad_norm": 1.50890052318573,
+      "learning_rate": 2e-05,
+      "loss": 0.04156789,
+      "step": 8083
+    },
+    {
+      "epoch": 16.168,
+      "grad_norm": 1.4338157176971436,
+      "learning_rate": 2e-05,
+      "loss": 0.05392047,
+      "step": 8084
+    },
+    {
+      "epoch": 16.17,
+      "grad_norm": 1.7338756322860718,
+      "learning_rate": 2e-05,
+      "loss": 0.04553423,
+      "step": 8085
+    },
+    {
+      "epoch": 16.172,
+      "grad_norm": 1.5352826118469238,
+      "learning_rate": 2e-05,
+      "loss": 0.04915183,
+      "step": 8086
+    },
+    {
+      "epoch": 16.174,
+      "grad_norm": 1.7722485065460205,
+      "learning_rate": 2e-05,
+      "loss": 0.03585491,
+      "step": 8087
+    },
+    {
+      "epoch": 16.176,
+      "grad_norm": 0.9967431426048279,
+      "learning_rate": 2e-05,
+      "loss": 0.02942163,
+      "step": 8088
+    },
+    {
+      "epoch": 16.178,
+      "grad_norm": 1.2177917957305908,
+      "learning_rate": 2e-05,
+      "loss": 0.04380299,
+      "step": 8089
+    },
+    {
+      "epoch": 16.18,
+      "grad_norm": 1.2559581995010376,
+      "learning_rate": 2e-05,
+      "loss": 0.02480926,
+      "step": 8090
+    },
+    {
+      "epoch": 16.182,
+      "grad_norm": 1.5222054719924927,
+      "learning_rate": 2e-05,
+      "loss": 0.04941765,
+      "step": 8091
+    },
+    {
+      "epoch": 16.184,
+      "grad_norm": 1.0515002012252808,
+      "learning_rate": 2e-05,
+      "loss": 0.04209872,
+      "step": 8092
+    },
+    {
+      "epoch": 16.186,
+      "grad_norm": 1.8783875703811646,
+      "learning_rate": 2e-05,
+      "loss": 0.03835857,
+      "step": 8093
+    },
+    {
+      "epoch": 16.188,
+      "grad_norm": 1.2935268878936768,
+      "learning_rate": 2e-05,
+      "loss": 0.05065227,
+      "step": 8094
+    },
+    {
+      "epoch": 16.19,
+      "grad_norm": 1.6219897270202637,
+      "learning_rate": 2e-05,
+      "loss": 0.02984177,
+      "step": 8095
+    },
+    {
+      "epoch": 16.192,
+      "grad_norm": 1.3317407369613647,
+      "learning_rate": 2e-05,
+      "loss": 0.0331154,
+      "step": 8096
+    },
+    {
+      "epoch": 16.194,
+      "grad_norm": 1.2899963855743408,
+      "learning_rate": 2e-05,
+      "loss": 0.03418003,
+      "step": 8097
+    },
+    {
+      "epoch": 16.196,
+      "grad_norm": 1.1565256118774414,
+      "learning_rate": 2e-05,
+      "loss": 0.03147539,
+      "step": 8098
+    },
+    {
+      "epoch": 16.198,
+      "grad_norm": 1.2315301895141602,
+      "learning_rate": 2e-05,
+      "loss": 0.0391296,
+      "step": 8099
+    },
+    {
+      "epoch": 16.2,
+      "grad_norm": 1.1125531196594238,
+      "learning_rate": 2e-05,
+      "loss": 0.03473473,
+      "step": 8100
+    },
+    {
+      "epoch": 16.202,
+      "grad_norm": 1.5574398040771484,
+      "learning_rate": 2e-05,
+      "loss": 0.0310582,
+      "step": 8101
+    },
+    {
+      "epoch": 16.204,
+      "grad_norm": 1.4571691751480103,
+      "learning_rate": 2e-05,
+      "loss": 0.03549339,
+      "step": 8102
+    },
+    {
+      "epoch": 16.206,
+      "grad_norm": 1.2197107076644897,
+      "learning_rate": 2e-05,
+      "loss": 0.04195347,
+      "step": 8103
+    },
+    {
+      "epoch": 16.208,
+      "grad_norm": 2.789921760559082,
+      "learning_rate": 2e-05,
+      "loss": 0.03652881,
+      "step": 8104
+    },
+    {
+      "epoch": 16.21,
+      "grad_norm": 4.851873397827148,
+      "learning_rate": 2e-05,
+      "loss": 0.03688388,
+      "step": 8105
+    },
+    {
+      "epoch": 16.212,
+      "grad_norm": 1.457202434539795,
+      "learning_rate": 2e-05,
+      "loss": 0.03209008,
+      "step": 8106
+    },
+    {
+      "epoch": 16.214,
+      "grad_norm": 1.0909388065338135,
+      "learning_rate": 2e-05,
+      "loss": 0.02918109,
+      "step": 8107
+    },
+    {
+      "epoch": 16.216,
+      "grad_norm": 2.0140578746795654,
+      "learning_rate": 2e-05,
+      "loss": 0.05105082,
+      "step": 8108
+    },
+    {
+      "epoch": 16.218,
+      "grad_norm": 1.9580172300338745,
+      "learning_rate": 2e-05,
+      "loss": 0.04800195,
+      "step": 8109
+    },
+    {
+      "epoch": 16.22,
+      "grad_norm": 3.5140187740325928,
+      "learning_rate": 2e-05,
+      "loss": 0.04388273,
+      "step": 8110
+    },
+    {
+      "epoch": 16.222,
+      "grad_norm": 1.116141676902771,
+      "learning_rate": 2e-05,
+      "loss": 0.03350087,
+      "step": 8111
+    },
+    {
+      "epoch": 16.224,
+      "grad_norm": 2.4265329837799072,
+      "learning_rate": 2e-05,
+      "loss": 0.04413512,
+      "step": 8112
+    },
+    {
+      "epoch": 16.226,
+      "grad_norm": 1.9018596410751343,
+      "learning_rate": 2e-05,
+      "loss": 0.0421547,
+      "step": 8113
+    },
+    {
+      "epoch": 16.228,
+      "grad_norm": 1.3331454992294312,
+      "learning_rate": 2e-05,
+      "loss": 0.04760406,
+      "step": 8114
+    },
+    {
+      "epoch": 16.23,
+      "grad_norm": 0.9366099834442139,
+      "learning_rate": 2e-05,
+      "loss": 0.02583786,
+      "step": 8115
+    },
+    {
+      "epoch": 16.232,
+      "grad_norm": 2.820237636566162,
+      "learning_rate": 2e-05,
+      "loss": 0.02830683,
+      "step": 8116
+    },
+    {
+      "epoch": 16.234,
+      "grad_norm": 1.1773334741592407,
+      "learning_rate": 2e-05,
+      "loss": 0.03467422,
+      "step": 8117
+    },
+    {
+      "epoch": 16.236,
+      "grad_norm": 1.685434103012085,
+      "learning_rate": 2e-05,
+      "loss": 0.04194804,
+      "step": 8118
+    },
+    {
+      "epoch": 16.238,
+      "grad_norm": 1.8741134405136108,
+      "learning_rate": 2e-05,
+      "loss": 0.0288431,
+      "step": 8119
+    },
+    {
+      "epoch": 16.24,
+      "grad_norm": 1.3268035650253296,
+      "learning_rate": 2e-05,
+      "loss": 0.0287002,
+      "step": 8120
+    },
+    {
+      "epoch": 16.242,
+      "grad_norm": 1.3486356735229492,
+      "learning_rate": 2e-05,
+      "loss": 0.03664606,
+      "step": 8121
+    },
+    {
+      "epoch": 16.244,
+      "grad_norm": 1.631987452507019,
+      "learning_rate": 2e-05,
+      "loss": 0.05182088,
+      "step": 8122
+    },
+    {
+      "epoch": 16.246,
+      "grad_norm": 2.88786244392395,
+      "learning_rate": 2e-05,
+      "loss": 0.05489088,
+      "step": 8123
+    },
+    {
+      "epoch": 16.248,
+      "grad_norm": 2.0187666416168213,
+      "learning_rate": 2e-05,
+      "loss": 0.04556795,
+      "step": 8124
+    },
+    {
+      "epoch": 16.25,
+      "grad_norm": 1.089003562927246,
+      "learning_rate": 2e-05,
+      "loss": 0.03417504,
+      "step": 8125
+    },
+    {
+      "epoch": 16.252,
+      "grad_norm": 1.238042950630188,
+      "learning_rate": 2e-05,
+      "loss": 0.03643361,
+      "step": 8126
+    },
+    {
+      "epoch": 16.254,
+      "grad_norm": 1.347186803817749,
+      "learning_rate": 2e-05,
+      "loss": 0.0338303,
+      "step": 8127
+    },
+    {
+      "epoch": 16.256,
+      "grad_norm": 1.772897720336914,
+      "learning_rate": 2e-05,
+      "loss": 0.05571172,
+      "step": 8128
+    },
+    {
+      "epoch": 16.258,
+      "grad_norm": 0.9345927834510803,
+      "learning_rate": 2e-05,
+      "loss": 0.02508531,
+      "step": 8129
+    },
+    {
+      "epoch": 16.26,
+      "grad_norm": 1.491607904434204,
+      "learning_rate": 2e-05,
+      "loss": 0.03616347,
+      "step": 8130
+    },
+    {
+      "epoch": 16.262,
+      "grad_norm": 1.141614317893982,
+      "learning_rate": 2e-05,
+      "loss": 0.03462991,
+      "step": 8131
+    },
+    {
+      "epoch": 16.264,
+      "grad_norm": 1.2147083282470703,
+      "learning_rate": 2e-05,
+      "loss": 0.03725596,
+      "step": 8132
+    },
+    {
+      "epoch": 16.266,
+      "grad_norm": 1.7106729745864868,
+      "learning_rate": 2e-05,
+      "loss": 0.03632404,
+      "step": 8133
+    },
+    {
+      "epoch": 16.268,
+      "grad_norm": 1.0277926921844482,
+      "learning_rate": 2e-05,
+      "loss": 0.0347311,
+      "step": 8134
+    },
+    {
+      "epoch": 16.27,
+      "grad_norm": 1.2781480550765991,
+      "learning_rate": 2e-05,
+      "loss": 0.03461632,
+      "step": 8135
+    },
+    {
+      "epoch": 16.272,
+      "grad_norm": 1.013649582862854,
+      "learning_rate": 2e-05,
+      "loss": 0.03419242,
+      "step": 8136
+    },
+    {
+      "epoch": 16.274,
+      "grad_norm": 1.7054558992385864,
+      "learning_rate": 2e-05,
+      "loss": 0.04766083,
+      "step": 8137
+    },
+    {
+      "epoch": 16.276,
+      "grad_norm": 1.1981877088546753,
+      "learning_rate": 2e-05,
+      "loss": 0.03875602,
+      "step": 8138
+    },
+    {
+      "epoch": 16.278,
+      "grad_norm": 1.4614924192428589,
+      "learning_rate": 2e-05,
+      "loss": 0.04386155,
+      "step": 8139
+    },
+    {
+      "epoch": 16.28,
+      "grad_norm": 0.8466894626617432,
+      "learning_rate": 2e-05,
+      "loss": 0.01991762,
+      "step": 8140
+    },
+    {
+      "epoch": 16.282,
+      "grad_norm": 2.6902008056640625,
+      "learning_rate": 2e-05,
+      "loss": 0.04461161,
+      "step": 8141
+    },
+    {
+      "epoch": 16.284,
+      "grad_norm": 1.712175965309143,
+      "learning_rate": 2e-05,
+      "loss": 0.04943609,
+      "step": 8142
+    },
+    {
+      "epoch": 16.286,
+      "grad_norm": 1.3116623163223267,
+      "learning_rate": 2e-05,
+      "loss": 0.03155245,
+      "step": 8143
+    },
+    {
+      "epoch": 16.288,
+      "grad_norm": 2.068293809890747,
+      "learning_rate": 2e-05,
+      "loss": 0.04026989,
+      "step": 8144
+    },
+    {
+      "epoch": 16.29,
+      "grad_norm": 2.457705020904541,
+      "learning_rate": 2e-05,
+      "loss": 0.05160169,
+      "step": 8145
+    },
+    {
+      "epoch": 16.292,
+      "grad_norm": 1.1721792221069336,
+      "learning_rate": 2e-05,
+      "loss": 0.03861858,
+      "step": 8146
+    },
+    {
+      "epoch": 16.294,
+      "grad_norm": 1.3003913164138794,
+      "learning_rate": 2e-05,
+      "loss": 0.03377434,
+      "step": 8147
+    },
+    {
+      "epoch": 16.296,
+      "grad_norm": 1.6648625135421753,
+      "learning_rate": 2e-05,
+      "loss": 0.05673098,
+      "step": 8148
+    },
+    {
+      "epoch": 16.298,
+      "grad_norm": 1.6256797313690186,
+      "learning_rate": 2e-05,
+      "loss": 0.03393706,
+      "step": 8149
+    },
+    {
+      "epoch": 16.3,
+      "grad_norm": 1.4245861768722534,
+      "learning_rate": 2e-05,
+      "loss": 0.03638397,
+      "step": 8150
+    },
+    {
+      "epoch": 16.302,
+      "grad_norm": 3.2517616748809814,
+      "learning_rate": 2e-05,
+      "loss": 0.04278335,
+      "step": 8151
+    },
+    {
+      "epoch": 16.304,
+      "grad_norm": 0.8710333108901978,
+      "learning_rate": 2e-05,
+      "loss": 0.02735967,
+      "step": 8152
+    },
+    {
+      "epoch": 16.306,
+      "grad_norm": 3.8048529624938965,
+      "learning_rate": 2e-05,
+      "loss": 0.04639032,
+      "step": 8153
+    },
+    {
+      "epoch": 16.308,
+      "grad_norm": 1.7357240915298462,
+      "learning_rate": 2e-05,
+      "loss": 0.04244914,
+      "step": 8154
+    },
+    {
+      "epoch": 16.31,
+      "grad_norm": 1.8791494369506836,
+      "learning_rate": 2e-05,
+      "loss": 0.04231622,
+      "step": 8155
+    },
+    {
+      "epoch": 16.312,
+      "grad_norm": 1.1524310111999512,
+      "learning_rate": 2e-05,
+      "loss": 0.03676322,
+      "step": 8156
+    },
+    {
+      "epoch": 16.314,
+      "grad_norm": 2.3864612579345703,
+      "learning_rate": 2e-05,
+      "loss": 0.04578413,
+      "step": 8157
+    },
+    {
+      "epoch": 16.316,
+      "grad_norm": 1.7564839124679565,
+      "learning_rate": 2e-05,
+      "loss": 0.04500315,
+      "step": 8158
+    },
+    {
+      "epoch": 16.318,
+      "grad_norm": 2.045438766479492,
+      "learning_rate": 2e-05,
+      "loss": 0.0306298,
+      "step": 8159
+    },
+    {
+      "epoch": 16.32,
+      "grad_norm": 0.8264663815498352,
+      "learning_rate": 2e-05,
+      "loss": 0.02070304,
+      "step": 8160
+    },
+    {
+      "epoch": 16.322,
+      "grad_norm": 1.1368653774261475,
+      "learning_rate": 2e-05,
+      "loss": 0.03928949,
+      "step": 8161
+    },
+    {
+      "epoch": 16.324,
+      "grad_norm": 0.9940206408500671,
+      "learning_rate": 2e-05,
+      "loss": 0.034535,
+      "step": 8162
+    },
+    {
+      "epoch": 16.326,
+      "grad_norm": 1.2441214323043823,
+      "learning_rate": 2e-05,
+      "loss": 0.0337163,
+      "step": 8163
+    },
+    {
+      "epoch": 16.328,
+      "grad_norm": 1.5334255695343018,
+      "learning_rate": 2e-05,
+      "loss": 0.04570875,
+      "step": 8164
+    },
+    {
+      "epoch": 16.33,
+      "grad_norm": 1.4286037683486938,
+      "learning_rate": 2e-05,
+      "loss": 0.05645541,
+      "step": 8165
+    },
+    {
+      "epoch": 16.332,
+      "grad_norm": 2.280270576477051,
+      "learning_rate": 2e-05,
+      "loss": 0.0485545,
+      "step": 8166
+    },
+    {
+      "epoch": 16.334,
+      "grad_norm": 1.9893189668655396,
+      "learning_rate": 2e-05,
+      "loss": 0.03430265,
+      "step": 8167
+    },
+    {
+      "epoch": 16.336,
+      "grad_norm": 1.4144779443740845,
+      "learning_rate": 2e-05,
+      "loss": 0.04236887,
+      "step": 8168
+    },
+    {
+      "epoch": 16.338,
+      "grad_norm": 1.0118204355239868,
+      "learning_rate": 2e-05,
+      "loss": 0.02984769,
+      "step": 8169
+    },
+    {
+      "epoch": 16.34,
+      "grad_norm": 1.6643238067626953,
+      "learning_rate": 2e-05,
+      "loss": 0.04789416,
+      "step": 8170
+    },
+    {
+      "epoch": 16.342,
+      "grad_norm": 0.8409438729286194,
+      "learning_rate": 2e-05,
+      "loss": 0.01973891,
+      "step": 8171
+    },
+    {
+      "epoch": 16.344,
+      "grad_norm": 1.0004379749298096,
+      "learning_rate": 2e-05,
+      "loss": 0.03158588,
+      "step": 8172
+    },
+    {
+      "epoch": 16.346,
+      "grad_norm": 2.194171667098999,
+      "learning_rate": 2e-05,
+      "loss": 0.06659742,
+      "step": 8173
+    },
+    {
+      "epoch": 16.348,
+      "grad_norm": 1.4098418951034546,
+      "learning_rate": 2e-05,
+      "loss": 0.03775009,
+      "step": 8174
+    },
+    {
+      "epoch": 16.35,
+      "grad_norm": 1.5424364805221558,
+      "learning_rate": 2e-05,
+      "loss": 0.04169576,
+      "step": 8175
+    },
+    {
+      "epoch": 16.352,
+      "grad_norm": 1.8581490516662598,
+      "learning_rate": 2e-05,
+      "loss": 0.07275459,
+      "step": 8176
+    },
+    {
+      "epoch": 16.354,
+      "grad_norm": 1.3709255456924438,
+      "learning_rate": 2e-05,
+      "loss": 0.03472374,
+      "step": 8177
+    },
+    {
+      "epoch": 16.356,
+      "grad_norm": 1.9072877168655396,
+      "learning_rate": 2e-05,
+      "loss": 0.0435157,
+      "step": 8178
+    },
+    {
+      "epoch": 16.358,
+      "grad_norm": 2.198082208633423,
+      "learning_rate": 2e-05,
+      "loss": 0.02928953,
+      "step": 8179
+    },
+    {
+      "epoch": 16.36,
+      "grad_norm": 0.9633461833000183,
+      "learning_rate": 2e-05,
+      "loss": 0.03266443,
+      "step": 8180
+    },
+    {
+      "epoch": 16.362,
+      "grad_norm": 2.41123366355896,
+      "learning_rate": 2e-05,
+      "loss": 0.04052667,
+      "step": 8181
+    },
+    {
+      "epoch": 16.364,
+      "grad_norm": 1.4949405193328857,
+      "learning_rate": 2e-05,
+      "loss": 0.05271856,
+      "step": 8182
+    },
+    {
+      "epoch": 16.366,
+      "grad_norm": 1.5633461475372314,
+      "learning_rate": 2e-05,
+      "loss": 0.04020818,
+      "step": 8183
+    },
+    {
+      "epoch": 16.368,
+      "grad_norm": 1.0727436542510986,
+      "learning_rate": 2e-05,
+      "loss": 0.03464281,
+      "step": 8184
+    },
+    {
+      "epoch": 16.37,
+      "grad_norm": 1.2527438402175903,
+      "learning_rate": 2e-05,
+      "loss": 0.04834267,
+      "step": 8185
+    },
+    {
+      "epoch": 16.372,
+      "grad_norm": 1.9610435962677002,
+      "learning_rate": 2e-05,
+      "loss": 0.05433511,
+      "step": 8186
+    },
+    {
+      "epoch": 16.374,
+      "grad_norm": 1.6253420114517212,
+      "learning_rate": 2e-05,
+      "loss": 0.04721621,
+      "step": 8187
+    },
+    {
+      "epoch": 16.376,
+      "grad_norm": 2.0063889026641846,
+      "learning_rate": 2e-05,
+      "loss": 0.04768866,
+      "step": 8188
+    },
+    {
+      "epoch": 16.378,
+      "grad_norm": 1.5373892784118652,
+      "learning_rate": 2e-05,
+      "loss": 0.04332329,
+      "step": 8189
+    },
+    {
+      "epoch": 16.38,
+      "grad_norm": 1.7891055345535278,
+      "learning_rate": 2e-05,
+      "loss": 0.0316738,
+      "step": 8190
+    },
+    {
+      "epoch": 16.382,
+      "grad_norm": 1.626381516456604,
+      "learning_rate": 2e-05,
+      "loss": 0.0384635,
+      "step": 8191
+    },
+    {
+      "epoch": 16.384,
+      "grad_norm": 1.2403956651687622,
+      "learning_rate": 2e-05,
+      "loss": 0.03793911,
+      "step": 8192
+    },
+    {
+      "epoch": 16.386,
+      "grad_norm": 1.388485074043274,
+      "learning_rate": 2e-05,
+      "loss": 0.05434792,
+      "step": 8193
+    },
+    {
+      "epoch": 16.388,
+      "grad_norm": 0.903146505355835,
+      "learning_rate": 2e-05,
+      "loss": 0.02984499,
+      "step": 8194
+    },
+    {
+      "epoch": 16.39,
+      "grad_norm": 1.14999258518219,
+      "learning_rate": 2e-05,
+      "loss": 0.0324681,
+      "step": 8195
+    },
+    {
+      "epoch": 16.392,
+      "grad_norm": 0.8718903660774231,
+      "learning_rate": 2e-05,
+      "loss": 0.03075159,
+      "step": 8196
+    },
+    {
+      "epoch": 16.394,
+      "grad_norm": 0.9332501888275146,
+      "learning_rate": 2e-05,
+      "loss": 0.03167652,
+      "step": 8197
+    },
+    {
+      "epoch": 16.396,
+      "grad_norm": 1.2964439392089844,
+      "learning_rate": 2e-05,
+      "loss": 0.05365559,
+      "step": 8198
+    },
+    {
+      "epoch": 16.398,
+      "grad_norm": 1.2190895080566406,
+      "learning_rate": 2e-05,
+      "loss": 0.04543667,
+      "step": 8199
+    },
+    {
+      "epoch": 16.4,
+      "grad_norm": 1.297627568244934,
+      "learning_rate": 2e-05,
+      "loss": 0.03421879,
+      "step": 8200
+    },
+    {
+      "epoch": 16.402,
+      "grad_norm": 1.3913570642471313,
+      "learning_rate": 2e-05,
+      "loss": 0.03622192,
+      "step": 8201
+    },
+    {
+      "epoch": 16.404,
+      "grad_norm": 0.967580258846283,
+      "learning_rate": 2e-05,
+      "loss": 0.03260601,
+      "step": 8202
+    },
+    {
+      "epoch": 16.406,
+      "grad_norm": 1.093234896659851,
+      "learning_rate": 2e-05,
+      "loss": 0.04168686,
+      "step": 8203
+    },
+    {
+      "epoch": 16.408,
+      "grad_norm": 1.2416380643844604,
+      "learning_rate": 2e-05,
+      "loss": 0.04545043,
+      "step": 8204
+    },
+    {
+      "epoch": 16.41,
+      "grad_norm": 2.081451892852783,
+      "learning_rate": 2e-05,
+      "loss": 0.03653099,
+      "step": 8205
+    },
+    {
+      "epoch": 16.412,
+      "grad_norm": 1.4831323623657227,
+      "learning_rate": 2e-05,
+      "loss": 0.04202024,
+      "step": 8206
+    },
+    {
+      "epoch": 16.414,
+      "grad_norm": 0.8882843852043152,
+      "learning_rate": 2e-05,
+      "loss": 0.02640081,
+      "step": 8207
+    },
+    {
+      "epoch": 16.416,
+      "grad_norm": 1.2519268989562988,
+      "learning_rate": 2e-05,
+      "loss": 0.03396625,
+      "step": 8208
+    },
+    {
+      "epoch": 16.418,
+      "grad_norm": 1.0374751091003418,
+      "learning_rate": 2e-05,
+      "loss": 0.03408733,
+      "step": 8209
+    },
+    {
+      "epoch": 16.42,
+      "grad_norm": 1.2353500127792358,
+      "learning_rate": 2e-05,
+      "loss": 0.04257344,
+      "step": 8210
+    },
+    {
+      "epoch": 16.422,
+      "grad_norm": 1.310669183731079,
+      "learning_rate": 2e-05,
+      "loss": 0.03693325,
+      "step": 8211
+    },
+    {
+      "epoch": 16.424,
+      "grad_norm": 2.0331413745880127,
+      "learning_rate": 2e-05,
+      "loss": 0.03563491,
+      "step": 8212
+    },
+    {
+      "epoch": 16.426,
+      "grad_norm": 1.2886682748794556,
+      "learning_rate": 2e-05,
+      "loss": 0.03784215,
+      "step": 8213
+    },
+    {
+      "epoch": 16.428,
+      "grad_norm": 1.1843023300170898,
+      "learning_rate": 2e-05,
+      "loss": 0.04167389,
+      "step": 8214
+    },
+    {
+      "epoch": 16.43,
+      "grad_norm": 2.5486996173858643,
+      "learning_rate": 2e-05,
+      "loss": 0.05189614,
+      "step": 8215
+    },
+    {
+      "epoch": 16.432,
+      "grad_norm": 1.908674955368042,
+      "learning_rate": 2e-05,
+      "loss": 0.02384158,
+      "step": 8216
+    },
+    {
+      "epoch": 16.434,
+      "grad_norm": 1.7661586999893188,
+      "learning_rate": 2e-05,
+      "loss": 0.03849943,
+      "step": 8217
+    },
+    {
+      "epoch": 16.436,
+      "grad_norm": 1.0904186964035034,
+      "learning_rate": 2e-05,
+      "loss": 0.04790416,
+      "step": 8218
+    },
+    {
+      "epoch": 16.438,
+      "grad_norm": 1.0966798067092896,
+      "learning_rate": 2e-05,
+      "loss": 0.03781455,
+      "step": 8219
+    },
+    {
+      "epoch": 16.44,
+      "grad_norm": 0.9001188278198242,
+      "learning_rate": 2e-05,
+      "loss": 0.02561498,
+      "step": 8220
+    },
+    {
+      "epoch": 16.442,
+      "grad_norm": 0.9903857111930847,
+      "learning_rate": 2e-05,
+      "loss": 0.0389318,
+      "step": 8221
+    },
+    {
+      "epoch": 16.444,
+      "grad_norm": 1.5551356077194214,
+      "learning_rate": 2e-05,
+      "loss": 0.05141037,
+      "step": 8222
+    },
+    {
+      "epoch": 16.446,
+      "grad_norm": 1.2812175750732422,
+      "learning_rate": 2e-05,
+      "loss": 0.03643136,
+      "step": 8223
+    },
+    {
+      "epoch": 16.448,
+      "grad_norm": 1.0215216875076294,
+      "learning_rate": 2e-05,
+      "loss": 0.02926203,
+      "step": 8224
+    },
+    {
+      "epoch": 16.45,
+      "grad_norm": 1.4955259561538696,
+      "learning_rate": 2e-05,
+      "loss": 0.03899873,
+      "step": 8225
+    },
+    {
+      "epoch": 16.452,
+      "grad_norm": 1.4788180589675903,
+      "learning_rate": 2e-05,
+      "loss": 0.05681079,
+      "step": 8226
+    },
+    {
+      "epoch": 16.454,
+      "grad_norm": 1.0436245203018188,
+      "learning_rate": 2e-05,
+      "loss": 0.03019481,
+      "step": 8227
+    },
+    {
+      "epoch": 16.456,
+      "grad_norm": 0.839566707611084,
+      "learning_rate": 2e-05,
+      "loss": 0.02787675,
+      "step": 8228
+    },
+    {
+      "epoch": 16.458,
+      "grad_norm": 1.0782071352005005,
+      "learning_rate": 2e-05,
+      "loss": 0.0280141,
+      "step": 8229
+    },
+    {
+      "epoch": 16.46,
+      "grad_norm": 1.0634349584579468,
+      "learning_rate": 2e-05,
+      "loss": 0.03917925,
+      "step": 8230
+    },
+    {
+      "epoch": 16.462,
+      "grad_norm": 1.0668877363204956,
+      "learning_rate": 2e-05,
+      "loss": 0.03235231,
+      "step": 8231
+    },
+    {
+      "epoch": 16.464,
+      "grad_norm": 1.0869591236114502,
+      "learning_rate": 2e-05,
+      "loss": 0.03207518,
+      "step": 8232
+    },
+    {
+      "epoch": 16.466,
+      "grad_norm": 1.7971811294555664,
+      "learning_rate": 2e-05,
+      "loss": 0.04344584,
+      "step": 8233
+    },
+    {
+      "epoch": 16.468,
+      "grad_norm": 2.0052719116210938,
+      "learning_rate": 2e-05,
+      "loss": 0.03197655,
+      "step": 8234
+    },
+    {
+      "epoch": 16.47,
+      "grad_norm": 1.3378233909606934,
+      "learning_rate": 2e-05,
+      "loss": 0.03234541,
+      "step": 8235
+    },
+    {
+      "epoch": 16.472,
+      "grad_norm": 2.523751735687256,
+      "learning_rate": 2e-05,
+      "loss": 0.04990536,
+      "step": 8236
+    },
+    {
+      "epoch": 16.474,
+      "grad_norm": 1.0454021692276,
+      "learning_rate": 2e-05,
+      "loss": 0.03331465,
+      "step": 8237
+    },
+    {
+      "epoch": 16.476,
+      "grad_norm": 1.1137912273406982,
+      "learning_rate": 2e-05,
+      "loss": 0.03970309,
+      "step": 8238
+    },
+    {
+      "epoch": 16.478,
+      "grad_norm": 1.585533618927002,
+      "learning_rate": 2e-05,
+      "loss": 0.0488726,
+      "step": 8239
+    },
+    {
+      "epoch": 16.48,
+      "grad_norm": 3.0131523609161377,
+      "learning_rate": 2e-05,
+      "loss": 0.04734076,
+      "step": 8240
+    },
+    {
+      "epoch": 16.482,
+      "grad_norm": 0.9826610684394836,
+      "learning_rate": 2e-05,
+      "loss": 0.0252513,
+      "step": 8241
+    },
+    {
+      "epoch": 16.484,
+      "grad_norm": 1.9169968366622925,
+      "learning_rate": 2e-05,
+      "loss": 0.03856509,
+      "step": 8242
+    },
+    {
+      "epoch": 16.486,
+      "grad_norm": 1.338022232055664,
+      "learning_rate": 2e-05,
+      "loss": 0.03761451,
+      "step": 8243
+    },
+    {
+      "epoch": 16.488,
+      "grad_norm": 1.4171205759048462,
+      "learning_rate": 2e-05,
+      "loss": 0.04478035,
+      "step": 8244
+    },
+    {
+      "epoch": 16.49,
+      "grad_norm": 1.2033617496490479,
+      "learning_rate": 2e-05,
+      "loss": 0.03369937,
+      "step": 8245
+    },
+    {
+      "epoch": 16.492,
+      "grad_norm": 1.676818609237671,
+      "learning_rate": 2e-05,
+      "loss": 0.03760742,
+      "step": 8246
+    },
+    {
+      "epoch": 16.494,
+      "grad_norm": 1.0126457214355469,
+      "learning_rate": 2e-05,
+      "loss": 0.03277501,
+      "step": 8247
+    },
+    {
+      "epoch": 16.496,
+      "grad_norm": 1.122615933418274,
+      "learning_rate": 2e-05,
+      "loss": 0.03475213,
+      "step": 8248
+    },
+    {
+      "epoch": 16.498,
+      "grad_norm": 1.2609951496124268,
+      "learning_rate": 2e-05,
+      "loss": 0.03638831,
+      "step": 8249
+    },
+    {
+      "epoch": 16.5,
+      "grad_norm": 1.3109709024429321,
+      "learning_rate": 2e-05,
+      "loss": 0.04073593,
+      "step": 8250
+    },
+    {
+      "epoch": 16.502,
+      "grad_norm": 1.6363704204559326,
+      "learning_rate": 2e-05,
+      "loss": 0.04546898,
+      "step": 8251
+    },
+    {
+      "epoch": 16.504,
+      "grad_norm": 1.0490771532058716,
+      "learning_rate": 2e-05,
+      "loss": 0.03132774,
+      "step": 8252
+    },
+    {
+      "epoch": 16.506,
+      "grad_norm": 1.9119375944137573,
+      "learning_rate": 2e-05,
+      "loss": 0.04283144,
+      "step": 8253
+    },
+    {
+      "epoch": 16.508,
+      "grad_norm": 1.7922325134277344,
+      "learning_rate": 2e-05,
+      "loss": 0.04706988,
+      "step": 8254
+    },
+    {
+      "epoch": 16.51,
+      "grad_norm": 2.0722665786743164,
+      "learning_rate": 2e-05,
+      "loss": 0.04967377,
+      "step": 8255
+    },
+    {
+      "epoch": 16.512,
+      "grad_norm": 1.6956374645233154,
+      "learning_rate": 2e-05,
+      "loss": 0.03436553,
+      "step": 8256
+    },
+    {
+      "epoch": 16.514,
+      "grad_norm": 1.2904715538024902,
+      "learning_rate": 2e-05,
+      "loss": 0.03789183,
+      "step": 8257
+    },
+    {
+      "epoch": 16.516,
+      "grad_norm": 1.4763702154159546,
+      "learning_rate": 2e-05,
+      "loss": 0.04574104,
+      "step": 8258
+    },
+    {
+      "epoch": 16.518,
+      "grad_norm": 1.6451772451400757,
+      "learning_rate": 2e-05,
+      "loss": 0.05261879,
+      "step": 8259
+    },
+    {
+      "epoch": 16.52,
+      "grad_norm": 1.7381635904312134,
+      "learning_rate": 2e-05,
+      "loss": 0.03526191,
+      "step": 8260
+    },
+    {
+      "epoch": 16.522,
+      "grad_norm": 1.263259768486023,
+      "learning_rate": 2e-05,
+      "loss": 0.03281062,
+      "step": 8261
+    },
+    {
+      "epoch": 16.524,
+      "grad_norm": 1.220733642578125,
+      "learning_rate": 2e-05,
+      "loss": 0.03505442,
+      "step": 8262
+    },
+    {
+      "epoch": 16.526,
+      "grad_norm": 1.2341818809509277,
+      "learning_rate": 2e-05,
+      "loss": 0.04311585,
+      "step": 8263
+    },
+    {
+      "epoch": 16.528,
+      "grad_norm": 0.9565370678901672,
+      "learning_rate": 2e-05,
+      "loss": 0.02928256,
+      "step": 8264
+    },
+    {
+      "epoch": 16.53,
+      "grad_norm": 0.8718287944793701,
+      "learning_rate": 2e-05,
+      "loss": 0.02539262,
+      "step": 8265
+    },
+    {
+      "epoch": 16.532,
+      "grad_norm": 1.1354038715362549,
+      "learning_rate": 2e-05,
+      "loss": 0.03417098,
+      "step": 8266
+    },
+    {
+      "epoch": 16.534,
+      "grad_norm": 1.1788674592971802,
+      "learning_rate": 2e-05,
+      "loss": 0.03965496,
+      "step": 8267
+    },
+    {
+      "epoch": 16.536,
+      "grad_norm": 1.3101824522018433,
+      "learning_rate": 2e-05,
+      "loss": 0.03684475,
+      "step": 8268
+    },
+    {
+      "epoch": 16.538,
+      "grad_norm": 1.3392512798309326,
+      "learning_rate": 2e-05,
+      "loss": 0.04036241,
+      "step": 8269
+    },
+    {
+      "epoch": 16.54,
+      "grad_norm": 2.521176338195801,
+      "learning_rate": 2e-05,
+      "loss": 0.04675987,
+      "step": 8270
+    },
+    {
+      "epoch": 16.542,
+      "grad_norm": 1.11541748046875,
+      "learning_rate": 2e-05,
+      "loss": 0.03730694,
+      "step": 8271
+    },
+    {
+      "epoch": 16.544,
+      "grad_norm": 1.528278112411499,
+      "learning_rate": 2e-05,
+      "loss": 0.06005888,
+      "step": 8272
+    },
+    {
+      "epoch": 16.546,
+      "grad_norm": 1.2424397468566895,
+      "learning_rate": 2e-05,
+      "loss": 0.05591384,
+      "step": 8273
+    },
+    {
+      "epoch": 16.548000000000002,
+      "grad_norm": 1.0814480781555176,
+      "learning_rate": 2e-05,
+      "loss": 0.03427697,
+      "step": 8274
+    },
+    {
+      "epoch": 16.55,
+      "grad_norm": 2.6472551822662354,
+      "learning_rate": 2e-05,
+      "loss": 0.0521223,
+      "step": 8275
+    },
+    {
+      "epoch": 16.552,
+      "grad_norm": 1.3621563911437988,
+      "learning_rate": 2e-05,
+      "loss": 0.04933754,
+      "step": 8276
+    },
+    {
+      "epoch": 16.554,
+      "grad_norm": 2.2583024501800537,
+      "learning_rate": 2e-05,
+      "loss": 0.05284938,
+      "step": 8277
+    },
+    {
+      "epoch": 16.556,
+      "grad_norm": 1.0987781286239624,
+      "learning_rate": 2e-05,
+      "loss": 0.02926936,
+      "step": 8278
+    },
+    {
+      "epoch": 16.558,
+      "grad_norm": 1.283735752105713,
+      "learning_rate": 2e-05,
+      "loss": 0.03353404,
+      "step": 8279
+    },
+    {
+      "epoch": 16.56,
+      "grad_norm": 1.0234546661376953,
+      "learning_rate": 2e-05,
+      "loss": 0.03110966,
+      "step": 8280
+    },
+    {
+      "epoch": 16.562,
+      "grad_norm": 1.325693130493164,
+      "learning_rate": 2e-05,
+      "loss": 0.04735526,
+      "step": 8281
+    },
+    {
+      "epoch": 16.564,
+      "grad_norm": 1.5526974201202393,
+      "learning_rate": 2e-05,
+      "loss": 0.04662727,
+      "step": 8282
+    },
+    {
+      "epoch": 16.566,
+      "grad_norm": 2.4452273845672607,
+      "learning_rate": 2e-05,
+      "loss": 0.03830102,
+      "step": 8283
+    },
+    {
+      "epoch": 16.568,
+      "grad_norm": 1.2660468816757202,
+      "learning_rate": 2e-05,
+      "loss": 0.04160655,
+      "step": 8284
+    },
+    {
+      "epoch": 16.57,
+      "grad_norm": 1.1161909103393555,
+      "learning_rate": 2e-05,
+      "loss": 0.04231364,
+      "step": 8285
+    },
+    {
+      "epoch": 16.572,
+      "grad_norm": 1.012315273284912,
+      "learning_rate": 2e-05,
+      "loss": 0.03184524,
+      "step": 8286
+    },
+    {
+      "epoch": 16.574,
+      "grad_norm": 1.1147305965423584,
+      "learning_rate": 2e-05,
+      "loss": 0.03896227,
+      "step": 8287
+    },
+    {
+      "epoch": 16.576,
+      "grad_norm": 1.0445970296859741,
+      "learning_rate": 2e-05,
+      "loss": 0.03200065,
+      "step": 8288
+    },
+    {
+      "epoch": 16.578,
+      "grad_norm": 1.7669461965560913,
+      "learning_rate": 2e-05,
+      "loss": 0.04407949,
+      "step": 8289
+    },
+    {
+      "epoch": 16.58,
+      "grad_norm": 1.550229549407959,
+      "learning_rate": 2e-05,
+      "loss": 0.02869645,
+      "step": 8290
+    },
+    {
+      "epoch": 16.582,
+      "grad_norm": 1.0623831748962402,
+      "learning_rate": 2e-05,
+      "loss": 0.0305838,
+      "step": 8291
+    },
+    {
+      "epoch": 16.584,
+      "grad_norm": 1.5914502143859863,
+      "learning_rate": 2e-05,
+      "loss": 0.04625592,
+      "step": 8292
+    },
+    {
+      "epoch": 16.586,
+      "grad_norm": 1.5053743124008179,
+      "learning_rate": 2e-05,
+      "loss": 0.03940097,
+      "step": 8293
+    },
+    {
+      "epoch": 16.588,
+      "grad_norm": 1.9061319828033447,
+      "learning_rate": 2e-05,
+      "loss": 0.05159453,
+      "step": 8294
+    },
+    {
+      "epoch": 16.59,
+      "grad_norm": 1.4841290712356567,
+      "learning_rate": 2e-05,
+      "loss": 0.05897345,
+      "step": 8295
+    },
+    {
+      "epoch": 16.592,
+      "grad_norm": 2.136908769607544,
+      "learning_rate": 2e-05,
+      "loss": 0.05323712,
+      "step": 8296
+    },
+    {
+      "epoch": 16.594,
+      "grad_norm": 1.3379402160644531,
+      "learning_rate": 2e-05,
+      "loss": 0.03460371,
+      "step": 8297
+    },
+    {
+      "epoch": 16.596,
+      "grad_norm": 1.020004391670227,
+      "learning_rate": 2e-05,
+      "loss": 0.03321714,
+      "step": 8298
+    },
+    {
+      "epoch": 16.598,
+      "grad_norm": 2.1319894790649414,
+      "learning_rate": 2e-05,
+      "loss": 0.04367604,
+      "step": 8299
+    },
+    {
+      "epoch": 16.6,
+      "grad_norm": 1.1898785829544067,
+      "learning_rate": 2e-05,
+      "loss": 0.03668055,
+      "step": 8300
+    },
+    {
+      "epoch": 16.602,
+      "grad_norm": 1.096915364265442,
+      "learning_rate": 2e-05,
+      "loss": 0.03318947,
+      "step": 8301
+    },
+    {
+      "epoch": 16.604,
+      "grad_norm": 1.0413960218429565,
+      "learning_rate": 2e-05,
+      "loss": 0.03442251,
+      "step": 8302
+    },
+    {
+      "epoch": 16.606,
+      "grad_norm": 1.2012441158294678,
+      "learning_rate": 2e-05,
+      "loss": 0.02705516,
+      "step": 8303
+    },
+    {
+      "epoch": 16.608,
+      "grad_norm": 1.2894916534423828,
+      "learning_rate": 2e-05,
+      "loss": 0.03271394,
+      "step": 8304
+    },
+    {
+      "epoch": 16.61,
+      "grad_norm": 1.2636003494262695,
+      "learning_rate": 2e-05,
+      "loss": 0.04137167,
+      "step": 8305
+    },
+    {
+      "epoch": 16.612,
+      "grad_norm": 1.2958347797393799,
+      "learning_rate": 2e-05,
+      "loss": 0.04770758,
+      "step": 8306
+    },
+    {
+      "epoch": 16.614,
+      "grad_norm": 2.002460479736328,
+      "learning_rate": 2e-05,
+      "loss": 0.0475505,
+      "step": 8307
+    },
+    {
+      "epoch": 16.616,
+      "grad_norm": 0.9963358640670776,
+      "learning_rate": 2e-05,
+      "loss": 0.03287155,
+      "step": 8308
+    },
+    {
+      "epoch": 16.618,
+      "grad_norm": 1.1301119327545166,
+      "learning_rate": 2e-05,
+      "loss": 0.04497526,
+      "step": 8309
+    },
+    {
+      "epoch": 16.62,
+      "grad_norm": 1.0189284086227417,
+      "learning_rate": 2e-05,
+      "loss": 0.0293178,
+      "step": 8310
+    },
+    {
+      "epoch": 16.622,
+      "grad_norm": 0.8906993269920349,
+      "learning_rate": 2e-05,
+      "loss": 0.01907645,
+      "step": 8311
+    },
+    {
+      "epoch": 16.624,
+      "grad_norm": 1.2180598974227905,
+      "learning_rate": 2e-05,
+      "loss": 0.03251302,
+      "step": 8312
+    },
+    {
+      "epoch": 16.626,
+      "grad_norm": 1.7638144493103027,
+      "learning_rate": 2e-05,
+      "loss": 0.03417107,
+      "step": 8313
+    },
+    {
+      "epoch": 16.628,
+      "grad_norm": 1.3163055181503296,
+      "learning_rate": 2e-05,
+      "loss": 0.03712931,
+      "step": 8314
+    },
+    {
+      "epoch": 16.63,
+      "grad_norm": 0.903484046459198,
+      "learning_rate": 2e-05,
+      "loss": 0.02302654,
+      "step": 8315
+    },
+    {
+      "epoch": 16.632,
+      "grad_norm": 2.890394926071167,
+      "learning_rate": 2e-05,
+      "loss": 0.05785254,
+      "step": 8316
+    },
+    {
+      "epoch": 16.634,
+      "grad_norm": 1.6316351890563965,
+      "learning_rate": 2e-05,
+      "loss": 0.05490664,
+      "step": 8317
+    },
+    {
+      "epoch": 16.636,
+      "grad_norm": 1.3253639936447144,
+      "learning_rate": 2e-05,
+      "loss": 0.03378769,
+      "step": 8318
+    },
+    {
+      "epoch": 16.638,
+      "grad_norm": 0.8841352462768555,
+      "learning_rate": 2e-05,
+      "loss": 0.02580001,
+      "step": 8319
+    },
+    {
+      "epoch": 16.64,
+      "grad_norm": 1.5405997037887573,
+      "learning_rate": 2e-05,
+      "loss": 0.0418197,
+      "step": 8320
+    },
+    {
+      "epoch": 16.642,
+      "grad_norm": 1.6014611721038818,
+      "learning_rate": 2e-05,
+      "loss": 0.05297592,
+      "step": 8321
+    },
+    {
+      "epoch": 16.644,
+      "grad_norm": 0.9699270129203796,
+      "learning_rate": 2e-05,
+      "loss": 0.03074204,
+      "step": 8322
+    },
+    {
+      "epoch": 16.646,
+      "grad_norm": 1.0691285133361816,
+      "learning_rate": 2e-05,
+      "loss": 0.02828666,
+      "step": 8323
+    },
+    {
+      "epoch": 16.648,
+      "grad_norm": 1.4978876113891602,
+      "learning_rate": 2e-05,
+      "loss": 0.04264086,
+      "step": 8324
+    },
+    {
+      "epoch": 16.65,
+      "grad_norm": 1.067944884300232,
+      "learning_rate": 2e-05,
+      "loss": 0.04064353,
+      "step": 8325
+    },
+    {
+      "epoch": 16.652,
+      "grad_norm": 1.609089732170105,
+      "learning_rate": 2e-05,
+      "loss": 0.03844542,
+      "step": 8326
+    },
+    {
+      "epoch": 16.654,
+      "grad_norm": 1.3730612993240356,
+      "learning_rate": 2e-05,
+      "loss": 0.03860483,
+      "step": 8327
+    },
+    {
+      "epoch": 16.656,
+      "grad_norm": 0.989010214805603,
+      "learning_rate": 2e-05,
+      "loss": 0.02395894,
+      "step": 8328
+    },
+    {
+      "epoch": 16.658,
+      "grad_norm": 0.89674311876297,
+      "learning_rate": 2e-05,
+      "loss": 0.02376273,
+      "step": 8329
+    },
+    {
+      "epoch": 16.66,
+      "grad_norm": 1.1137092113494873,
+      "learning_rate": 2e-05,
+      "loss": 0.01789567,
+      "step": 8330
+    },
+    {
+      "epoch": 16.662,
+      "grad_norm": 1.0320215225219727,
+      "learning_rate": 2e-05,
+      "loss": 0.0323988,
+      "step": 8331
+    },
+    {
+      "epoch": 16.664,
+      "grad_norm": 2.489551067352295,
+      "learning_rate": 2e-05,
+      "loss": 0.05527273,
+      "step": 8332
+    },
+    {
+      "epoch": 16.666,
+      "grad_norm": 1.2354681491851807,
+      "learning_rate": 2e-05,
+      "loss": 0.03027255,
+      "step": 8333
+    },
+    {
+      "epoch": 16.668,
+      "grad_norm": 1.5036636590957642,
+      "learning_rate": 2e-05,
+      "loss": 0.03653241,
+      "step": 8334
+    },
+    {
+      "epoch": 16.67,
+      "grad_norm": 1.5149270296096802,
+      "learning_rate": 2e-05,
+      "loss": 0.04814313,
+      "step": 8335
+    },
+    {
+      "epoch": 16.672,
+      "grad_norm": 1.3045963048934937,
+      "learning_rate": 2e-05,
+      "loss": 0.03031799,
+      "step": 8336
+    },
+    {
+      "epoch": 16.674,
+      "grad_norm": 1.2806516885757446,
+      "learning_rate": 2e-05,
+      "loss": 0.04692192,
+      "step": 8337
+    },
+    {
+      "epoch": 16.676,
+      "grad_norm": 1.1512733697891235,
+      "learning_rate": 2e-05,
+      "loss": 0.04004252,
+      "step": 8338
+    },
+    {
+      "epoch": 16.678,
+      "grad_norm": 3.1383249759674072,
+      "learning_rate": 2e-05,
+      "loss": 0.0324553,
+      "step": 8339
+    },
+    {
+      "epoch": 16.68,
+      "grad_norm": 1.2203539609909058,
+      "learning_rate": 2e-05,
+      "loss": 0.03611985,
+      "step": 8340
+    },
+    {
+      "epoch": 16.682,
+      "grad_norm": 1.2353651523590088,
+      "learning_rate": 2e-05,
+      "loss": 0.03453249,
+      "step": 8341
+    },
+    {
+      "epoch": 16.684,
+      "grad_norm": 1.2120214700698853,
+      "learning_rate": 2e-05,
+      "loss": 0.04656072,
+      "step": 8342
+    },
+    {
+      "epoch": 16.686,
+      "grad_norm": 1.1942620277404785,
+      "learning_rate": 2e-05,
+      "loss": 0.04031057,
+      "step": 8343
+    },
+    {
+      "epoch": 16.688,
+      "grad_norm": 1.174004077911377,
+      "learning_rate": 2e-05,
+      "loss": 0.03165752,
+      "step": 8344
+    },
+    {
+      "epoch": 16.69,
+      "grad_norm": 2.1396734714508057,
+      "learning_rate": 2e-05,
+      "loss": 0.04839662,
+      "step": 8345
+    },
+    {
+      "epoch": 16.692,
+      "grad_norm": 1.1642301082611084,
+      "learning_rate": 2e-05,
+      "loss": 0.0288971,
+      "step": 8346
+    },
+    {
+      "epoch": 16.694,
+      "grad_norm": 1.6050286293029785,
+      "learning_rate": 2e-05,
+      "loss": 0.03548468,
+      "step": 8347
+    },
+    {
+      "epoch": 16.696,
+      "grad_norm": 1.416585922241211,
+      "learning_rate": 2e-05,
+      "loss": 0.03986949,
+      "step": 8348
+    },
+    {
+      "epoch": 16.698,
+      "grad_norm": 1.3496794700622559,
+      "learning_rate": 2e-05,
+      "loss": 0.04378842,
+      "step": 8349
+    },
+    {
+      "epoch": 16.7,
+      "grad_norm": 3.526726722717285,
+      "learning_rate": 2e-05,
+      "loss": 0.04224211,
+      "step": 8350
+    },
+    {
+      "epoch": 16.701999999999998,
+      "grad_norm": 2.085320234298706,
+      "learning_rate": 2e-05,
+      "loss": 0.02257501,
+      "step": 8351
+    },
+    {
+      "epoch": 16.704,
+      "grad_norm": 1.418104648590088,
+      "learning_rate": 2e-05,
+      "loss": 0.03896306,
+      "step": 8352
+    },
+    {
+      "epoch": 16.706,
+      "grad_norm": 1.8346154689788818,
+      "learning_rate": 2e-05,
+      "loss": 0.04356946,
+      "step": 8353
+    },
+    {
+      "epoch": 16.708,
+      "grad_norm": 1.2784098386764526,
+      "learning_rate": 2e-05,
+      "loss": 0.04127508,
+      "step": 8354
+    },
+    {
+      "epoch": 16.71,
+      "grad_norm": 1.3677000999450684,
+      "learning_rate": 2e-05,
+      "loss": 0.03775769,
+      "step": 8355
+    },
+    {
+      "epoch": 16.712,
+      "grad_norm": 1.7542495727539062,
+      "learning_rate": 2e-05,
+      "loss": 0.05272632,
+      "step": 8356
+    },
+    {
+      "epoch": 16.714,
+      "grad_norm": 1.9329322576522827,
+      "learning_rate": 2e-05,
+      "loss": 0.041315,
+      "step": 8357
+    },
+    {
+      "epoch": 16.716,
+      "grad_norm": 1.4365530014038086,
+      "learning_rate": 2e-05,
+      "loss": 0.03562573,
+      "step": 8358
+    },
+    {
+      "epoch": 16.718,
+      "grad_norm": 1.7461820840835571,
+      "learning_rate": 2e-05,
+      "loss": 0.04044671,
+      "step": 8359
+    },
+    {
+      "epoch": 16.72,
+      "grad_norm": 5.9637556076049805,
+      "learning_rate": 2e-05,
+      "loss": 0.04565287,
+      "step": 8360
+    },
+    {
+      "epoch": 16.722,
+      "grad_norm": 1.428763747215271,
+      "learning_rate": 2e-05,
+      "loss": 0.03354582,
+      "step": 8361
+    },
+    {
+      "epoch": 16.724,
+      "grad_norm": 1.4038552045822144,
+      "learning_rate": 2e-05,
+      "loss": 0.0424789,
+      "step": 8362
+    },
+    {
+      "epoch": 16.726,
+      "grad_norm": 1.2637503147125244,
+      "learning_rate": 2e-05,
+      "loss": 0.03532393,
+      "step": 8363
+    },
+    {
+      "epoch": 16.728,
+      "grad_norm": 1.817488670349121,
+      "learning_rate": 2e-05,
+      "loss": 0.04140956,
+      "step": 8364
+    },
+    {
+      "epoch": 16.73,
+      "grad_norm": 1.4110076427459717,
+      "learning_rate": 2e-05,
+      "loss": 0.03238542,
+      "step": 8365
+    },
+    {
+      "epoch": 16.732,
+      "grad_norm": 1.412273645401001,
+      "learning_rate": 2e-05,
+      "loss": 0.0464592,
+      "step": 8366
+    },
+    {
+      "epoch": 16.734,
+      "grad_norm": 1.112044334411621,
+      "learning_rate": 2e-05,
+      "loss": 0.0404929,
+      "step": 8367
+    },
+    {
+      "epoch": 16.736,
+      "grad_norm": 1.867051362991333,
+      "learning_rate": 2e-05,
+      "loss": 0.06586152,
+      "step": 8368
+    },
+    {
+      "epoch": 16.738,
+      "grad_norm": 6.789832592010498,
+      "learning_rate": 2e-05,
+      "loss": 0.03430699,
+      "step": 8369
+    },
+    {
+      "epoch": 16.74,
+      "grad_norm": 1.4201772212982178,
+      "learning_rate": 2e-05,
+      "loss": 0.03015824,
+      "step": 8370
+    },
+    {
+      "epoch": 16.742,
+      "grad_norm": 1.1724557876586914,
+      "learning_rate": 2e-05,
+      "loss": 0.03560322,
+      "step": 8371
+    },
+    {
+      "epoch": 16.744,
+      "grad_norm": 1.1797523498535156,
+      "learning_rate": 2e-05,
+      "loss": 0.02636663,
+      "step": 8372
+    },
+    {
+      "epoch": 16.746,
+      "grad_norm": 1.3166332244873047,
+      "learning_rate": 2e-05,
+      "loss": 0.03184573,
+      "step": 8373
+    },
+    {
+      "epoch": 16.748,
+      "grad_norm": 1.034335970878601,
+      "learning_rate": 2e-05,
+      "loss": 0.02896601,
+      "step": 8374
+    },
+    {
+      "epoch": 16.75,
+      "grad_norm": 1.0973848104476929,
+      "learning_rate": 2e-05,
+      "loss": 0.03897912,
+      "step": 8375
+    },
+    {
+      "epoch": 16.752,
+      "grad_norm": 1.1487135887145996,
+      "learning_rate": 2e-05,
+      "loss": 0.03742709,
+      "step": 8376
+    },
+    {
+      "epoch": 16.754,
+      "grad_norm": 1.0410828590393066,
+      "learning_rate": 2e-05,
+      "loss": 0.03710461,
+      "step": 8377
+    },
+    {
+      "epoch": 16.756,
+      "grad_norm": 1.2444871664047241,
+      "learning_rate": 2e-05,
+      "loss": 0.03904767,
+      "step": 8378
+    },
+    {
+      "epoch": 16.758,
+      "grad_norm": 1.5975366830825806,
+      "learning_rate": 2e-05,
+      "loss": 0.04068123,
+      "step": 8379
+    },
+    {
+      "epoch": 16.76,
+      "grad_norm": 1.1661255359649658,
+      "learning_rate": 2e-05,
+      "loss": 0.02516171,
+      "step": 8380
+    },
+    {
+      "epoch": 16.762,
+      "grad_norm": 2.5861120223999023,
+      "learning_rate": 2e-05,
+      "loss": 0.04131169,
+      "step": 8381
+    },
+    {
+      "epoch": 16.764,
+      "grad_norm": 2.497673749923706,
+      "learning_rate": 2e-05,
+      "loss": 0.04161269,
+      "step": 8382
+    },
+    {
+      "epoch": 16.766,
+      "grad_norm": 0.9345721006393433,
+      "learning_rate": 2e-05,
+      "loss": 0.02630493,
+      "step": 8383
+    },
+    {
+      "epoch": 16.768,
+      "grad_norm": 1.4174412488937378,
+      "learning_rate": 2e-05,
+      "loss": 0.03265426,
+      "step": 8384
+    },
+    {
+      "epoch": 16.77,
+      "grad_norm": 1.2070684432983398,
+      "learning_rate": 2e-05,
+      "loss": 0.03450275,
+      "step": 8385
+    },
+    {
+      "epoch": 16.772,
+      "grad_norm": 1.3117280006408691,
+      "learning_rate": 2e-05,
+      "loss": 0.04178444,
+      "step": 8386
+    },
+    {
+      "epoch": 16.774,
+      "grad_norm": 1.5097566843032837,
+      "learning_rate": 2e-05,
+      "loss": 0.04680046,
+      "step": 8387
+    },
+    {
+      "epoch": 16.776,
+      "grad_norm": 1.4381500482559204,
+      "learning_rate": 2e-05,
+      "loss": 0.04752283,
+      "step": 8388
+    },
+    {
+      "epoch": 16.778,
+      "grad_norm": 1.1138615608215332,
+      "learning_rate": 2e-05,
+      "loss": 0.03561661,
+      "step": 8389
+    },
+    {
+      "epoch": 16.78,
+      "grad_norm": 1.2291978597640991,
+      "learning_rate": 2e-05,
+      "loss": 0.03155664,
+      "step": 8390
+    },
+    {
+      "epoch": 16.782,
+      "grad_norm": 1.3546652793884277,
+      "learning_rate": 2e-05,
+      "loss": 0.03348469,
+      "step": 8391
+    },
+    {
+      "epoch": 16.784,
+      "grad_norm": 1.1350245475769043,
+      "learning_rate": 2e-05,
+      "loss": 0.04207388,
+      "step": 8392
+    },
+    {
+      "epoch": 16.786,
+      "grad_norm": 1.6995071172714233,
+      "learning_rate": 2e-05,
+      "loss": 0.05210853,
+      "step": 8393
+    },
+    {
+      "epoch": 16.788,
+      "grad_norm": 2.271347761154175,
+      "learning_rate": 2e-05,
+      "loss": 0.05933637,
+      "step": 8394
+    },
+    {
+      "epoch": 16.79,
+      "grad_norm": 0.8941410779953003,
+      "learning_rate": 2e-05,
+      "loss": 0.02386924,
+      "step": 8395
+    },
+    {
+      "epoch": 16.792,
+      "grad_norm": 1.6003661155700684,
+      "learning_rate": 2e-05,
+      "loss": 0.05291647,
+      "step": 8396
+    },
+    {
+      "epoch": 16.794,
+      "grad_norm": 1.3317174911499023,
+      "learning_rate": 2e-05,
+      "loss": 0.03087507,
+      "step": 8397
+    },
+    {
+      "epoch": 16.796,
+      "grad_norm": 0.9487977623939514,
+      "learning_rate": 2e-05,
+      "loss": 0.02690414,
+      "step": 8398
+    },
+    {
+      "epoch": 16.798000000000002,
+      "grad_norm": 1.009123682975769,
+      "learning_rate": 2e-05,
+      "loss": 0.02644867,
+      "step": 8399
+    },
+    {
+      "epoch": 16.8,
+      "grad_norm": 1.2167582511901855,
+      "learning_rate": 2e-05,
+      "loss": 0.03899329,
+      "step": 8400
+    },
+    {
+      "epoch": 16.802,
+      "grad_norm": 1.482376217842102,
+      "learning_rate": 2e-05,
+      "loss": 0.02873115,
+      "step": 8401
+    },
+    {
+      "epoch": 16.804,
+      "grad_norm": 1.1082066297531128,
+      "learning_rate": 2e-05,
+      "loss": 0.03039532,
+      "step": 8402
+    },
+    {
+      "epoch": 16.806,
+      "grad_norm": 1.2599750757217407,
+      "learning_rate": 2e-05,
+      "loss": 0.04293299,
+      "step": 8403
+    },
+    {
+      "epoch": 16.808,
+      "grad_norm": 0.9543814063072205,
+      "learning_rate": 2e-05,
+      "loss": 0.03469695,
+      "step": 8404
+    },
+    {
+      "epoch": 16.81,
+      "grad_norm": 1.0824414491653442,
+      "learning_rate": 2e-05,
+      "loss": 0.03982303,
+      "step": 8405
+    },
+    {
+      "epoch": 16.812,
+      "grad_norm": 1.099599838256836,
+      "learning_rate": 2e-05,
+      "loss": 0.02591001,
+      "step": 8406
+    },
+    {
+      "epoch": 16.814,
+      "grad_norm": 1.7804385423660278,
+      "learning_rate": 2e-05,
+      "loss": 0.0323559,
+      "step": 8407
+    },
+    {
+      "epoch": 16.816,
+      "grad_norm": 1.8087066411972046,
+      "learning_rate": 2e-05,
+      "loss": 0.04229058,
+      "step": 8408
+    },
+    {
+      "epoch": 16.818,
+      "grad_norm": 1.2992652654647827,
+      "learning_rate": 2e-05,
+      "loss": 0.0409187,
+      "step": 8409
+    },
+    {
+      "epoch": 16.82,
+      "grad_norm": 1.1120244264602661,
+      "learning_rate": 2e-05,
+      "loss": 0.03508178,
+      "step": 8410
+    },
+    {
+      "epoch": 16.822,
+      "grad_norm": 0.9184687733650208,
+      "learning_rate": 2e-05,
+      "loss": 0.02861979,
+      "step": 8411
+    },
+    {
+      "epoch": 16.824,
+      "grad_norm": 1.2980962991714478,
+      "learning_rate": 2e-05,
+      "loss": 0.02535828,
+      "step": 8412
+    },
+    {
+      "epoch": 16.826,
+      "grad_norm": 1.9372985363006592,
+      "learning_rate": 2e-05,
+      "loss": 0.03568355,
+      "step": 8413
+    },
+    {
+      "epoch": 16.828,
+      "grad_norm": 1.514122486114502,
+      "learning_rate": 2e-05,
+      "loss": 0.03849922,
+      "step": 8414
+    },
+    {
+      "epoch": 16.83,
+      "grad_norm": 1.285515308380127,
+      "learning_rate": 2e-05,
+      "loss": 0.03315242,
+      "step": 8415
+    },
+    {
+      "epoch": 16.832,
+      "grad_norm": 0.9878953099250793,
+      "learning_rate": 2e-05,
+      "loss": 0.02930378,
+      "step": 8416
+    },
+    {
+      "epoch": 16.834,
+      "grad_norm": 1.765859842300415,
+      "learning_rate": 2e-05,
+      "loss": 0.05877309,
+      "step": 8417
+    },
+    {
+      "epoch": 16.836,
+      "grad_norm": 1.3087562322616577,
+      "learning_rate": 2e-05,
+      "loss": 0.04466981,
+      "step": 8418
+    },
+    {
+      "epoch": 16.838,
+      "grad_norm": 3.0149006843566895,
+      "learning_rate": 2e-05,
+      "loss": 0.05240371,
+      "step": 8419
+    },
+    {
+      "epoch": 16.84,
+      "grad_norm": 1.1996604204177856,
+      "learning_rate": 2e-05,
+      "loss": 0.04400566,
+      "step": 8420
+    },
+    {
+      "epoch": 16.842,
+      "grad_norm": 1.4025583267211914,
+      "learning_rate": 2e-05,
+      "loss": 0.04432599,
+      "step": 8421
+    },
+    {
+      "epoch": 16.844,
+      "grad_norm": 2.9091427326202393,
+      "learning_rate": 2e-05,
+      "loss": 0.04849138,
+      "step": 8422
+    },
+    {
+      "epoch": 16.846,
+      "grad_norm": 2.77856707572937,
+      "learning_rate": 2e-05,
+      "loss": 0.05462526,
+      "step": 8423
+    },
+    {
+      "epoch": 16.848,
+      "grad_norm": 1.108500361442566,
+      "learning_rate": 2e-05,
+      "loss": 0.03238573,
+      "step": 8424
+    },
+    {
+      "epoch": 16.85,
+      "grad_norm": 1.0004764795303345,
+      "learning_rate": 2e-05,
+      "loss": 0.02566298,
+      "step": 8425
+    },
+    {
+      "epoch": 16.852,
+      "grad_norm": 6.166355133056641,
+      "learning_rate": 2e-05,
+      "loss": 0.05295902,
+      "step": 8426
+    },
+    {
+      "epoch": 16.854,
+      "grad_norm": 1.1335303783416748,
+      "learning_rate": 2e-05,
+      "loss": 0.0286171,
+      "step": 8427
+    },
+    {
+      "epoch": 16.856,
+      "grad_norm": 1.1951411962509155,
+      "learning_rate": 2e-05,
+      "loss": 0.03944296,
+      "step": 8428
+    },
+    {
+      "epoch": 16.858,
+      "grad_norm": 1.430793046951294,
+      "learning_rate": 2e-05,
+      "loss": 0.03612977,
+      "step": 8429
+    },
+    {
+      "epoch": 16.86,
+      "grad_norm": 1.3680472373962402,
+      "learning_rate": 2e-05,
+      "loss": 0.03450888,
+      "step": 8430
+    },
+    {
+      "epoch": 16.862,
+      "grad_norm": 1.0445443391799927,
+      "learning_rate": 2e-05,
+      "loss": 0.03357092,
+      "step": 8431
+    },
+    {
+      "epoch": 16.864,
+      "grad_norm": 1.1048623323440552,
+      "learning_rate": 2e-05,
+      "loss": 0.02782473,
+      "step": 8432
+    },
+    {
+      "epoch": 16.866,
+      "grad_norm": 1.264748215675354,
+      "learning_rate": 2e-05,
+      "loss": 0.05136178,
+      "step": 8433
+    },
+    {
+      "epoch": 16.868,
+      "grad_norm": 3.1316561698913574,
+      "learning_rate": 2e-05,
+      "loss": 0.0381931,
+      "step": 8434
+    },
+    {
+      "epoch": 16.87,
+      "grad_norm": 0.8829813599586487,
+      "learning_rate": 2e-05,
+      "loss": 0.02602118,
+      "step": 8435
+    },
+    {
+      "epoch": 16.872,
+      "grad_norm": 1.240877389907837,
+      "learning_rate": 2e-05,
+      "loss": 0.0411177,
+      "step": 8436
+    },
+    {
+      "epoch": 16.874,
+      "grad_norm": 0.9366046786308289,
+      "learning_rate": 2e-05,
+      "loss": 0.02752039,
+      "step": 8437
+    },
+    {
+      "epoch": 16.876,
+      "grad_norm": 1.1578431129455566,
+      "learning_rate": 2e-05,
+      "loss": 0.03548286,
+      "step": 8438
+    },
+    {
+      "epoch": 16.878,
+      "grad_norm": 1.1822301149368286,
+      "learning_rate": 2e-05,
+      "loss": 0.03064628,
+      "step": 8439
+    },
+    {
+      "epoch": 16.88,
+      "grad_norm": 2.111692428588867,
+      "learning_rate": 2e-05,
+      "loss": 0.05242971,
+      "step": 8440
+    },
+    {
+      "epoch": 16.882,
+      "grad_norm": 1.7009152173995972,
+      "learning_rate": 2e-05,
+      "loss": 0.03527777,
+      "step": 8441
+    },
+    {
+      "epoch": 16.884,
+      "grad_norm": 1.7795860767364502,
+      "learning_rate": 2e-05,
+      "loss": 0.04584951,
+      "step": 8442
+    },
+    {
+      "epoch": 16.886,
+      "grad_norm": 1.7482233047485352,
+      "learning_rate": 2e-05,
+      "loss": 0.04331917,
+      "step": 8443
+    },
+    {
+      "epoch": 16.888,
+      "grad_norm": 1.1555484533309937,
+      "learning_rate": 2e-05,
+      "loss": 0.03320149,
+      "step": 8444
+    },
+    {
+      "epoch": 16.89,
+      "grad_norm": 1.5613912343978882,
+      "learning_rate": 2e-05,
+      "loss": 0.03231899,
+      "step": 8445
+    },
+    {
+      "epoch": 16.892,
+      "grad_norm": 1.4290865659713745,
+      "learning_rate": 2e-05,
+      "loss": 0.03894955,
+      "step": 8446
+    },
+    {
+      "epoch": 16.894,
+      "grad_norm": 1.1222296953201294,
+      "learning_rate": 2e-05,
+      "loss": 0.03836842,
+      "step": 8447
+    },
+    {
+      "epoch": 16.896,
+      "grad_norm": 2.30586838722229,
+      "learning_rate": 2e-05,
+      "loss": 0.04976833,
+      "step": 8448
+    },
+    {
+      "epoch": 16.898,
+      "grad_norm": 1.0983035564422607,
+      "learning_rate": 2e-05,
+      "loss": 0.03981961,
+      "step": 8449
+    },
+    {
+      "epoch": 16.9,
+      "grad_norm": 1.6317425966262817,
+      "learning_rate": 2e-05,
+      "loss": 0.05492754,
+      "step": 8450
+    },
+    {
+      "epoch": 16.902,
+      "grad_norm": 1.9192370176315308,
+      "learning_rate": 2e-05,
+      "loss": 0.03952098,
+      "step": 8451
+    },
+    {
+      "epoch": 16.904,
+      "grad_norm": 1.1875005960464478,
+      "learning_rate": 2e-05,
+      "loss": 0.03277217,
+      "step": 8452
+    },
+    {
+      "epoch": 16.906,
+      "grad_norm": 1.4212887287139893,
+      "learning_rate": 2e-05,
+      "loss": 0.02984127,
+      "step": 8453
+    },
+    {
+      "epoch": 16.908,
+      "grad_norm": 1.5091922283172607,
+      "learning_rate": 2e-05,
+      "loss": 0.03316611,
+      "step": 8454
+    },
+    {
+      "epoch": 16.91,
+      "grad_norm": 2.024031400680542,
+      "learning_rate": 2e-05,
+      "loss": 0.04787176,
+      "step": 8455
+    },
+    {
+      "epoch": 16.912,
+      "grad_norm": 2.3478596210479736,
+      "learning_rate": 2e-05,
+      "loss": 0.0493004,
+      "step": 8456
+    },
+    {
+      "epoch": 16.914,
+      "grad_norm": 1.2963180541992188,
+      "learning_rate": 2e-05,
+      "loss": 0.02935142,
+      "step": 8457
+    },
+    {
+      "epoch": 16.916,
+      "grad_norm": 1.0010120868682861,
+      "learning_rate": 2e-05,
+      "loss": 0.02706702,
+      "step": 8458
+    },
+    {
+      "epoch": 16.918,
+      "grad_norm": 1.579359531402588,
+      "learning_rate": 2e-05,
+      "loss": 0.03226943,
+      "step": 8459
+    },
+    {
+      "epoch": 16.92,
+      "grad_norm": 1.0996006727218628,
+      "learning_rate": 2e-05,
+      "loss": 0.03097832,
+      "step": 8460
+    },
+    {
+      "epoch": 16.922,
+      "grad_norm": 1.2066538333892822,
+      "learning_rate": 2e-05,
+      "loss": 0.03681121,
+      "step": 8461
+    },
+    {
+      "epoch": 16.924,
+      "grad_norm": 1.255723476409912,
+      "learning_rate": 2e-05,
+      "loss": 0.03210659,
+      "step": 8462
+    },
+    {
+      "epoch": 16.926,
+      "grad_norm": 1.2862356901168823,
+      "learning_rate": 2e-05,
+      "loss": 0.03344401,
+      "step": 8463
+    },
+    {
+      "epoch": 16.928,
+      "grad_norm": 1.046955943107605,
+      "learning_rate": 2e-05,
+      "loss": 0.02581613,
+      "step": 8464
+    },
+    {
+      "epoch": 16.93,
+      "grad_norm": 1.5476090908050537,
+      "learning_rate": 2e-05,
+      "loss": 0.04508192,
+      "step": 8465
+    },
+    {
+      "epoch": 16.932,
+      "grad_norm": 1.1604325771331787,
+      "learning_rate": 2e-05,
+      "loss": 0.03390781,
+      "step": 8466
+    },
+    {
+      "epoch": 16.934,
+      "grad_norm": 1.613224983215332,
+      "learning_rate": 2e-05,
+      "loss": 0.05440042,
+      "step": 8467
+    },
+    {
+      "epoch": 16.936,
+      "grad_norm": 1.0437114238739014,
+      "learning_rate": 2e-05,
+      "loss": 0.03564305,
+      "step": 8468
+    },
+    {
+      "epoch": 16.938,
+      "grad_norm": 1.7123421430587769,
+      "learning_rate": 2e-05,
+      "loss": 0.03162618,
+      "step": 8469
+    },
+    {
+      "epoch": 16.94,
+      "grad_norm": 1.2123863697052002,
+      "learning_rate": 2e-05,
+      "loss": 0.03680079,
+      "step": 8470
+    },
+    {
+      "epoch": 16.942,
+      "grad_norm": 1.7789182662963867,
+      "learning_rate": 2e-05,
+      "loss": 0.05218393,
+      "step": 8471
+    },
+    {
+      "epoch": 16.944,
+      "grad_norm": 1.4410932064056396,
+      "learning_rate": 2e-05,
+      "loss": 0.03325985,
+      "step": 8472
+    },
+    {
+      "epoch": 16.946,
+      "grad_norm": 0.8621723055839539,
+      "learning_rate": 2e-05,
+      "loss": 0.0242934,
+      "step": 8473
+    },
+    {
+      "epoch": 16.948,
+      "grad_norm": 1.3893159627914429,
+      "learning_rate": 2e-05,
+      "loss": 0.05090608,
+      "step": 8474
+    },
+    {
+      "epoch": 16.95,
+      "grad_norm": 1.0170800685882568,
+      "learning_rate": 2e-05,
+      "loss": 0.03791419,
+      "step": 8475
+    },
+    {
+      "epoch": 16.951999999999998,
+      "grad_norm": 1.146407127380371,
+      "learning_rate": 2e-05,
+      "loss": 0.04361141,
+      "step": 8476
+    },
+    {
+      "epoch": 16.954,
+      "grad_norm": 1.6457027196884155,
+      "learning_rate": 2e-05,
+      "loss": 0.04759943,
+      "step": 8477
+    },
+    {
+      "epoch": 16.956,
+      "grad_norm": 1.1141819953918457,
+      "learning_rate": 2e-05,
+      "loss": 0.02985225,
+      "step": 8478
+    },
+    {
+      "epoch": 16.958,
+      "grad_norm": 1.0538251399993896,
+      "learning_rate": 2e-05,
+      "loss": 0.02504754,
+      "step": 8479
+    },
+    {
+      "epoch": 16.96,
+      "grad_norm": 1.3906612396240234,
+      "learning_rate": 2e-05,
+      "loss": 0.04034654,
+      "step": 8480
+    },
+    {
+      "epoch": 16.962,
+      "grad_norm": 1.228425145149231,
+      "learning_rate": 2e-05,
+      "loss": 0.05102978,
+      "step": 8481
+    },
+    {
+      "epoch": 16.964,
+      "grad_norm": 1.1037037372589111,
+      "learning_rate": 2e-05,
+      "loss": 0.02688233,
+      "step": 8482
+    },
+    {
+      "epoch": 16.966,
+      "grad_norm": 1.2161399126052856,
+      "learning_rate": 2e-05,
+      "loss": 0.03387916,
+      "step": 8483
+    },
+    {
+      "epoch": 16.968,
+      "grad_norm": 2.475914716720581,
+      "learning_rate": 2e-05,
+      "loss": 0.04503019,
+      "step": 8484
+    },
+    {
+      "epoch": 16.97,
+      "grad_norm": 2.1681854724884033,
+      "learning_rate": 2e-05,
+      "loss": 0.03996513,
+      "step": 8485
+    },
+    {
+      "epoch": 16.972,
+      "grad_norm": 2.7576403617858887,
+      "learning_rate": 2e-05,
+      "loss": 0.06164122,
+      "step": 8486
+    },
+    {
+      "epoch": 16.974,
+      "grad_norm": 1.2548010349273682,
+      "learning_rate": 2e-05,
+      "loss": 0.0354467,
+      "step": 8487
+    },
+    {
+      "epoch": 16.976,
+      "grad_norm": 1.4516505002975464,
+      "learning_rate": 2e-05,
+      "loss": 0.03556127,
+      "step": 8488
+    },
+    {
+      "epoch": 16.978,
+      "grad_norm": 2.353919744491577,
+      "learning_rate": 2e-05,
+      "loss": 0.03970582,
+      "step": 8489
+    },
+    {
+      "epoch": 16.98,
+      "grad_norm": 1.0244505405426025,
+      "learning_rate": 2e-05,
+      "loss": 0.03089332,
+      "step": 8490
+    },
+    {
+      "epoch": 16.982,
+      "grad_norm": 1.5581693649291992,
+      "learning_rate": 2e-05,
+      "loss": 0.03672716,
+      "step": 8491
+    },
+    {
+      "epoch": 16.984,
+      "grad_norm": 1.7653076648712158,
+      "learning_rate": 2e-05,
+      "loss": 0.04039037,
+      "step": 8492
+    },
+    {
+      "epoch": 16.986,
+      "grad_norm": 1.0333586931228638,
+      "learning_rate": 2e-05,
+      "loss": 0.03089204,
+      "step": 8493
+    },
+    {
+      "epoch": 16.988,
+      "grad_norm": 1.0320618152618408,
+      "learning_rate": 2e-05,
+      "loss": 0.0225993,
+      "step": 8494
+    },
+    {
+      "epoch": 16.99,
+      "grad_norm": 1.119234561920166,
+      "learning_rate": 2e-05,
+      "loss": 0.03324748,
+      "step": 8495
+    },
+    {
+      "epoch": 16.992,
+      "grad_norm": 1.9788683652877808,
+      "learning_rate": 2e-05,
+      "loss": 0.03121308,
+      "step": 8496
+    },
+    {
+      "epoch": 16.994,
+      "grad_norm": 1.3806347846984863,
+      "learning_rate": 2e-05,
+      "loss": 0.04654794,
+      "step": 8497
+    },
+    {
+      "epoch": 16.996,
+      "grad_norm": 1.1207362413406372,
+      "learning_rate": 2e-05,
+      "loss": 0.04140043,
+      "step": 8498
+    },
+    {
+      "epoch": 16.998,
+      "grad_norm": 1.039371371269226,
+      "learning_rate": 2e-05,
+      "loss": 0.03301085,
+      "step": 8499
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 2.4387052059173584,
+      "learning_rate": 2e-05,
+      "loss": 0.04641984,
+      "step": 8500
+    },
+    {
+      "epoch": 17.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.998,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9760479041916168,
+        "Equal_1": 0.996,
+        "Equal_2": 0.9680638722554891,
+        "Equal_3": 0.9181636726546906,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 0.998003992015968,
+        "LineComparison_3": 0.9940119760479041,
+        "Parallel_1": 0.9939879759519038,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.99,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.986,
+        "Perpendicular_3": 0.7755511022044088,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9936666666666667,
+        "PointLiesOnCircle_3": 0.99,
+        "PointLiesOnLine_1": 0.9959919839679359,
+        "PointLiesOnLine_2": 0.9919839679358717,
+        "PointLiesOnLine_3": 0.9820359281437125
+      },
+      "eval_runtime": 320.1697,
+      "eval_samples_per_second": 32.795,
+      "eval_steps_per_second": 0.656,
+      "step": 8500
+    },
+    {
+      "epoch": 17.002,
+      "grad_norm": 1.4980146884918213,
+      "learning_rate": 2e-05,
+      "loss": 0.04539901,
+      "step": 8501
+    },
+    {
+      "epoch": 17.004,
+      "grad_norm": 1.0780891180038452,
+      "learning_rate": 2e-05,
+      "loss": 0.03377684,
+      "step": 8502
+    },
+    {
+      "epoch": 17.006,
+      "grad_norm": 1.0425775051116943,
+      "learning_rate": 2e-05,
+      "loss": 0.02465243,
+      "step": 8503
+    },
+    {
+      "epoch": 17.008,
+      "grad_norm": 1.0172406435012817,
+      "learning_rate": 2e-05,
+      "loss": 0.02660679,
+      "step": 8504
+    },
+    {
+      "epoch": 17.01,
+      "grad_norm": 1.2166515588760376,
+      "learning_rate": 2e-05,
+      "loss": 0.04529484,
+      "step": 8505
+    },
+    {
+      "epoch": 17.012,
+      "grad_norm": 1.4045565128326416,
+      "learning_rate": 2e-05,
+      "loss": 0.04070731,
+      "step": 8506
+    },
+    {
+      "epoch": 17.014,
+      "grad_norm": 1.53168785572052,
+      "learning_rate": 2e-05,
+      "loss": 0.04683904,
+      "step": 8507
+    },
+    {
+      "epoch": 17.016,
+      "grad_norm": 1.4374054670333862,
+      "learning_rate": 2e-05,
+      "loss": 0.05705344,
+      "step": 8508
+    },
+    {
+      "epoch": 17.018,
+      "grad_norm": 1.0057858228683472,
+      "learning_rate": 2e-05,
+      "loss": 0.03421285,
+      "step": 8509
+    },
+    {
+      "epoch": 17.02,
+      "grad_norm": 1.1364305019378662,
+      "learning_rate": 2e-05,
+      "loss": 0.03909264,
+      "step": 8510
+    },
+    {
+      "epoch": 17.022,
+      "grad_norm": 1.0018776655197144,
+      "learning_rate": 2e-05,
+      "loss": 0.03444162,
+      "step": 8511
+    },
+    {
+      "epoch": 17.024,
+      "grad_norm": 1.4022345542907715,
+      "learning_rate": 2e-05,
+      "loss": 0.05768448,
+      "step": 8512
+    },
+    {
+      "epoch": 17.026,
+      "grad_norm": 2.5111844539642334,
+      "learning_rate": 2e-05,
+      "loss": 0.0339235,
+      "step": 8513
+    },
+    {
+      "epoch": 17.028,
+      "grad_norm": 1.921712040901184,
+      "learning_rate": 2e-05,
+      "loss": 0.03785212,
+      "step": 8514
+    },
+    {
+      "epoch": 17.03,
+      "grad_norm": 1.1148395538330078,
+      "learning_rate": 2e-05,
+      "loss": 0.02421642,
+      "step": 8515
+    },
+    {
+      "epoch": 17.032,
+      "grad_norm": 1.0764633417129517,
+      "learning_rate": 2e-05,
+      "loss": 0.03142115,
+      "step": 8516
+    },
+    {
+      "epoch": 17.034,
+      "grad_norm": 1.4113231897354126,
+      "learning_rate": 2e-05,
+      "loss": 0.03779317,
+      "step": 8517
+    },
+    {
+      "epoch": 17.036,
+      "grad_norm": 1.0931544303894043,
+      "learning_rate": 2e-05,
+      "loss": 0.02576648,
+      "step": 8518
+    },
+    {
+      "epoch": 17.038,
+      "grad_norm": 1.8094255924224854,
+      "learning_rate": 2e-05,
+      "loss": 0.04606591,
+      "step": 8519
+    },
+    {
+      "epoch": 17.04,
+      "grad_norm": 1.4644229412078857,
+      "learning_rate": 2e-05,
+      "loss": 0.043492,
+      "step": 8520
+    },
+    {
+      "epoch": 17.042,
+      "grad_norm": 1.4266523122787476,
+      "learning_rate": 2e-05,
+      "loss": 0.03921729,
+      "step": 8521
+    },
+    {
+      "epoch": 17.044,
+      "grad_norm": 1.3539372682571411,
+      "learning_rate": 2e-05,
+      "loss": 0.03920222,
+      "step": 8522
+    },
+    {
+      "epoch": 17.046,
+      "grad_norm": 1.028294563293457,
+      "learning_rate": 2e-05,
+      "loss": 0.03639862,
+      "step": 8523
+    },
+    {
+      "epoch": 17.048,
+      "grad_norm": 1.2716553211212158,
+      "learning_rate": 2e-05,
+      "loss": 0.03918849,
+      "step": 8524
+    },
+    {
+      "epoch": 17.05,
+      "grad_norm": 0.8702855706214905,
+      "learning_rate": 2e-05,
+      "loss": 0.02787583,
+      "step": 8525
+    },
+    {
+      "epoch": 17.052,
+      "grad_norm": 1.2607735395431519,
+      "learning_rate": 2e-05,
+      "loss": 0.04798802,
+      "step": 8526
+    },
+    {
+      "epoch": 17.054,
+      "grad_norm": 1.8052009344100952,
+      "learning_rate": 2e-05,
+      "loss": 0.04934546,
+      "step": 8527
+    },
+    {
+      "epoch": 17.056,
+      "grad_norm": 1.3691691160202026,
+      "learning_rate": 2e-05,
+      "loss": 0.02806879,
+      "step": 8528
+    },
+    {
+      "epoch": 17.058,
+      "grad_norm": 2.2015974521636963,
+      "learning_rate": 2e-05,
+      "loss": 0.03393783,
+      "step": 8529
+    },
+    {
+      "epoch": 17.06,
+      "grad_norm": 1.7185814380645752,
+      "learning_rate": 2e-05,
+      "loss": 0.04044708,
+      "step": 8530
+    },
+    {
+      "epoch": 17.062,
+      "grad_norm": 1.6913961172103882,
+      "learning_rate": 2e-05,
+      "loss": 0.04489831,
+      "step": 8531
+    },
+    {
+      "epoch": 17.064,
+      "grad_norm": 1.6114678382873535,
+      "learning_rate": 2e-05,
+      "loss": 0.06014886,
+      "step": 8532
+    },
+    {
+      "epoch": 17.066,
+      "grad_norm": 1.866416335105896,
+      "learning_rate": 2e-05,
+      "loss": 0.05753294,
+      "step": 8533
+    },
+    {
+      "epoch": 17.068,
+      "grad_norm": 2.823669910430908,
+      "learning_rate": 2e-05,
+      "loss": 0.0364582,
+      "step": 8534
+    },
+    {
+      "epoch": 17.07,
+      "grad_norm": 1.3200300931930542,
+      "learning_rate": 2e-05,
+      "loss": 0.02338468,
+      "step": 8535
+    },
+    {
+      "epoch": 17.072,
+      "grad_norm": 1.4321677684783936,
+      "learning_rate": 2e-05,
+      "loss": 0.03681827,
+      "step": 8536
+    },
+    {
+      "epoch": 17.074,
+      "grad_norm": 1.1837363243103027,
+      "learning_rate": 2e-05,
+      "loss": 0.03107641,
+      "step": 8537
+    },
+    {
+      "epoch": 17.076,
+      "grad_norm": 1.5008882284164429,
+      "learning_rate": 2e-05,
+      "loss": 0.04763446,
+      "step": 8538
+    },
+    {
+      "epoch": 17.078,
+      "grad_norm": 0.9194531440734863,
+      "learning_rate": 2e-05,
+      "loss": 0.03133226,
+      "step": 8539
+    },
+    {
+      "epoch": 17.08,
+      "grad_norm": 1.7025022506713867,
+      "learning_rate": 2e-05,
+      "loss": 0.0493839,
+      "step": 8540
+    },
+    {
+      "epoch": 17.082,
+      "grad_norm": 1.8602598905563354,
+      "learning_rate": 2e-05,
+      "loss": 0.0486247,
+      "step": 8541
+    },
+    {
+      "epoch": 17.084,
+      "grad_norm": 6.84015417098999,
+      "learning_rate": 2e-05,
+      "loss": 0.02411661,
+      "step": 8542
+    },
+    {
+      "epoch": 17.086,
+      "grad_norm": 1.2023508548736572,
+      "learning_rate": 2e-05,
+      "loss": 0.03706846,
+      "step": 8543
+    },
+    {
+      "epoch": 17.088,
+      "grad_norm": 1.4911589622497559,
+      "learning_rate": 2e-05,
+      "loss": 0.05094575,
+      "step": 8544
+    },
+    {
+      "epoch": 17.09,
+      "grad_norm": 0.8137642741203308,
+      "learning_rate": 2e-05,
+      "loss": 0.02076185,
+      "step": 8545
+    },
+    {
+      "epoch": 17.092,
+      "grad_norm": 1.6235312223434448,
+      "learning_rate": 2e-05,
+      "loss": 0.04601521,
+      "step": 8546
+    },
+    {
+      "epoch": 17.094,
+      "grad_norm": 2.2012600898742676,
+      "learning_rate": 2e-05,
+      "loss": 0.04425218,
+      "step": 8547
+    },
+    {
+      "epoch": 17.096,
+      "grad_norm": 1.039987564086914,
+      "learning_rate": 2e-05,
+      "loss": 0.02848525,
+      "step": 8548
+    },
+    {
+      "epoch": 17.098,
+      "grad_norm": 1.7109042406082153,
+      "learning_rate": 2e-05,
+      "loss": 0.06454369,
+      "step": 8549
+    },
+    {
+      "epoch": 17.1,
+      "grad_norm": 1.8760957717895508,
+      "learning_rate": 2e-05,
+      "loss": 0.02889769,
+      "step": 8550
+    },
+    {
+      "epoch": 17.102,
+      "grad_norm": 1.0814695358276367,
+      "learning_rate": 2e-05,
+      "loss": 0.0295868,
+      "step": 8551
+    },
+    {
+      "epoch": 17.104,
+      "grad_norm": 1.5878161191940308,
+      "learning_rate": 2e-05,
+      "loss": 0.03735021,
+      "step": 8552
+    },
+    {
+      "epoch": 17.106,
+      "grad_norm": 1.4773187637329102,
+      "learning_rate": 2e-05,
+      "loss": 0.04390514,
+      "step": 8553
+    },
+    {
+      "epoch": 17.108,
+      "grad_norm": 1.8776631355285645,
+      "learning_rate": 2e-05,
+      "loss": 0.04321947,
+      "step": 8554
+    },
+    {
+      "epoch": 17.11,
+      "grad_norm": 1.309813141822815,
+      "learning_rate": 2e-05,
+      "loss": 0.03866652,
+      "step": 8555
+    },
+    {
+      "epoch": 17.112,
+      "grad_norm": 0.9536056518554688,
+      "learning_rate": 2e-05,
+      "loss": 0.02671996,
+      "step": 8556
+    },
+    {
+      "epoch": 17.114,
+      "grad_norm": 1.3391567468643188,
+      "learning_rate": 2e-05,
+      "loss": 0.02907237,
+      "step": 8557
+    },
+    {
+      "epoch": 17.116,
+      "grad_norm": 1.0060744285583496,
+      "learning_rate": 2e-05,
+      "loss": 0.03439762,
+      "step": 8558
+    },
+    {
+      "epoch": 17.118,
+      "grad_norm": 0.9535341262817383,
+      "learning_rate": 2e-05,
+      "loss": 0.02899779,
+      "step": 8559
+    },
+    {
+      "epoch": 17.12,
+      "grad_norm": 0.7711392045021057,
+      "learning_rate": 2e-05,
+      "loss": 0.02512438,
+      "step": 8560
+    },
+    {
+      "epoch": 17.122,
+      "grad_norm": 1.0597705841064453,
+      "learning_rate": 2e-05,
+      "loss": 0.03412605,
+      "step": 8561
+    },
+    {
+      "epoch": 17.124,
+      "grad_norm": 1.8042422533035278,
+      "learning_rate": 2e-05,
+      "loss": 0.05712911,
+      "step": 8562
+    },
+    {
+      "epoch": 17.126,
+      "grad_norm": 1.9867520332336426,
+      "learning_rate": 2e-05,
+      "loss": 0.04097403,
+      "step": 8563
+    },
+    {
+      "epoch": 17.128,
+      "grad_norm": 1.8169209957122803,
+      "learning_rate": 2e-05,
+      "loss": 0.05447955,
+      "step": 8564
+    },
+    {
+      "epoch": 17.13,
+      "grad_norm": 1.8252878189086914,
+      "learning_rate": 2e-05,
+      "loss": 0.03339032,
+      "step": 8565
+    },
+    {
+      "epoch": 17.132,
+      "grad_norm": 2.2705142498016357,
+      "learning_rate": 2e-05,
+      "loss": 0.04022797,
+      "step": 8566
+    },
+    {
+      "epoch": 17.134,
+      "grad_norm": 1.9331085681915283,
+      "learning_rate": 2e-05,
+      "loss": 0.03970163,
+      "step": 8567
+    },
+    {
+      "epoch": 17.136,
+      "grad_norm": 1.5278258323669434,
+      "learning_rate": 2e-05,
+      "loss": 0.03794595,
+      "step": 8568
+    },
+    {
+      "epoch": 17.138,
+      "grad_norm": 2.026817560195923,
+      "learning_rate": 2e-05,
+      "loss": 0.0471566,
+      "step": 8569
+    },
+    {
+      "epoch": 17.14,
+      "grad_norm": 1.5275219678878784,
+      "learning_rate": 2e-05,
+      "loss": 0.04161655,
+      "step": 8570
+    },
+    {
+      "epoch": 17.142,
+      "grad_norm": 1.6281226873397827,
+      "learning_rate": 2e-05,
+      "loss": 0.03525121,
+      "step": 8571
+    },
+    {
+      "epoch": 17.144,
+      "grad_norm": 2.1644468307495117,
+      "learning_rate": 2e-05,
+      "loss": 0.05372618,
+      "step": 8572
+    },
+    {
+      "epoch": 17.146,
+      "grad_norm": 1.3852871656417847,
+      "learning_rate": 2e-05,
+      "loss": 0.04645745,
+      "step": 8573
+    },
+    {
+      "epoch": 17.148,
+      "grad_norm": 1.2143161296844482,
+      "learning_rate": 2e-05,
+      "loss": 0.02777028,
+      "step": 8574
+    },
+    {
+      "epoch": 17.15,
+      "grad_norm": 1.1853028535842896,
+      "learning_rate": 2e-05,
+      "loss": 0.03904351,
+      "step": 8575
+    },
+    {
+      "epoch": 17.152,
+      "grad_norm": 1.9432642459869385,
+      "learning_rate": 2e-05,
+      "loss": 0.04074139,
+      "step": 8576
+    },
+    {
+      "epoch": 17.154,
+      "grad_norm": 1.6851083040237427,
+      "learning_rate": 2e-05,
+      "loss": 0.05192745,
+      "step": 8577
+    },
+    {
+      "epoch": 17.156,
+      "grad_norm": 1.3564956188201904,
+      "learning_rate": 2e-05,
+      "loss": 0.03743386,
+      "step": 8578
+    },
+    {
+      "epoch": 17.158,
+      "grad_norm": 1.7471429109573364,
+      "learning_rate": 2e-05,
+      "loss": 0.0443368,
+      "step": 8579
+    },
+    {
+      "epoch": 17.16,
+      "grad_norm": 1.41841459274292,
+      "learning_rate": 2e-05,
+      "loss": 0.04000156,
+      "step": 8580
+    },
+    {
+      "epoch": 17.162,
+      "grad_norm": 0.9492871761322021,
+      "learning_rate": 2e-05,
+      "loss": 0.0318904,
+      "step": 8581
+    },
+    {
+      "epoch": 17.164,
+      "grad_norm": 1.0863397121429443,
+      "learning_rate": 2e-05,
+      "loss": 0.02380545,
+      "step": 8582
+    },
+    {
+      "epoch": 17.166,
+      "grad_norm": 3.9473729133605957,
+      "learning_rate": 2e-05,
+      "loss": 0.03920302,
+      "step": 8583
+    },
+    {
+      "epoch": 17.168,
+      "grad_norm": 1.176370620727539,
+      "learning_rate": 2e-05,
+      "loss": 0.04776629,
+      "step": 8584
+    },
+    {
+      "epoch": 17.17,
+      "grad_norm": 1.3243099451065063,
+      "learning_rate": 2e-05,
+      "loss": 0.03770189,
+      "step": 8585
+    },
+    {
+      "epoch": 17.172,
+      "grad_norm": 1.1418389081954956,
+      "learning_rate": 2e-05,
+      "loss": 0.02544946,
+      "step": 8586
+    },
+    {
+      "epoch": 17.174,
+      "grad_norm": 1.9030910730361938,
+      "learning_rate": 2e-05,
+      "loss": 0.03700698,
+      "step": 8587
+    },
+    {
+      "epoch": 17.176,
+      "grad_norm": 1.3709733486175537,
+      "learning_rate": 2e-05,
+      "loss": 0.04343133,
+      "step": 8588
+    },
+    {
+      "epoch": 17.178,
+      "grad_norm": 1.4506902694702148,
+      "learning_rate": 2e-05,
+      "loss": 0.03628192,
+      "step": 8589
+    },
+    {
+      "epoch": 17.18,
+      "grad_norm": 1.3070485591888428,
+      "learning_rate": 2e-05,
+      "loss": 0.04146688,
+      "step": 8590
+    },
+    {
+      "epoch": 17.182,
+      "grad_norm": 1.234598994255066,
+      "learning_rate": 2e-05,
+      "loss": 0.03088631,
+      "step": 8591
+    },
+    {
+      "epoch": 17.184,
+      "grad_norm": 2.7125167846679688,
+      "learning_rate": 2e-05,
+      "loss": 0.03890502,
+      "step": 8592
+    },
+    {
+      "epoch": 17.186,
+      "grad_norm": 1.535182237625122,
+      "learning_rate": 2e-05,
+      "loss": 0.04280488,
+      "step": 8593
+    },
+    {
+      "epoch": 17.188,
+      "grad_norm": 1.0927315950393677,
+      "learning_rate": 2e-05,
+      "loss": 0.0398199,
+      "step": 8594
+    },
+    {
+      "epoch": 17.19,
+      "grad_norm": 2.3022119998931885,
+      "learning_rate": 2e-05,
+      "loss": 0.04995221,
+      "step": 8595
+    },
+    {
+      "epoch": 17.192,
+      "grad_norm": 1.2546242475509644,
+      "learning_rate": 2e-05,
+      "loss": 0.03015491,
+      "step": 8596
+    },
+    {
+      "epoch": 17.194,
+      "grad_norm": 1.2892979383468628,
+      "learning_rate": 2e-05,
+      "loss": 0.02786731,
+      "step": 8597
+    },
+    {
+      "epoch": 17.196,
+      "grad_norm": 1.2983757257461548,
+      "learning_rate": 2e-05,
+      "loss": 0.03557572,
+      "step": 8598
+    },
+    {
+      "epoch": 17.198,
+      "grad_norm": 1.1861157417297363,
+      "learning_rate": 2e-05,
+      "loss": 0.03920743,
+      "step": 8599
+    },
+    {
+      "epoch": 17.2,
+      "grad_norm": 1.5425999164581299,
+      "learning_rate": 2e-05,
+      "loss": 0.03938346,
+      "step": 8600
+    },
+    {
+      "epoch": 17.202,
+      "grad_norm": 1.861615538597107,
+      "learning_rate": 2e-05,
+      "loss": 0.03425018,
+      "step": 8601
+    },
+    {
+      "epoch": 17.204,
+      "grad_norm": 1.7203572988510132,
+      "learning_rate": 2e-05,
+      "loss": 0.04107557,
+      "step": 8602
+    },
+    {
+      "epoch": 17.206,
+      "grad_norm": 1.869676113128662,
+      "learning_rate": 2e-05,
+      "loss": 0.02804127,
+      "step": 8603
+    },
+    {
+      "epoch": 17.208,
+      "grad_norm": 1.1111414432525635,
+      "learning_rate": 2e-05,
+      "loss": 0.03163296,
+      "step": 8604
+    },
+    {
+      "epoch": 17.21,
+      "grad_norm": 1.1464769840240479,
+      "learning_rate": 2e-05,
+      "loss": 0.03462142,
+      "step": 8605
+    },
+    {
+      "epoch": 17.212,
+      "grad_norm": 1.6060582399368286,
+      "learning_rate": 2e-05,
+      "loss": 0.06490205,
+      "step": 8606
+    },
+    {
+      "epoch": 17.214,
+      "grad_norm": 1.082814335823059,
+      "learning_rate": 2e-05,
+      "loss": 0.02265838,
+      "step": 8607
+    },
+    {
+      "epoch": 17.216,
+      "grad_norm": 1.1353710889816284,
+      "learning_rate": 2e-05,
+      "loss": 0.03856842,
+      "step": 8608
+    },
+    {
+      "epoch": 17.218,
+      "grad_norm": 1.3957148790359497,
+      "learning_rate": 2e-05,
+      "loss": 0.03138808,
+      "step": 8609
+    },
+    {
+      "epoch": 17.22,
+      "grad_norm": 1.2920960187911987,
+      "learning_rate": 2e-05,
+      "loss": 0.0430844,
+      "step": 8610
+    },
+    {
+      "epoch": 17.222,
+      "grad_norm": 1.2752081155776978,
+      "learning_rate": 2e-05,
+      "loss": 0.04843235,
+      "step": 8611
+    },
+    {
+      "epoch": 17.224,
+      "grad_norm": 1.4055542945861816,
+      "learning_rate": 2e-05,
+      "loss": 0.03793164,
+      "step": 8612
+    },
+    {
+      "epoch": 17.226,
+      "grad_norm": 1.3925482034683228,
+      "learning_rate": 2e-05,
+      "loss": 0.03343806,
+      "step": 8613
+    },
+    {
+      "epoch": 17.228,
+      "grad_norm": 1.2577686309814453,
+      "learning_rate": 2e-05,
+      "loss": 0.03357682,
+      "step": 8614
+    },
+    {
+      "epoch": 17.23,
+      "grad_norm": 1.5118682384490967,
+      "learning_rate": 2e-05,
+      "loss": 0.05048358,
+      "step": 8615
+    },
+    {
+      "epoch": 17.232,
+      "grad_norm": 1.094104528427124,
+      "learning_rate": 2e-05,
+      "loss": 0.02639047,
+      "step": 8616
+    },
+    {
+      "epoch": 17.234,
+      "grad_norm": 1.337664008140564,
+      "learning_rate": 2e-05,
+      "loss": 0.03451654,
+      "step": 8617
+    },
+    {
+      "epoch": 17.236,
+      "grad_norm": 1.4902300834655762,
+      "learning_rate": 2e-05,
+      "loss": 0.03678834,
+      "step": 8618
+    },
+    {
+      "epoch": 17.238,
+      "grad_norm": 0.9843270778656006,
+      "learning_rate": 2e-05,
+      "loss": 0.02527812,
+      "step": 8619
+    },
+    {
+      "epoch": 17.24,
+      "grad_norm": 1.1602085828781128,
+      "learning_rate": 2e-05,
+      "loss": 0.02774549,
+      "step": 8620
+    },
+    {
+      "epoch": 17.242,
+      "grad_norm": 1.4860682487487793,
+      "learning_rate": 2e-05,
+      "loss": 0.04466415,
+      "step": 8621
+    },
+    {
+      "epoch": 17.244,
+      "grad_norm": 1.3373606204986572,
+      "learning_rate": 2e-05,
+      "loss": 0.03324435,
+      "step": 8622
+    },
+    {
+      "epoch": 17.246,
+      "grad_norm": 1.4892737865447998,
+      "learning_rate": 2e-05,
+      "loss": 0.03874163,
+      "step": 8623
+    },
+    {
+      "epoch": 17.248,
+      "grad_norm": 1.1000738143920898,
+      "learning_rate": 2e-05,
+      "loss": 0.03442953,
+      "step": 8624
+    },
+    {
+      "epoch": 17.25,
+      "grad_norm": 1.588263988494873,
+      "learning_rate": 2e-05,
+      "loss": 0.05064608,
+      "step": 8625
+    },
+    {
+      "epoch": 17.252,
+      "grad_norm": 2.7729272842407227,
+      "learning_rate": 2e-05,
+      "loss": 0.03416018,
+      "step": 8626
+    },
+    {
+      "epoch": 17.254,
+      "grad_norm": 1.785430669784546,
+      "learning_rate": 2e-05,
+      "loss": 0.04151814,
+      "step": 8627
+    },
+    {
+      "epoch": 17.256,
+      "grad_norm": 1.0420557260513306,
+      "learning_rate": 2e-05,
+      "loss": 0.03002853,
+      "step": 8628
+    },
+    {
+      "epoch": 17.258,
+      "grad_norm": 1.5617362260818481,
+      "learning_rate": 2e-05,
+      "loss": 0.0326589,
+      "step": 8629
+    },
+    {
+      "epoch": 17.26,
+      "grad_norm": 1.3772300481796265,
+      "learning_rate": 2e-05,
+      "loss": 0.03914788,
+      "step": 8630
+    },
+    {
+      "epoch": 17.262,
+      "grad_norm": 1.2935206890106201,
+      "learning_rate": 2e-05,
+      "loss": 0.03735418,
+      "step": 8631
+    },
+    {
+      "epoch": 17.264,
+      "grad_norm": 1.2990622520446777,
+      "learning_rate": 2e-05,
+      "loss": 0.03097917,
+      "step": 8632
+    },
+    {
+      "epoch": 17.266,
+      "grad_norm": 1.4346680641174316,
+      "learning_rate": 2e-05,
+      "loss": 0.03221875,
+      "step": 8633
+    },
+    {
+      "epoch": 17.268,
+      "grad_norm": 1.139218807220459,
+      "learning_rate": 2e-05,
+      "loss": 0.03914971,
+      "step": 8634
+    },
+    {
+      "epoch": 17.27,
+      "grad_norm": 1.8043378591537476,
+      "learning_rate": 2e-05,
+      "loss": 0.05375669,
+      "step": 8635
+    },
+    {
+      "epoch": 17.272,
+      "grad_norm": 1.302628993988037,
+      "learning_rate": 2e-05,
+      "loss": 0.03872466,
+      "step": 8636
+    },
+    {
+      "epoch": 17.274,
+      "grad_norm": 1.0800704956054688,
+      "learning_rate": 2e-05,
+      "loss": 0.03145566,
+      "step": 8637
+    },
+    {
+      "epoch": 17.276,
+      "grad_norm": 1.0463593006134033,
+      "learning_rate": 2e-05,
+      "loss": 0.03460503,
+      "step": 8638
+    },
+    {
+      "epoch": 17.278,
+      "grad_norm": 1.3378190994262695,
+      "learning_rate": 2e-05,
+      "loss": 0.05418106,
+      "step": 8639
+    },
+    {
+      "epoch": 17.28,
+      "grad_norm": 1.3952049016952515,
+      "learning_rate": 2e-05,
+      "loss": 0.04103397,
+      "step": 8640
+    },
+    {
+      "epoch": 17.282,
+      "grad_norm": 1.4025013446807861,
+      "learning_rate": 2e-05,
+      "loss": 0.03793554,
+      "step": 8641
+    },
+    {
+      "epoch": 17.284,
+      "grad_norm": 1.229583978652954,
+      "learning_rate": 2e-05,
+      "loss": 0.05052347,
+      "step": 8642
+    },
+    {
+      "epoch": 17.286,
+      "grad_norm": 1.4108766317367554,
+      "learning_rate": 2e-05,
+      "loss": 0.02928887,
+      "step": 8643
+    },
+    {
+      "epoch": 17.288,
+      "grad_norm": 1.744924783706665,
+      "learning_rate": 2e-05,
+      "loss": 0.04452213,
+      "step": 8644
+    },
+    {
+      "epoch": 17.29,
+      "grad_norm": 1.015102744102478,
+      "learning_rate": 2e-05,
+      "loss": 0.0326981,
+      "step": 8645
+    },
+    {
+      "epoch": 17.292,
+      "grad_norm": 2.1634633541107178,
+      "learning_rate": 2e-05,
+      "loss": 0.04052316,
+      "step": 8646
+    },
+    {
+      "epoch": 17.294,
+      "grad_norm": 0.9109699130058289,
+      "learning_rate": 2e-05,
+      "loss": 0.02265767,
+      "step": 8647
+    },
+    {
+      "epoch": 17.296,
+      "grad_norm": 1.964977502822876,
+      "learning_rate": 2e-05,
+      "loss": 0.07206996,
+      "step": 8648
+    },
+    {
+      "epoch": 17.298,
+      "grad_norm": 1.3658807277679443,
+      "learning_rate": 2e-05,
+      "loss": 0.04043622,
+      "step": 8649
+    },
+    {
+      "epoch": 17.3,
+      "grad_norm": 1.3623096942901611,
+      "learning_rate": 2e-05,
+      "loss": 0.02930884,
+      "step": 8650
+    },
+    {
+      "epoch": 17.302,
+      "grad_norm": 1.2337470054626465,
+      "learning_rate": 2e-05,
+      "loss": 0.03874534,
+      "step": 8651
+    },
+    {
+      "epoch": 17.304,
+      "grad_norm": 0.892915666103363,
+      "learning_rate": 2e-05,
+      "loss": 0.03198083,
+      "step": 8652
+    },
+    {
+      "epoch": 17.306,
+      "grad_norm": 0.9995888471603394,
+      "learning_rate": 2e-05,
+      "loss": 0.03072457,
+      "step": 8653
+    },
+    {
+      "epoch": 17.308,
+      "grad_norm": 1.4063588380813599,
+      "learning_rate": 2e-05,
+      "loss": 0.04875659,
+      "step": 8654
+    },
+    {
+      "epoch": 17.31,
+      "grad_norm": 1.2465662956237793,
+      "learning_rate": 2e-05,
+      "loss": 0.03184005,
+      "step": 8655
+    },
+    {
+      "epoch": 17.312,
+      "grad_norm": 1.2158682346343994,
+      "learning_rate": 2e-05,
+      "loss": 0.03214359,
+      "step": 8656
+    },
+    {
+      "epoch": 17.314,
+      "grad_norm": 0.8416528701782227,
+      "learning_rate": 2e-05,
+      "loss": 0.02059457,
+      "step": 8657
+    },
+    {
+      "epoch": 17.316,
+      "grad_norm": 1.3326908349990845,
+      "learning_rate": 2e-05,
+      "loss": 0.04002319,
+      "step": 8658
+    },
+    {
+      "epoch": 17.318,
+      "grad_norm": 1.201299786567688,
+      "learning_rate": 2e-05,
+      "loss": 0.03972005,
+      "step": 8659
+    },
+    {
+      "epoch": 17.32,
+      "grad_norm": 0.8614779114723206,
+      "learning_rate": 2e-05,
+      "loss": 0.02517816,
+      "step": 8660
+    },
+    {
+      "epoch": 17.322,
+      "grad_norm": 1.0429071187973022,
+      "learning_rate": 2e-05,
+      "loss": 0.0332981,
+      "step": 8661
+    },
+    {
+      "epoch": 17.324,
+      "grad_norm": 1.0974829196929932,
+      "learning_rate": 2e-05,
+      "loss": 0.0395813,
+      "step": 8662
+    },
+    {
+      "epoch": 17.326,
+      "grad_norm": 1.7790201902389526,
+      "learning_rate": 2e-05,
+      "loss": 0.04459688,
+      "step": 8663
+    },
+    {
+      "epoch": 17.328,
+      "grad_norm": 2.299643039703369,
+      "learning_rate": 2e-05,
+      "loss": 0.04014507,
+      "step": 8664
+    },
+    {
+      "epoch": 17.33,
+      "grad_norm": 1.0453588962554932,
+      "learning_rate": 2e-05,
+      "loss": 0.02698977,
+      "step": 8665
+    },
+    {
+      "epoch": 17.332,
+      "grad_norm": 0.8015127778053284,
+      "learning_rate": 2e-05,
+      "loss": 0.02366449,
+      "step": 8666
+    },
+    {
+      "epoch": 17.334,
+      "grad_norm": 1.241165280342102,
+      "learning_rate": 2e-05,
+      "loss": 0.03793594,
+      "step": 8667
+    },
+    {
+      "epoch": 17.336,
+      "grad_norm": 3.191179037094116,
+      "learning_rate": 2e-05,
+      "loss": 0.05078402,
+      "step": 8668
+    },
+    {
+      "epoch": 17.338,
+      "grad_norm": 1.3432191610336304,
+      "learning_rate": 2e-05,
+      "loss": 0.03950462,
+      "step": 8669
+    },
+    {
+      "epoch": 17.34,
+      "grad_norm": 1.2250608205795288,
+      "learning_rate": 2e-05,
+      "loss": 0.04277131,
+      "step": 8670
+    },
+    {
+      "epoch": 17.342,
+      "grad_norm": 1.4782729148864746,
+      "learning_rate": 2e-05,
+      "loss": 0.03562623,
+      "step": 8671
+    },
+    {
+      "epoch": 17.344,
+      "grad_norm": 1.5145196914672852,
+      "learning_rate": 2e-05,
+      "loss": 0.03654065,
+      "step": 8672
+    },
+    {
+      "epoch": 17.346,
+      "grad_norm": 1.0246665477752686,
+      "learning_rate": 2e-05,
+      "loss": 0.03273585,
+      "step": 8673
+    },
+    {
+      "epoch": 17.348,
+      "grad_norm": 0.9382162690162659,
+      "learning_rate": 2e-05,
+      "loss": 0.03068813,
+      "step": 8674
+    },
+    {
+      "epoch": 17.35,
+      "grad_norm": 1.0336867570877075,
+      "learning_rate": 2e-05,
+      "loss": 0.03111012,
+      "step": 8675
+    },
+    {
+      "epoch": 17.352,
+      "grad_norm": 1.593807339668274,
+      "learning_rate": 2e-05,
+      "loss": 0.0379719,
+      "step": 8676
+    },
+    {
+      "epoch": 17.354,
+      "grad_norm": 1.819577693939209,
+      "learning_rate": 2e-05,
+      "loss": 0.03360122,
+      "step": 8677
+    },
+    {
+      "epoch": 17.356,
+      "grad_norm": 1.5745559930801392,
+      "learning_rate": 2e-05,
+      "loss": 0.05366817,
+      "step": 8678
+    },
+    {
+      "epoch": 17.358,
+      "grad_norm": 1.227429986000061,
+      "learning_rate": 2e-05,
+      "loss": 0.04186604,
+      "step": 8679
+    },
+    {
+      "epoch": 17.36,
+      "grad_norm": 1.5559841394424438,
+      "learning_rate": 2e-05,
+      "loss": 0.0397791,
+      "step": 8680
+    },
+    {
+      "epoch": 17.362,
+      "grad_norm": 1.0777933597564697,
+      "learning_rate": 2e-05,
+      "loss": 0.02661498,
+      "step": 8681
+    },
+    {
+      "epoch": 17.364,
+      "grad_norm": 1.2830750942230225,
+      "learning_rate": 2e-05,
+      "loss": 0.03172344,
+      "step": 8682
+    },
+    {
+      "epoch": 17.366,
+      "grad_norm": 0.9228249788284302,
+      "learning_rate": 2e-05,
+      "loss": 0.03080669,
+      "step": 8683
+    },
+    {
+      "epoch": 17.368,
+      "grad_norm": 1.545691967010498,
+      "learning_rate": 2e-05,
+      "loss": 0.04729048,
+      "step": 8684
+    },
+    {
+      "epoch": 17.37,
+      "grad_norm": 2.842621326446533,
+      "learning_rate": 2e-05,
+      "loss": 0.03637499,
+      "step": 8685
+    },
+    {
+      "epoch": 17.372,
+      "grad_norm": 2.176117181777954,
+      "learning_rate": 2e-05,
+      "loss": 0.04287196,
+      "step": 8686
+    },
+    {
+      "epoch": 17.374,
+      "grad_norm": 2.0805728435516357,
+      "learning_rate": 2e-05,
+      "loss": 0.03709119,
+      "step": 8687
+    },
+    {
+      "epoch": 17.376,
+      "grad_norm": 1.0376954078674316,
+      "learning_rate": 2e-05,
+      "loss": 0.02746601,
+      "step": 8688
+    },
+    {
+      "epoch": 17.378,
+      "grad_norm": 4.563030242919922,
+      "learning_rate": 2e-05,
+      "loss": 0.0574781,
+      "step": 8689
+    },
+    {
+      "epoch": 17.38,
+      "grad_norm": 1.4875022172927856,
+      "learning_rate": 2e-05,
+      "loss": 0.0384874,
+      "step": 8690
+    },
+    {
+      "epoch": 17.382,
+      "grad_norm": 1.295417070388794,
+      "learning_rate": 2e-05,
+      "loss": 0.04185217,
+      "step": 8691
+    },
+    {
+      "epoch": 17.384,
+      "grad_norm": 1.428702712059021,
+      "learning_rate": 2e-05,
+      "loss": 0.04110858,
+      "step": 8692
+    },
+    {
+      "epoch": 17.386,
+      "grad_norm": 1.7342407703399658,
+      "learning_rate": 2e-05,
+      "loss": 0.04000304,
+      "step": 8693
+    },
+    {
+      "epoch": 17.388,
+      "grad_norm": 1.2365176677703857,
+      "learning_rate": 2e-05,
+      "loss": 0.0342959,
+      "step": 8694
+    },
+    {
+      "epoch": 17.39,
+      "grad_norm": 2.3056838512420654,
+      "learning_rate": 2e-05,
+      "loss": 0.04070599,
+      "step": 8695
+    },
+    {
+      "epoch": 17.392,
+      "grad_norm": 1.948683261871338,
+      "learning_rate": 2e-05,
+      "loss": 0.03871055,
+      "step": 8696
+    },
+    {
+      "epoch": 17.394,
+      "grad_norm": 3.8803670406341553,
+      "learning_rate": 2e-05,
+      "loss": 0.03387075,
+      "step": 8697
+    },
+    {
+      "epoch": 17.396,
+      "grad_norm": 1.0368173122406006,
+      "learning_rate": 2e-05,
+      "loss": 0.03241309,
+      "step": 8698
+    },
+    {
+      "epoch": 17.398,
+      "grad_norm": 3.3204972743988037,
+      "learning_rate": 2e-05,
+      "loss": 0.06573039,
+      "step": 8699
+    },
+    {
+      "epoch": 17.4,
+      "grad_norm": 1.3388797044754028,
+      "learning_rate": 2e-05,
+      "loss": 0.03789306,
+      "step": 8700
+    },
+    {
+      "epoch": 17.402,
+      "grad_norm": 2.9801526069641113,
+      "learning_rate": 2e-05,
+      "loss": 0.03678794,
+      "step": 8701
+    },
+    {
+      "epoch": 17.404,
+      "grad_norm": 1.5335742235183716,
+      "learning_rate": 2e-05,
+      "loss": 0.0323316,
+      "step": 8702
+    },
+    {
+      "epoch": 17.406,
+      "grad_norm": 1.2214679718017578,
+      "learning_rate": 2e-05,
+      "loss": 0.02849279,
+      "step": 8703
+    },
+    {
+      "epoch": 17.408,
+      "grad_norm": 1.08639395236969,
+      "learning_rate": 2e-05,
+      "loss": 0.04005399,
+      "step": 8704
+    },
+    {
+      "epoch": 17.41,
+      "grad_norm": 1.7272155284881592,
+      "learning_rate": 2e-05,
+      "loss": 0.04677524,
+      "step": 8705
+    },
+    {
+      "epoch": 17.412,
+      "grad_norm": 1.4193323850631714,
+      "learning_rate": 2e-05,
+      "loss": 0.04003004,
+      "step": 8706
+    },
+    {
+      "epoch": 17.414,
+      "grad_norm": 1.1759463548660278,
+      "learning_rate": 2e-05,
+      "loss": 0.0381237,
+      "step": 8707
+    },
+    {
+      "epoch": 17.416,
+      "grad_norm": 2.2209203243255615,
+      "learning_rate": 2e-05,
+      "loss": 0.04388089,
+      "step": 8708
+    },
+    {
+      "epoch": 17.418,
+      "grad_norm": 1.5851181745529175,
+      "learning_rate": 2e-05,
+      "loss": 0.02392046,
+      "step": 8709
+    },
+    {
+      "epoch": 17.42,
+      "grad_norm": 0.9647265076637268,
+      "learning_rate": 2e-05,
+      "loss": 0.02776775,
+      "step": 8710
+    },
+    {
+      "epoch": 17.422,
+      "grad_norm": 1.6217008829116821,
+      "learning_rate": 2e-05,
+      "loss": 0.04731276,
+      "step": 8711
+    },
+    {
+      "epoch": 17.424,
+      "grad_norm": 0.8883227109909058,
+      "learning_rate": 2e-05,
+      "loss": 0.02925518,
+      "step": 8712
+    },
+    {
+      "epoch": 17.426,
+      "grad_norm": 1.7081272602081299,
+      "learning_rate": 2e-05,
+      "loss": 0.03178982,
+      "step": 8713
+    },
+    {
+      "epoch": 17.428,
+      "grad_norm": 1.0577318668365479,
+      "learning_rate": 2e-05,
+      "loss": 0.02738732,
+      "step": 8714
+    },
+    {
+      "epoch": 17.43,
+      "grad_norm": 1.5216280221939087,
+      "learning_rate": 2e-05,
+      "loss": 0.0358897,
+      "step": 8715
+    },
+    {
+      "epoch": 17.432,
+      "grad_norm": 1.733811855316162,
+      "learning_rate": 2e-05,
+      "loss": 0.05327031,
+      "step": 8716
+    },
+    {
+      "epoch": 17.434,
+      "grad_norm": 1.0504800081253052,
+      "learning_rate": 2e-05,
+      "loss": 0.02647152,
+      "step": 8717
+    },
+    {
+      "epoch": 17.436,
+      "grad_norm": 1.4641132354736328,
+      "learning_rate": 2e-05,
+      "loss": 0.03930076,
+      "step": 8718
+    },
+    {
+      "epoch": 17.438,
+      "grad_norm": 0.9525814652442932,
+      "learning_rate": 2e-05,
+      "loss": 0.03154223,
+      "step": 8719
+    },
+    {
+      "epoch": 17.44,
+      "grad_norm": 1.2682076692581177,
+      "learning_rate": 2e-05,
+      "loss": 0.04108154,
+      "step": 8720
+    },
+    {
+      "epoch": 17.442,
+      "grad_norm": 0.9335964918136597,
+      "learning_rate": 2e-05,
+      "loss": 0.0300378,
+      "step": 8721
+    },
+    {
+      "epoch": 17.444,
+      "grad_norm": 2.1530227661132812,
+      "learning_rate": 2e-05,
+      "loss": 0.0590164,
+      "step": 8722
+    },
+    {
+      "epoch": 17.446,
+      "grad_norm": 1.495179295539856,
+      "learning_rate": 2e-05,
+      "loss": 0.04352273,
+      "step": 8723
+    },
+    {
+      "epoch": 17.448,
+      "grad_norm": 1.4746037721633911,
+      "learning_rate": 2e-05,
+      "loss": 0.04175236,
+      "step": 8724
+    },
+    {
+      "epoch": 17.45,
+      "grad_norm": 1.263691782951355,
+      "learning_rate": 2e-05,
+      "loss": 0.03465231,
+      "step": 8725
+    },
+    {
+      "epoch": 17.452,
+      "grad_norm": 1.277445912361145,
+      "learning_rate": 2e-05,
+      "loss": 0.0429166,
+      "step": 8726
+    },
+    {
+      "epoch": 17.454,
+      "grad_norm": 0.9855414628982544,
+      "learning_rate": 2e-05,
+      "loss": 0.02733266,
+      "step": 8727
+    },
+    {
+      "epoch": 17.456,
+      "grad_norm": 1.5489836931228638,
+      "learning_rate": 2e-05,
+      "loss": 0.04620903,
+      "step": 8728
+    },
+    {
+      "epoch": 17.458,
+      "grad_norm": 1.0529900789260864,
+      "learning_rate": 2e-05,
+      "loss": 0.03364294,
+      "step": 8729
+    },
+    {
+      "epoch": 17.46,
+      "grad_norm": 1.1789546012878418,
+      "learning_rate": 2e-05,
+      "loss": 0.03275872,
+      "step": 8730
+    },
+    {
+      "epoch": 17.462,
+      "grad_norm": 0.9303401112556458,
+      "learning_rate": 2e-05,
+      "loss": 0.0350357,
+      "step": 8731
+    },
+    {
+      "epoch": 17.464,
+      "grad_norm": 0.8787309527397156,
+      "learning_rate": 2e-05,
+      "loss": 0.02384308,
+      "step": 8732
+    },
+    {
+      "epoch": 17.466,
+      "grad_norm": 1.6333367824554443,
+      "learning_rate": 2e-05,
+      "loss": 0.03150726,
+      "step": 8733
+    },
+    {
+      "epoch": 17.468,
+      "grad_norm": 1.1511269807815552,
+      "learning_rate": 2e-05,
+      "loss": 0.03846743,
+      "step": 8734
+    },
+    {
+      "epoch": 17.47,
+      "grad_norm": 1.0783816576004028,
+      "learning_rate": 2e-05,
+      "loss": 0.02969732,
+      "step": 8735
+    },
+    {
+      "epoch": 17.472,
+      "grad_norm": 1.126899003982544,
+      "learning_rate": 2e-05,
+      "loss": 0.03826762,
+      "step": 8736
+    },
+    {
+      "epoch": 17.474,
+      "grad_norm": 0.972001314163208,
+      "learning_rate": 2e-05,
+      "loss": 0.0226456,
+      "step": 8737
+    },
+    {
+      "epoch": 17.476,
+      "grad_norm": 2.0911567211151123,
+      "learning_rate": 2e-05,
+      "loss": 0.04673334,
+      "step": 8738
+    },
+    {
+      "epoch": 17.478,
+      "grad_norm": 1.2967848777770996,
+      "learning_rate": 2e-05,
+      "loss": 0.03325557,
+      "step": 8739
+    },
+    {
+      "epoch": 17.48,
+      "grad_norm": 1.4275823831558228,
+      "learning_rate": 2e-05,
+      "loss": 0.03418783,
+      "step": 8740
+    },
+    {
+      "epoch": 17.482,
+      "grad_norm": 1.6219322681427002,
+      "learning_rate": 2e-05,
+      "loss": 0.03391339,
+      "step": 8741
+    },
+    {
+      "epoch": 17.484,
+      "grad_norm": 2.5044708251953125,
+      "learning_rate": 2e-05,
+      "loss": 0.04077835,
+      "step": 8742
+    },
+    {
+      "epoch": 17.486,
+      "grad_norm": 1.286624789237976,
+      "learning_rate": 2e-05,
+      "loss": 0.03713045,
+      "step": 8743
+    },
+    {
+      "epoch": 17.488,
+      "grad_norm": 1.0048493146896362,
+      "learning_rate": 2e-05,
+      "loss": 0.02878037,
+      "step": 8744
+    },
+    {
+      "epoch": 17.49,
+      "grad_norm": 1.4934651851654053,
+      "learning_rate": 2e-05,
+      "loss": 0.03150274,
+      "step": 8745
+    },
+    {
+      "epoch": 17.492,
+      "grad_norm": 1.1032167673110962,
+      "learning_rate": 2e-05,
+      "loss": 0.03681361,
+      "step": 8746
+    },
+    {
+      "epoch": 17.494,
+      "grad_norm": 1.4902517795562744,
+      "learning_rate": 2e-05,
+      "loss": 0.04430451,
+      "step": 8747
+    },
+    {
+      "epoch": 17.496,
+      "grad_norm": 2.1936278343200684,
+      "learning_rate": 2e-05,
+      "loss": 0.05962869,
+      "step": 8748
+    },
+    {
+      "epoch": 17.498,
+      "grad_norm": 1.479569673538208,
+      "learning_rate": 2e-05,
+      "loss": 0.04059211,
+      "step": 8749
+    },
+    {
+      "epoch": 17.5,
+      "grad_norm": 1.139105200767517,
+      "learning_rate": 2e-05,
+      "loss": 0.03791492,
+      "step": 8750
+    },
+    {
+      "epoch": 17.502,
+      "grad_norm": 1.4118876457214355,
+      "learning_rate": 2e-05,
+      "loss": 0.03690355,
+      "step": 8751
+    },
+    {
+      "epoch": 17.504,
+      "grad_norm": 1.1906758546829224,
+      "learning_rate": 2e-05,
+      "loss": 0.03069144,
+      "step": 8752
+    },
+    {
+      "epoch": 17.506,
+      "grad_norm": 1.3194007873535156,
+      "learning_rate": 2e-05,
+      "loss": 0.04116385,
+      "step": 8753
+    },
+    {
+      "epoch": 17.508,
+      "grad_norm": 2.3798933029174805,
+      "learning_rate": 2e-05,
+      "loss": 0.03834019,
+      "step": 8754
+    },
+    {
+      "epoch": 17.51,
+      "grad_norm": 1.7592543363571167,
+      "learning_rate": 2e-05,
+      "loss": 0.05264074,
+      "step": 8755
+    },
+    {
+      "epoch": 17.512,
+      "grad_norm": 1.1170209646224976,
+      "learning_rate": 2e-05,
+      "loss": 0.03822346,
+      "step": 8756
+    },
+    {
+      "epoch": 17.514,
+      "grad_norm": 1.0373886823654175,
+      "learning_rate": 2e-05,
+      "loss": 0.03700188,
+      "step": 8757
+    },
+    {
+      "epoch": 17.516,
+      "grad_norm": 1.619546890258789,
+      "learning_rate": 2e-05,
+      "loss": 0.03848478,
+      "step": 8758
+    },
+    {
+      "epoch": 17.518,
+      "grad_norm": 0.9552158117294312,
+      "learning_rate": 2e-05,
+      "loss": 0.0324291,
+      "step": 8759
+    },
+    {
+      "epoch": 17.52,
+      "grad_norm": 1.3211631774902344,
+      "learning_rate": 2e-05,
+      "loss": 0.04355022,
+      "step": 8760
+    },
+    {
+      "epoch": 17.522,
+      "grad_norm": 1.0997872352600098,
+      "learning_rate": 2e-05,
+      "loss": 0.03464405,
+      "step": 8761
+    },
+    {
+      "epoch": 17.524,
+      "grad_norm": 1.546210527420044,
+      "learning_rate": 2e-05,
+      "loss": 0.03780903,
+      "step": 8762
+    },
+    {
+      "epoch": 17.526,
+      "grad_norm": 1.3221256732940674,
+      "learning_rate": 2e-05,
+      "loss": 0.03689417,
+      "step": 8763
+    },
+    {
+      "epoch": 17.528,
+      "grad_norm": 1.8055446147918701,
+      "learning_rate": 2e-05,
+      "loss": 0.04009835,
+      "step": 8764
+    },
+    {
+      "epoch": 17.53,
+      "grad_norm": 1.3256739377975464,
+      "learning_rate": 2e-05,
+      "loss": 0.02718241,
+      "step": 8765
+    },
+    {
+      "epoch": 17.532,
+      "grad_norm": 1.0728418827056885,
+      "learning_rate": 2e-05,
+      "loss": 0.02585372,
+      "step": 8766
+    },
+    {
+      "epoch": 17.534,
+      "grad_norm": 2.8010523319244385,
+      "learning_rate": 2e-05,
+      "loss": 0.03689571,
+      "step": 8767
+    },
+    {
+      "epoch": 17.536,
+      "grad_norm": 1.8129520416259766,
+      "learning_rate": 2e-05,
+      "loss": 0.04276496,
+      "step": 8768
+    },
+    {
+      "epoch": 17.538,
+      "grad_norm": 1.5692745447158813,
+      "learning_rate": 2e-05,
+      "loss": 0.03856734,
+      "step": 8769
+    },
+    {
+      "epoch": 17.54,
+      "grad_norm": 1.0610696077346802,
+      "learning_rate": 2e-05,
+      "loss": 0.02780731,
+      "step": 8770
+    },
+    {
+      "epoch": 17.542,
+      "grad_norm": 2.569103479385376,
+      "learning_rate": 2e-05,
+      "loss": 0.04206748,
+      "step": 8771
+    },
+    {
+      "epoch": 17.544,
+      "grad_norm": 2.11444354057312,
+      "learning_rate": 2e-05,
+      "loss": 0.03847713,
+      "step": 8772
+    },
+    {
+      "epoch": 17.546,
+      "grad_norm": 1.2605483531951904,
+      "learning_rate": 2e-05,
+      "loss": 0.03328384,
+      "step": 8773
+    },
+    {
+      "epoch": 17.548000000000002,
+      "grad_norm": 2.3000173568725586,
+      "learning_rate": 2e-05,
+      "loss": 0.0356028,
+      "step": 8774
+    },
+    {
+      "epoch": 17.55,
+      "grad_norm": 1.8785638809204102,
+      "learning_rate": 2e-05,
+      "loss": 0.04467007,
+      "step": 8775
+    },
+    {
+      "epoch": 17.552,
+      "grad_norm": 1.260977864265442,
+      "learning_rate": 2e-05,
+      "loss": 0.02562046,
+      "step": 8776
+    },
+    {
+      "epoch": 17.554,
+      "grad_norm": 9.660441398620605,
+      "learning_rate": 2e-05,
+      "loss": 0.04144814,
+      "step": 8777
+    },
+    {
+      "epoch": 17.556,
+      "grad_norm": 1.8776023387908936,
+      "learning_rate": 2e-05,
+      "loss": 0.04732294,
+      "step": 8778
+    },
+    {
+      "epoch": 17.558,
+      "grad_norm": 1.1368457078933716,
+      "learning_rate": 2e-05,
+      "loss": 0.04169676,
+      "step": 8779
+    },
+    {
+      "epoch": 17.56,
+      "grad_norm": 1.5295716524124146,
+      "learning_rate": 2e-05,
+      "loss": 0.05010715,
+      "step": 8780
+    },
+    {
+      "epoch": 17.562,
+      "grad_norm": 0.9344932436943054,
+      "learning_rate": 2e-05,
+      "loss": 0.02512026,
+      "step": 8781
+    },
+    {
+      "epoch": 17.564,
+      "grad_norm": 1.0945249795913696,
+      "learning_rate": 2e-05,
+      "loss": 0.03761927,
+      "step": 8782
+    },
+    {
+      "epoch": 17.566,
+      "grad_norm": 1.4906353950500488,
+      "learning_rate": 2e-05,
+      "loss": 0.04850148,
+      "step": 8783
+    },
+    {
+      "epoch": 17.568,
+      "grad_norm": 1.0239425897598267,
+      "learning_rate": 2e-05,
+      "loss": 0.0349384,
+      "step": 8784
+    },
+    {
+      "epoch": 17.57,
+      "grad_norm": 1.3439842462539673,
+      "learning_rate": 2e-05,
+      "loss": 0.02740002,
+      "step": 8785
+    },
+    {
+      "epoch": 17.572,
+      "grad_norm": 1.8573307991027832,
+      "learning_rate": 2e-05,
+      "loss": 0.03505997,
+      "step": 8786
+    },
+    {
+      "epoch": 17.574,
+      "grad_norm": 1.1892509460449219,
+      "learning_rate": 2e-05,
+      "loss": 0.03444887,
+      "step": 8787
+    },
+    {
+      "epoch": 17.576,
+      "grad_norm": 1.6804590225219727,
+      "learning_rate": 2e-05,
+      "loss": 0.03725114,
+      "step": 8788
+    },
+    {
+      "epoch": 17.578,
+      "grad_norm": 1.5094798803329468,
+      "learning_rate": 2e-05,
+      "loss": 0.03691319,
+      "step": 8789
+    },
+    {
+      "epoch": 17.58,
+      "grad_norm": 1.433280110359192,
+      "learning_rate": 2e-05,
+      "loss": 0.03601858,
+      "step": 8790
+    },
+    {
+      "epoch": 17.582,
+      "grad_norm": 0.9142979383468628,
+      "learning_rate": 2e-05,
+      "loss": 0.03045445,
+      "step": 8791
+    },
+    {
+      "epoch": 17.584,
+      "grad_norm": 0.9239221811294556,
+      "learning_rate": 2e-05,
+      "loss": 0.02360635,
+      "step": 8792
+    },
+    {
+      "epoch": 17.586,
+      "grad_norm": 2.742828369140625,
+      "learning_rate": 2e-05,
+      "loss": 0.04096455,
+      "step": 8793
+    },
+    {
+      "epoch": 17.588,
+      "grad_norm": 1.469237208366394,
+      "learning_rate": 2e-05,
+      "loss": 0.03894402,
+      "step": 8794
+    },
+    {
+      "epoch": 17.59,
+      "grad_norm": 1.224602222442627,
+      "learning_rate": 2e-05,
+      "loss": 0.03480183,
+      "step": 8795
+    },
+    {
+      "epoch": 17.592,
+      "grad_norm": 2.1132779121398926,
+      "learning_rate": 2e-05,
+      "loss": 0.04112736,
+      "step": 8796
+    },
+    {
+      "epoch": 17.594,
+      "grad_norm": 1.4963428974151611,
+      "learning_rate": 2e-05,
+      "loss": 0.03678083,
+      "step": 8797
+    },
+    {
+      "epoch": 17.596,
+      "grad_norm": 1.1321431398391724,
+      "learning_rate": 2e-05,
+      "loss": 0.02224684,
+      "step": 8798
+    },
+    {
+      "epoch": 17.598,
+      "grad_norm": 1.7322660684585571,
+      "learning_rate": 2e-05,
+      "loss": 0.04057566,
+      "step": 8799
+    },
+    {
+      "epoch": 17.6,
+      "grad_norm": 1.437656283378601,
+      "learning_rate": 2e-05,
+      "loss": 0.03213751,
+      "step": 8800
+    },
+    {
+      "epoch": 17.602,
+      "grad_norm": 1.932984471321106,
+      "learning_rate": 2e-05,
+      "loss": 0.04008548,
+      "step": 8801
+    },
+    {
+      "epoch": 17.604,
+      "grad_norm": 1.0793958902359009,
+      "learning_rate": 2e-05,
+      "loss": 0.02873497,
+      "step": 8802
+    },
+    {
+      "epoch": 17.606,
+      "grad_norm": 1.9617685079574585,
+      "learning_rate": 2e-05,
+      "loss": 0.04457264,
+      "step": 8803
+    },
+    {
+      "epoch": 17.608,
+      "grad_norm": 1.7142366170883179,
+      "learning_rate": 2e-05,
+      "loss": 0.03925028,
+      "step": 8804
+    },
+    {
+      "epoch": 17.61,
+      "grad_norm": 1.638058066368103,
+      "learning_rate": 2e-05,
+      "loss": 0.04914237,
+      "step": 8805
+    },
+    {
+      "epoch": 17.612,
+      "grad_norm": 1.3433510065078735,
+      "learning_rate": 2e-05,
+      "loss": 0.03332677,
+      "step": 8806
+    },
+    {
+      "epoch": 17.614,
+      "grad_norm": 0.9329344034194946,
+      "learning_rate": 2e-05,
+      "loss": 0.03115493,
+      "step": 8807
+    },
+    {
+      "epoch": 17.616,
+      "grad_norm": 0.9696981310844421,
+      "learning_rate": 2e-05,
+      "loss": 0.0317462,
+      "step": 8808
+    },
+    {
+      "epoch": 17.618,
+      "grad_norm": 2.5762224197387695,
+      "learning_rate": 2e-05,
+      "loss": 0.04256003,
+      "step": 8809
+    },
+    {
+      "epoch": 17.62,
+      "grad_norm": 1.8832440376281738,
+      "learning_rate": 2e-05,
+      "loss": 0.03043187,
+      "step": 8810
+    },
+    {
+      "epoch": 17.622,
+      "grad_norm": 2.1648950576782227,
+      "learning_rate": 2e-05,
+      "loss": 0.04386858,
+      "step": 8811
+    },
+    {
+      "epoch": 17.624,
+      "grad_norm": 1.578770637512207,
+      "learning_rate": 2e-05,
+      "loss": 0.03511237,
+      "step": 8812
+    },
+    {
+      "epoch": 17.626,
+      "grad_norm": 1.0861693620681763,
+      "learning_rate": 2e-05,
+      "loss": 0.04000392,
+      "step": 8813
+    },
+    {
+      "epoch": 17.628,
+      "grad_norm": 1.1005021333694458,
+      "learning_rate": 2e-05,
+      "loss": 0.03653765,
+      "step": 8814
+    },
+    {
+      "epoch": 17.63,
+      "grad_norm": 1.5705111026763916,
+      "learning_rate": 2e-05,
+      "loss": 0.04100737,
+      "step": 8815
+    },
+    {
+      "epoch": 17.632,
+      "grad_norm": 0.9939731359481812,
+      "learning_rate": 2e-05,
+      "loss": 0.02468416,
+      "step": 8816
+    },
+    {
+      "epoch": 17.634,
+      "grad_norm": 1.6492228507995605,
+      "learning_rate": 2e-05,
+      "loss": 0.03122869,
+      "step": 8817
+    },
+    {
+      "epoch": 17.636,
+      "grad_norm": 1.1242082118988037,
+      "learning_rate": 2e-05,
+      "loss": 0.03778775,
+      "step": 8818
+    },
+    {
+      "epoch": 17.638,
+      "grad_norm": 1.0076687335968018,
+      "learning_rate": 2e-05,
+      "loss": 0.03823631,
+      "step": 8819
+    },
+    {
+      "epoch": 17.64,
+      "grad_norm": 1.2714289426803589,
+      "learning_rate": 2e-05,
+      "loss": 0.03733125,
+      "step": 8820
+    },
+    {
+      "epoch": 17.642,
+      "grad_norm": 1.4574763774871826,
+      "learning_rate": 2e-05,
+      "loss": 0.04100376,
+      "step": 8821
+    },
+    {
+      "epoch": 17.644,
+      "grad_norm": 1.4745090007781982,
+      "learning_rate": 2e-05,
+      "loss": 0.04590669,
+      "step": 8822
+    },
+    {
+      "epoch": 17.646,
+      "grad_norm": 1.2290929555892944,
+      "learning_rate": 2e-05,
+      "loss": 0.02955338,
+      "step": 8823
+    },
+    {
+      "epoch": 17.648,
+      "grad_norm": 1.5989232063293457,
+      "learning_rate": 2e-05,
+      "loss": 0.04057942,
+      "step": 8824
+    },
+    {
+      "epoch": 17.65,
+      "grad_norm": 1.0911918878555298,
+      "learning_rate": 2e-05,
+      "loss": 0.0314585,
+      "step": 8825
+    },
+    {
+      "epoch": 17.652,
+      "grad_norm": 1.1875574588775635,
+      "learning_rate": 2e-05,
+      "loss": 0.02796382,
+      "step": 8826
+    },
+    {
+      "epoch": 17.654,
+      "grad_norm": 0.8497743010520935,
+      "learning_rate": 2e-05,
+      "loss": 0.02476349,
+      "step": 8827
+    },
+    {
+      "epoch": 17.656,
+      "grad_norm": 1.1606796979904175,
+      "learning_rate": 2e-05,
+      "loss": 0.03331729,
+      "step": 8828
+    },
+    {
+      "epoch": 17.658,
+      "grad_norm": 1.929567575454712,
+      "learning_rate": 2e-05,
+      "loss": 0.04759387,
+      "step": 8829
+    },
+    {
+      "epoch": 17.66,
+      "grad_norm": 1.10890531539917,
+      "learning_rate": 2e-05,
+      "loss": 0.03875232,
+      "step": 8830
+    },
+    {
+      "epoch": 17.662,
+      "grad_norm": 1.1237422227859497,
+      "learning_rate": 2e-05,
+      "loss": 0.02639496,
+      "step": 8831
+    },
+    {
+      "epoch": 17.664,
+      "grad_norm": 1.3612616062164307,
+      "learning_rate": 2e-05,
+      "loss": 0.04055498,
+      "step": 8832
+    },
+    {
+      "epoch": 17.666,
+      "grad_norm": 1.0668309926986694,
+      "learning_rate": 2e-05,
+      "loss": 0.02782569,
+      "step": 8833
+    },
+    {
+      "epoch": 17.668,
+      "grad_norm": 1.186408519744873,
+      "learning_rate": 2e-05,
+      "loss": 0.03598,
+      "step": 8834
+    },
+    {
+      "epoch": 17.67,
+      "grad_norm": 2.71864914894104,
+      "learning_rate": 2e-05,
+      "loss": 0.05295214,
+      "step": 8835
+    },
+    {
+      "epoch": 17.672,
+      "grad_norm": 2.222414016723633,
+      "learning_rate": 2e-05,
+      "loss": 0.04588395,
+      "step": 8836
+    },
+    {
+      "epoch": 17.674,
+      "grad_norm": 1.272736668586731,
+      "learning_rate": 2e-05,
+      "loss": 0.03839596,
+      "step": 8837
+    },
+    {
+      "epoch": 17.676,
+      "grad_norm": 1.5483832359313965,
+      "learning_rate": 2e-05,
+      "loss": 0.0436643,
+      "step": 8838
+    },
+    {
+      "epoch": 17.678,
+      "grad_norm": 3.551170587539673,
+      "learning_rate": 2e-05,
+      "loss": 0.03059383,
+      "step": 8839
+    },
+    {
+      "epoch": 17.68,
+      "grad_norm": 1.3416359424591064,
+      "learning_rate": 2e-05,
+      "loss": 0.03237988,
+      "step": 8840
+    },
+    {
+      "epoch": 17.682,
+      "grad_norm": 2.128422975540161,
+      "learning_rate": 2e-05,
+      "loss": 0.04019484,
+      "step": 8841
+    },
+    {
+      "epoch": 17.684,
+      "grad_norm": 1.4956728219985962,
+      "learning_rate": 2e-05,
+      "loss": 0.03178658,
+      "step": 8842
+    },
+    {
+      "epoch": 17.686,
+      "grad_norm": 1.644662618637085,
+      "learning_rate": 2e-05,
+      "loss": 0.05027148,
+      "step": 8843
+    },
+    {
+      "epoch": 17.688,
+      "grad_norm": 3.315823554992676,
+      "learning_rate": 2e-05,
+      "loss": 0.03754386,
+      "step": 8844
+    },
+    {
+      "epoch": 17.69,
+      "grad_norm": 2.4120471477508545,
+      "learning_rate": 2e-05,
+      "loss": 0.03160707,
+      "step": 8845
+    },
+    {
+      "epoch": 17.692,
+      "grad_norm": 3.103656053543091,
+      "learning_rate": 2e-05,
+      "loss": 0.04504135,
+      "step": 8846
+    },
+    {
+      "epoch": 17.694,
+      "grad_norm": 1.1503428220748901,
+      "learning_rate": 2e-05,
+      "loss": 0.03259352,
+      "step": 8847
+    },
+    {
+      "epoch": 17.696,
+      "grad_norm": 1.2756330966949463,
+      "learning_rate": 2e-05,
+      "loss": 0.03456133,
+      "step": 8848
+    },
+    {
+      "epoch": 17.698,
+      "grad_norm": 1.1104017496109009,
+      "learning_rate": 2e-05,
+      "loss": 0.02761788,
+      "step": 8849
+    },
+    {
+      "epoch": 17.7,
+      "grad_norm": 0.9424893856048584,
+      "learning_rate": 2e-05,
+      "loss": 0.02640409,
+      "step": 8850
+    },
+    {
+      "epoch": 17.701999999999998,
+      "grad_norm": 1.0416557788848877,
+      "learning_rate": 2e-05,
+      "loss": 0.0349526,
+      "step": 8851
+    },
+    {
+      "epoch": 17.704,
+      "grad_norm": 1.6049892902374268,
+      "learning_rate": 2e-05,
+      "loss": 0.02749394,
+      "step": 8852
+    },
+    {
+      "epoch": 17.706,
+      "grad_norm": 1.6006180047988892,
+      "learning_rate": 2e-05,
+      "loss": 0.03950756,
+      "step": 8853
+    },
+    {
+      "epoch": 17.708,
+      "grad_norm": 1.0300740003585815,
+      "learning_rate": 2e-05,
+      "loss": 0.03133941,
+      "step": 8854
+    },
+    {
+      "epoch": 17.71,
+      "grad_norm": 3.491128444671631,
+      "learning_rate": 2e-05,
+      "loss": 0.04869714,
+      "step": 8855
+    },
+    {
+      "epoch": 17.712,
+      "grad_norm": 1.389901041984558,
+      "learning_rate": 2e-05,
+      "loss": 0.04399059,
+      "step": 8856
+    },
+    {
+      "epoch": 17.714,
+      "grad_norm": 1.09212327003479,
+      "learning_rate": 2e-05,
+      "loss": 0.03761019,
+      "step": 8857
+    },
+    {
+      "epoch": 17.716,
+      "grad_norm": 1.3363914489746094,
+      "learning_rate": 2e-05,
+      "loss": 0.0412604,
+      "step": 8858
+    },
+    {
+      "epoch": 17.718,
+      "grad_norm": 1.353605031967163,
+      "learning_rate": 2e-05,
+      "loss": 0.04291609,
+      "step": 8859
+    },
+    {
+      "epoch": 17.72,
+      "grad_norm": 1.0849765539169312,
+      "learning_rate": 2e-05,
+      "loss": 0.03004822,
+      "step": 8860
+    },
+    {
+      "epoch": 17.722,
+      "grad_norm": 1.2996422052383423,
+      "learning_rate": 2e-05,
+      "loss": 0.04499619,
+      "step": 8861
+    },
+    {
+      "epoch": 17.724,
+      "grad_norm": 1.1636970043182373,
+      "learning_rate": 2e-05,
+      "loss": 0.03615176,
+      "step": 8862
+    },
+    {
+      "epoch": 17.726,
+      "grad_norm": 0.9753794074058533,
+      "learning_rate": 2e-05,
+      "loss": 0.03914835,
+      "step": 8863
+    },
+    {
+      "epoch": 17.728,
+      "grad_norm": 1.3837984800338745,
+      "learning_rate": 2e-05,
+      "loss": 0.05555439,
+      "step": 8864
+    },
+    {
+      "epoch": 17.73,
+      "grad_norm": 1.9436976909637451,
+      "learning_rate": 2e-05,
+      "loss": 0.05068126,
+      "step": 8865
+    },
+    {
+      "epoch": 17.732,
+      "grad_norm": 1.1782814264297485,
+      "learning_rate": 2e-05,
+      "loss": 0.03269964,
+      "step": 8866
+    },
+    {
+      "epoch": 17.734,
+      "grad_norm": 1.342545747756958,
+      "learning_rate": 2e-05,
+      "loss": 0.051434,
+      "step": 8867
+    },
+    {
+      "epoch": 17.736,
+      "grad_norm": 1.1777751445770264,
+      "learning_rate": 2e-05,
+      "loss": 0.03173701,
+      "step": 8868
+    },
+    {
+      "epoch": 17.738,
+      "grad_norm": 1.666817307472229,
+      "learning_rate": 2e-05,
+      "loss": 0.03970077,
+      "step": 8869
+    },
+    {
+      "epoch": 17.74,
+      "grad_norm": 1.2977221012115479,
+      "learning_rate": 2e-05,
+      "loss": 0.04394735,
+      "step": 8870
+    },
+    {
+      "epoch": 17.742,
+      "grad_norm": 1.076141357421875,
+      "learning_rate": 2e-05,
+      "loss": 0.03446401,
+      "step": 8871
+    },
+    {
+      "epoch": 17.744,
+      "grad_norm": 1.037126898765564,
+      "learning_rate": 2e-05,
+      "loss": 0.03374257,
+      "step": 8872
+    },
+    {
+      "epoch": 17.746,
+      "grad_norm": 1.9888211488723755,
+      "learning_rate": 2e-05,
+      "loss": 0.05602897,
+      "step": 8873
+    },
+    {
+      "epoch": 17.748,
+      "grad_norm": 1.2631540298461914,
+      "learning_rate": 2e-05,
+      "loss": 0.03305461,
+      "step": 8874
+    },
+    {
+      "epoch": 17.75,
+      "grad_norm": 1.383660078048706,
+      "learning_rate": 2e-05,
+      "loss": 0.03023957,
+      "step": 8875
+    },
+    {
+      "epoch": 17.752,
+      "grad_norm": 1.6274633407592773,
+      "learning_rate": 2e-05,
+      "loss": 0.02720205,
+      "step": 8876
+    },
+    {
+      "epoch": 17.754,
+      "grad_norm": 2.328664541244507,
+      "learning_rate": 2e-05,
+      "loss": 0.04531138,
+      "step": 8877
+    },
+    {
+      "epoch": 17.756,
+      "grad_norm": 1.0562291145324707,
+      "learning_rate": 2e-05,
+      "loss": 0.02581457,
+      "step": 8878
+    },
+    {
+      "epoch": 17.758,
+      "grad_norm": 1.2635678052902222,
+      "learning_rate": 2e-05,
+      "loss": 0.03207887,
+      "step": 8879
+    },
+    {
+      "epoch": 17.76,
+      "grad_norm": 1.1033519506454468,
+      "learning_rate": 2e-05,
+      "loss": 0.03914376,
+      "step": 8880
+    },
+    {
+      "epoch": 17.762,
+      "grad_norm": 2.1870508193969727,
+      "learning_rate": 2e-05,
+      "loss": 0.04750444,
+      "step": 8881
+    },
+    {
+      "epoch": 17.764,
+      "grad_norm": 1.3173456192016602,
+      "learning_rate": 2e-05,
+      "loss": 0.04112824,
+      "step": 8882
+    },
+    {
+      "epoch": 17.766,
+      "grad_norm": 1.5373280048370361,
+      "learning_rate": 2e-05,
+      "loss": 0.03573046,
+      "step": 8883
+    },
+    {
+      "epoch": 17.768,
+      "grad_norm": 1.249330759048462,
+      "learning_rate": 2e-05,
+      "loss": 0.03755806,
+      "step": 8884
+    },
+    {
+      "epoch": 17.77,
+      "grad_norm": 1.2358646392822266,
+      "learning_rate": 2e-05,
+      "loss": 0.02392829,
+      "step": 8885
+    },
+    {
+      "epoch": 17.772,
+      "grad_norm": 0.6719095706939697,
+      "learning_rate": 2e-05,
+      "loss": 0.01649454,
+      "step": 8886
+    },
+    {
+      "epoch": 17.774,
+      "grad_norm": 1.3092347383499146,
+      "learning_rate": 2e-05,
+      "loss": 0.04142372,
+      "step": 8887
+    },
+    {
+      "epoch": 17.776,
+      "grad_norm": 1.4076433181762695,
+      "learning_rate": 2e-05,
+      "loss": 0.03731369,
+      "step": 8888
+    },
+    {
+      "epoch": 17.778,
+      "grad_norm": 1.1862058639526367,
+      "learning_rate": 2e-05,
+      "loss": 0.02469937,
+      "step": 8889
+    },
+    {
+      "epoch": 17.78,
+      "grad_norm": 1.0884705781936646,
+      "learning_rate": 2e-05,
+      "loss": 0.03976347,
+      "step": 8890
+    },
+    {
+      "epoch": 17.782,
+      "grad_norm": 1.3288317918777466,
+      "learning_rate": 2e-05,
+      "loss": 0.03909069,
+      "step": 8891
+    },
+    {
+      "epoch": 17.784,
+      "grad_norm": 1.437835693359375,
+      "learning_rate": 2e-05,
+      "loss": 0.04604793,
+      "step": 8892
+    },
+    {
+      "epoch": 17.786,
+      "grad_norm": 1.2429791688919067,
+      "learning_rate": 2e-05,
+      "loss": 0.03789811,
+      "step": 8893
+    },
+    {
+      "epoch": 17.788,
+      "grad_norm": 0.8929753303527832,
+      "learning_rate": 2e-05,
+      "loss": 0.02409481,
+      "step": 8894
+    },
+    {
+      "epoch": 17.79,
+      "grad_norm": 1.292466163635254,
+      "learning_rate": 2e-05,
+      "loss": 0.04581561,
+      "step": 8895
+    },
+    {
+      "epoch": 17.792,
+      "grad_norm": 1.1110408306121826,
+      "learning_rate": 2e-05,
+      "loss": 0.03448091,
+      "step": 8896
+    },
+    {
+      "epoch": 17.794,
+      "grad_norm": 1.2385011911392212,
+      "learning_rate": 2e-05,
+      "loss": 0.03106187,
+      "step": 8897
+    },
+    {
+      "epoch": 17.796,
+      "grad_norm": 2.670729637145996,
+      "learning_rate": 2e-05,
+      "loss": 0.04714116,
+      "step": 8898
+    },
+    {
+      "epoch": 17.798000000000002,
+      "grad_norm": 0.9738274216651917,
+      "learning_rate": 2e-05,
+      "loss": 0.02809829,
+      "step": 8899
+    },
+    {
+      "epoch": 17.8,
+      "grad_norm": 1.753575325012207,
+      "learning_rate": 2e-05,
+      "loss": 0.04472911,
+      "step": 8900
+    },
+    {
+      "epoch": 17.802,
+      "grad_norm": 1.6126110553741455,
+      "learning_rate": 2e-05,
+      "loss": 0.0548849,
+      "step": 8901
+    },
+    {
+      "epoch": 17.804,
+      "grad_norm": 1.9280973672866821,
+      "learning_rate": 2e-05,
+      "loss": 0.0414143,
+      "step": 8902
+    },
+    {
+      "epoch": 17.806,
+      "grad_norm": 0.8208984136581421,
+      "learning_rate": 2e-05,
+      "loss": 0.01933501,
+      "step": 8903
+    },
+    {
+      "epoch": 17.808,
+      "grad_norm": 2.380815029144287,
+      "learning_rate": 2e-05,
+      "loss": 0.03992394,
+      "step": 8904
+    },
+    {
+      "epoch": 17.81,
+      "grad_norm": 2.3603808879852295,
+      "learning_rate": 2e-05,
+      "loss": 0.06256126,
+      "step": 8905
+    },
+    {
+      "epoch": 17.812,
+      "grad_norm": 1.170880675315857,
+      "learning_rate": 2e-05,
+      "loss": 0.0337522,
+      "step": 8906
+    },
+    {
+      "epoch": 17.814,
+      "grad_norm": 1.7238221168518066,
+      "learning_rate": 2e-05,
+      "loss": 0.0442452,
+      "step": 8907
+    },
+    {
+      "epoch": 17.816,
+      "grad_norm": 1.3815233707427979,
+      "learning_rate": 2e-05,
+      "loss": 0.03570531,
+      "step": 8908
+    },
+    {
+      "epoch": 17.818,
+      "grad_norm": 1.077028751373291,
+      "learning_rate": 2e-05,
+      "loss": 0.02792567,
+      "step": 8909
+    },
+    {
+      "epoch": 17.82,
+      "grad_norm": 1.8009798526763916,
+      "learning_rate": 2e-05,
+      "loss": 0.03502089,
+      "step": 8910
+    },
+    {
+      "epoch": 17.822,
+      "grad_norm": 1.4753614664077759,
+      "learning_rate": 2e-05,
+      "loss": 0.04613937,
+      "step": 8911
+    },
+    {
+      "epoch": 17.824,
+      "grad_norm": 1.8474282026290894,
+      "learning_rate": 2e-05,
+      "loss": 0.05605477,
+      "step": 8912
+    },
+    {
+      "epoch": 17.826,
+      "grad_norm": 1.3313781023025513,
+      "learning_rate": 2e-05,
+      "loss": 0.04386368,
+      "step": 8913
+    },
+    {
+      "epoch": 17.828,
+      "grad_norm": 1.3753377199172974,
+      "learning_rate": 2e-05,
+      "loss": 0.0385115,
+      "step": 8914
+    },
+    {
+      "epoch": 17.83,
+      "grad_norm": 0.8378182053565979,
+      "learning_rate": 2e-05,
+      "loss": 0.02854342,
+      "step": 8915
+    },
+    {
+      "epoch": 17.832,
+      "grad_norm": 1.0748872756958008,
+      "learning_rate": 2e-05,
+      "loss": 0.02578593,
+      "step": 8916
+    },
+    {
+      "epoch": 17.834,
+      "grad_norm": 0.881872832775116,
+      "learning_rate": 2e-05,
+      "loss": 0.02822345,
+      "step": 8917
+    },
+    {
+      "epoch": 17.836,
+      "grad_norm": 2.561361074447632,
+      "learning_rate": 2e-05,
+      "loss": 0.04112708,
+      "step": 8918
+    },
+    {
+      "epoch": 17.838,
+      "grad_norm": 1.0299837589263916,
+      "learning_rate": 2e-05,
+      "loss": 0.030092,
+      "step": 8919
+    },
+    {
+      "epoch": 17.84,
+      "grad_norm": 1.040921688079834,
+      "learning_rate": 2e-05,
+      "loss": 0.02578293,
+      "step": 8920
+    },
+    {
+      "epoch": 17.842,
+      "grad_norm": 1.1591360569000244,
+      "learning_rate": 2e-05,
+      "loss": 0.04461095,
+      "step": 8921
+    },
+    {
+      "epoch": 17.844,
+      "grad_norm": 0.929568886756897,
+      "learning_rate": 2e-05,
+      "loss": 0.03212486,
+      "step": 8922
+    },
+    {
+      "epoch": 17.846,
+      "grad_norm": 1.0860724449157715,
+      "learning_rate": 2e-05,
+      "loss": 0.02192564,
+      "step": 8923
+    },
+    {
+      "epoch": 17.848,
+      "grad_norm": 1.2266968488693237,
+      "learning_rate": 2e-05,
+      "loss": 0.0363798,
+      "step": 8924
+    },
+    {
+      "epoch": 17.85,
+      "grad_norm": 1.622202754020691,
+      "learning_rate": 2e-05,
+      "loss": 0.03073082,
+      "step": 8925
+    },
+    {
+      "epoch": 17.852,
+      "grad_norm": 1.0483282804489136,
+      "learning_rate": 2e-05,
+      "loss": 0.03715392,
+      "step": 8926
+    },
+    {
+      "epoch": 17.854,
+      "grad_norm": 1.1092106103897095,
+      "learning_rate": 2e-05,
+      "loss": 0.03876258,
+      "step": 8927
+    },
+    {
+      "epoch": 17.856,
+      "grad_norm": 0.9673099517822266,
+      "learning_rate": 2e-05,
+      "loss": 0.02583649,
+      "step": 8928
+    },
+    {
+      "epoch": 17.858,
+      "grad_norm": 1.635252833366394,
+      "learning_rate": 2e-05,
+      "loss": 0.05052019,
+      "step": 8929
+    },
+    {
+      "epoch": 17.86,
+      "grad_norm": 1.153358817100525,
+      "learning_rate": 2e-05,
+      "loss": 0.03466346,
+      "step": 8930
+    },
+    {
+      "epoch": 17.862,
+      "grad_norm": 2.039275884628296,
+      "learning_rate": 2e-05,
+      "loss": 0.05241006,
+      "step": 8931
+    },
+    {
+      "epoch": 17.864,
+      "grad_norm": 1.3057268857955933,
+      "learning_rate": 2e-05,
+      "loss": 0.04647853,
+      "step": 8932
+    },
+    {
+      "epoch": 17.866,
+      "grad_norm": 1.776149868965149,
+      "learning_rate": 2e-05,
+      "loss": 0.03174097,
+      "step": 8933
+    },
+    {
+      "epoch": 17.868,
+      "grad_norm": 1.7656913995742798,
+      "learning_rate": 2e-05,
+      "loss": 0.03435258,
+      "step": 8934
+    },
+    {
+      "epoch": 17.87,
+      "grad_norm": 1.3331656455993652,
+      "learning_rate": 2e-05,
+      "loss": 0.04026199,
+      "step": 8935
+    },
+    {
+      "epoch": 17.872,
+      "grad_norm": 1.0311089754104614,
+      "learning_rate": 2e-05,
+      "loss": 0.02966706,
+      "step": 8936
+    },
+    {
+      "epoch": 17.874,
+      "grad_norm": 1.0116089582443237,
+      "learning_rate": 2e-05,
+      "loss": 0.03720299,
+      "step": 8937
+    },
+    {
+      "epoch": 17.876,
+      "grad_norm": 1.1523536443710327,
+      "learning_rate": 2e-05,
+      "loss": 0.0370546,
+      "step": 8938
+    },
+    {
+      "epoch": 17.878,
+      "grad_norm": 1.1194231510162354,
+      "learning_rate": 2e-05,
+      "loss": 0.03367734,
+      "step": 8939
+    },
+    {
+      "epoch": 17.88,
+      "grad_norm": 1.4645055532455444,
+      "learning_rate": 2e-05,
+      "loss": 0.04109664,
+      "step": 8940
+    },
+    {
+      "epoch": 17.882,
+      "grad_norm": 1.3934130668640137,
+      "learning_rate": 2e-05,
+      "loss": 0.04116001,
+      "step": 8941
+    },
+    {
+      "epoch": 17.884,
+      "grad_norm": 1.2432390451431274,
+      "learning_rate": 2e-05,
+      "loss": 0.0318905,
+      "step": 8942
+    },
+    {
+      "epoch": 17.886,
+      "grad_norm": 1.9031294584274292,
+      "learning_rate": 2e-05,
+      "loss": 0.04899106,
+      "step": 8943
+    },
+    {
+      "epoch": 17.888,
+      "grad_norm": 1.3559234142303467,
+      "learning_rate": 2e-05,
+      "loss": 0.04302301,
+      "step": 8944
+    },
+    {
+      "epoch": 17.89,
+      "grad_norm": 1.1823681592941284,
+      "learning_rate": 2e-05,
+      "loss": 0.03750154,
+      "step": 8945
+    },
+    {
+      "epoch": 17.892,
+      "grad_norm": 2.3975961208343506,
+      "learning_rate": 2e-05,
+      "loss": 0.04103041,
+      "step": 8946
+    },
+    {
+      "epoch": 17.894,
+      "grad_norm": 1.2826277017593384,
+      "learning_rate": 2e-05,
+      "loss": 0.03688804,
+      "step": 8947
+    },
+    {
+      "epoch": 17.896,
+      "grad_norm": 2.822355270385742,
+      "learning_rate": 2e-05,
+      "loss": 0.05060627,
+      "step": 8948
+    },
+    {
+      "epoch": 17.898,
+      "grad_norm": 1.0254136323928833,
+      "learning_rate": 2e-05,
+      "loss": 0.03011078,
+      "step": 8949
+    },
+    {
+      "epoch": 17.9,
+      "grad_norm": 1.1992510557174683,
+      "learning_rate": 2e-05,
+      "loss": 0.04118895,
+      "step": 8950
+    },
+    {
+      "epoch": 17.902,
+      "grad_norm": 1.697759747505188,
+      "learning_rate": 2e-05,
+      "loss": 0.04727785,
+      "step": 8951
+    },
+    {
+      "epoch": 17.904,
+      "grad_norm": 1.5532244443893433,
+      "learning_rate": 2e-05,
+      "loss": 0.05363618,
+      "step": 8952
+    },
+    {
+      "epoch": 17.906,
+      "grad_norm": 1.0093700885772705,
+      "learning_rate": 2e-05,
+      "loss": 0.03118528,
+      "step": 8953
+    },
+    {
+      "epoch": 17.908,
+      "grad_norm": 1.895651936531067,
+      "learning_rate": 2e-05,
+      "loss": 0.06020194,
+      "step": 8954
+    },
+    {
+      "epoch": 17.91,
+      "grad_norm": 1.0814530849456787,
+      "learning_rate": 2e-05,
+      "loss": 0.0355964,
+      "step": 8955
+    },
+    {
+      "epoch": 17.912,
+      "grad_norm": 1.521867275238037,
+      "learning_rate": 2e-05,
+      "loss": 0.04056902,
+      "step": 8956
+    },
+    {
+      "epoch": 17.914,
+      "grad_norm": 1.2673386335372925,
+      "learning_rate": 2e-05,
+      "loss": 0.03534425,
+      "step": 8957
+    },
+    {
+      "epoch": 17.916,
+      "grad_norm": 2.1670548915863037,
+      "learning_rate": 2e-05,
+      "loss": 0.03837797,
+      "step": 8958
+    },
+    {
+      "epoch": 17.918,
+      "grad_norm": 0.9772452116012573,
+      "learning_rate": 2e-05,
+      "loss": 0.02949987,
+      "step": 8959
+    },
+    {
+      "epoch": 17.92,
+      "grad_norm": 1.940874695777893,
+      "learning_rate": 2e-05,
+      "loss": 0.06990305,
+      "step": 8960
+    },
+    {
+      "epoch": 17.922,
+      "grad_norm": 2.148874044418335,
+      "learning_rate": 2e-05,
+      "loss": 0.03925563,
+      "step": 8961
+    },
+    {
+      "epoch": 17.924,
+      "grad_norm": 1.2279850244522095,
+      "learning_rate": 2e-05,
+      "loss": 0.03267869,
+      "step": 8962
+    },
+    {
+      "epoch": 17.926,
+      "grad_norm": 1.2713209390640259,
+      "learning_rate": 2e-05,
+      "loss": 0.03438141,
+      "step": 8963
+    },
+    {
+      "epoch": 17.928,
+      "grad_norm": 1.0285924673080444,
+      "learning_rate": 2e-05,
+      "loss": 0.03503043,
+      "step": 8964
+    },
+    {
+      "epoch": 17.93,
+      "grad_norm": 1.2333204746246338,
+      "learning_rate": 2e-05,
+      "loss": 0.03405837,
+      "step": 8965
+    },
+    {
+      "epoch": 17.932,
+      "grad_norm": 2.2821717262268066,
+      "learning_rate": 2e-05,
+      "loss": 0.03725312,
+      "step": 8966
+    },
+    {
+      "epoch": 17.934,
+      "grad_norm": 3.1810193061828613,
+      "learning_rate": 2e-05,
+      "loss": 0.0413457,
+      "step": 8967
+    },
+    {
+      "epoch": 17.936,
+      "grad_norm": 1.8636484146118164,
+      "learning_rate": 2e-05,
+      "loss": 0.0562732,
+      "step": 8968
+    },
+    {
+      "epoch": 17.938,
+      "grad_norm": 2.1253907680511475,
+      "learning_rate": 2e-05,
+      "loss": 0.04173985,
+      "step": 8969
+    },
+    {
+      "epoch": 17.94,
+      "grad_norm": 1.4649451971054077,
+      "learning_rate": 2e-05,
+      "loss": 0.04346614,
+      "step": 8970
+    },
+    {
+      "epoch": 17.942,
+      "grad_norm": 1.5256109237670898,
+      "learning_rate": 2e-05,
+      "loss": 0.05348255,
+      "step": 8971
+    },
+    {
+      "epoch": 17.944,
+      "grad_norm": 1.0634779930114746,
+      "learning_rate": 2e-05,
+      "loss": 0.04019362,
+      "step": 8972
+    },
+    {
+      "epoch": 17.946,
+      "grad_norm": 1.4219046831130981,
+      "learning_rate": 2e-05,
+      "loss": 0.03420949,
+      "step": 8973
+    },
+    {
+      "epoch": 17.948,
+      "grad_norm": 1.1320511102676392,
+      "learning_rate": 2e-05,
+      "loss": 0.03867266,
+      "step": 8974
+    },
+    {
+      "epoch": 17.95,
+      "grad_norm": 1.6765214204788208,
+      "learning_rate": 2e-05,
+      "loss": 0.04724558,
+      "step": 8975
+    },
+    {
+      "epoch": 17.951999999999998,
+      "grad_norm": 1.8259295225143433,
+      "learning_rate": 2e-05,
+      "loss": 0.03461309,
+      "step": 8976
+    },
+    {
+      "epoch": 17.954,
+      "grad_norm": 1.2042784690856934,
+      "learning_rate": 2e-05,
+      "loss": 0.04839892,
+      "step": 8977
+    },
+    {
+      "epoch": 17.956,
+      "grad_norm": 1.5858327150344849,
+      "learning_rate": 2e-05,
+      "loss": 0.0500534,
+      "step": 8978
+    },
+    {
+      "epoch": 17.958,
+      "grad_norm": 1.3547513484954834,
+      "learning_rate": 2e-05,
+      "loss": 0.02594247,
+      "step": 8979
+    },
+    {
+      "epoch": 17.96,
+      "grad_norm": 1.0181814432144165,
+      "learning_rate": 2e-05,
+      "loss": 0.02916558,
+      "step": 8980
+    },
+    {
+      "epoch": 17.962,
+      "grad_norm": 1.466145634651184,
+      "learning_rate": 2e-05,
+      "loss": 0.04672647,
+      "step": 8981
+    },
+    {
+      "epoch": 17.964,
+      "grad_norm": 1.3053315877914429,
+      "learning_rate": 2e-05,
+      "loss": 0.03587544,
+      "step": 8982
+    },
+    {
+      "epoch": 17.966,
+      "grad_norm": 1.0494365692138672,
+      "learning_rate": 2e-05,
+      "loss": 0.02380068,
+      "step": 8983
+    },
+    {
+      "epoch": 17.968,
+      "grad_norm": 1.8024916648864746,
+      "learning_rate": 2e-05,
+      "loss": 0.03359849,
+      "step": 8984
+    },
+    {
+      "epoch": 17.97,
+      "grad_norm": 0.9630800485610962,
+      "learning_rate": 2e-05,
+      "loss": 0.03356433,
+      "step": 8985
+    },
+    {
+      "epoch": 17.972,
+      "grad_norm": 1.1515108346939087,
+      "learning_rate": 2e-05,
+      "loss": 0.03505271,
+      "step": 8986
+    },
+    {
+      "epoch": 17.974,
+      "grad_norm": 1.3020286560058594,
+      "learning_rate": 2e-05,
+      "loss": 0.05472892,
+      "step": 8987
+    },
+    {
+      "epoch": 17.976,
+      "grad_norm": 1.7581522464752197,
+      "learning_rate": 2e-05,
+      "loss": 0.0482644,
+      "step": 8988
+    },
+    {
+      "epoch": 17.978,
+      "grad_norm": 1.5518864393234253,
+      "learning_rate": 2e-05,
+      "loss": 0.04105756,
+      "step": 8989
+    },
+    {
+      "epoch": 17.98,
+      "grad_norm": 0.9530649781227112,
+      "learning_rate": 2e-05,
+      "loss": 0.02671498,
+      "step": 8990
+    },
+    {
+      "epoch": 17.982,
+      "grad_norm": 1.1680177450180054,
+      "learning_rate": 2e-05,
+      "loss": 0.02665569,
+      "step": 8991
+    },
+    {
+      "epoch": 17.984,
+      "grad_norm": 1.260867714881897,
+      "learning_rate": 2e-05,
+      "loss": 0.02883239,
+      "step": 8992
+    },
+    {
+      "epoch": 17.986,
+      "grad_norm": 1.3985310792922974,
+      "learning_rate": 2e-05,
+      "loss": 0.03463531,
+      "step": 8993
+    },
+    {
+      "epoch": 17.988,
+      "grad_norm": 1.3712165355682373,
+      "learning_rate": 2e-05,
+      "loss": 0.04792679,
+      "step": 8994
+    },
+    {
+      "epoch": 17.99,
+      "grad_norm": 1.8040891885757446,
+      "learning_rate": 2e-05,
+      "loss": 0.04169717,
+      "step": 8995
+    },
+    {
+      "epoch": 17.992,
+      "grad_norm": 2.077030897140503,
+      "learning_rate": 2e-05,
+      "loss": 0.04134944,
+      "step": 8996
+    },
+    {
+      "epoch": 17.994,
+      "grad_norm": 0.8118720054626465,
+      "learning_rate": 2e-05,
+      "loss": 0.01849389,
+      "step": 8997
+    },
+    {
+      "epoch": 17.996,
+      "grad_norm": 0.983228862285614,
+      "learning_rate": 2e-05,
+      "loss": 0.02795386,
+      "step": 8998
+    },
+    {
+      "epoch": 17.998,
+      "grad_norm": 1.1231356859207153,
+      "learning_rate": 2e-05,
+      "loss": 0.02436145,
+      "step": 8999
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 1.8219008445739746,
+      "learning_rate": 2e-05,
+      "loss": 0.03983248,
+      "step": 9000
+    },
+    {
+      "epoch": 18.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 0.998,
+        "AngleClassification_3": 0.9740518962075848,
+        "Equal_1": 0.996,
+        "Equal_2": 0.9700598802395209,
+        "Equal_3": 0.93812375249501,
+        "LineComparison_1": 0.998,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9940119760479041,
+        "Parallel_1": 0.9879759519038076,
+        "Parallel_2": 0.9979959919839679,
+        "Parallel_3": 0.992,
+        "Perpendicular_1": 0.996,
+        "Perpendicular_2": 0.978,
+        "Perpendicular_3": 0.7464929859719439,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9973333333333333,
+        "PointLiesOnCircle_3": 0.9896666666666666,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9740518962075848
+      },
+      "eval_runtime": 320.1045,
+      "eval_samples_per_second": 32.802,
+      "eval_steps_per_second": 0.656,
+      "step": 9000
+    },
+    {
+      "epoch": 18.002,
+      "grad_norm": 1.8216357231140137,
+      "learning_rate": 2e-05,
+      "loss": 0.03083301,
+      "step": 9001
+    },
+    {
+      "epoch": 18.004,
+      "grad_norm": 0.8909358978271484,
+      "learning_rate": 2e-05,
+      "loss": 0.02501141,
+      "step": 9002
+    },
+    {
+      "epoch": 18.006,
+      "grad_norm": 1.4257256984710693,
+      "learning_rate": 2e-05,
+      "loss": 0.03546458,
+      "step": 9003
+    },
+    {
+      "epoch": 18.008,
+      "grad_norm": 1.231162667274475,
+      "learning_rate": 2e-05,
+      "loss": 0.04436444,
+      "step": 9004
+    },
+    {
+      "epoch": 18.01,
+      "grad_norm": 2.071355104446411,
+      "learning_rate": 2e-05,
+      "loss": 0.05257452,
+      "step": 9005
+    },
+    {
+      "epoch": 18.012,
+      "grad_norm": 2.303741455078125,
+      "learning_rate": 2e-05,
+      "loss": 0.04662026,
+      "step": 9006
+    },
+    {
+      "epoch": 18.014,
+      "grad_norm": 1.0220463275909424,
+      "learning_rate": 2e-05,
+      "loss": 0.03330886,
+      "step": 9007
+    },
+    {
+      "epoch": 18.016,
+      "grad_norm": 2.040062665939331,
+      "learning_rate": 2e-05,
+      "loss": 0.04663801,
+      "step": 9008
+    },
+    {
+      "epoch": 18.018,
+      "grad_norm": 2.4291489124298096,
+      "learning_rate": 2e-05,
+      "loss": 0.05256157,
+      "step": 9009
+    },
+    {
+      "epoch": 18.02,
+      "grad_norm": 2.1016221046447754,
+      "learning_rate": 2e-05,
+      "loss": 0.04151364,
+      "step": 9010
+    },
+    {
+      "epoch": 18.022,
+      "grad_norm": 1.0601978302001953,
+      "learning_rate": 2e-05,
+      "loss": 0.04119922,
+      "step": 9011
+    },
+    {
+      "epoch": 18.024,
+      "grad_norm": 1.0514633655548096,
+      "learning_rate": 2e-05,
+      "loss": 0.03135135,
+      "step": 9012
+    },
+    {
+      "epoch": 18.026,
+      "grad_norm": 1.4121065139770508,
+      "learning_rate": 2e-05,
+      "loss": 0.04619808,
+      "step": 9013
+    },
+    {
+      "epoch": 18.028,
+      "grad_norm": 0.9003342986106873,
+      "learning_rate": 2e-05,
+      "loss": 0.02542493,
+      "step": 9014
+    },
+    {
+      "epoch": 18.03,
+      "grad_norm": 2.492168664932251,
+      "learning_rate": 2e-05,
+      "loss": 0.04310292,
+      "step": 9015
+    },
+    {
+      "epoch": 18.032,
+      "grad_norm": 1.2616229057312012,
+      "learning_rate": 2e-05,
+      "loss": 0.03716831,
+      "step": 9016
+    },
+    {
+      "epoch": 18.034,
+      "grad_norm": 1.5388479232788086,
+      "learning_rate": 2e-05,
+      "loss": 0.05547727,
+      "step": 9017
+    },
+    {
+      "epoch": 18.036,
+      "grad_norm": 2.8383405208587646,
+      "learning_rate": 2e-05,
+      "loss": 0.04588626,
+      "step": 9018
+    },
+    {
+      "epoch": 18.038,
+      "grad_norm": 1.0203205347061157,
+      "learning_rate": 2e-05,
+      "loss": 0.03126208,
+      "step": 9019
+    },
+    {
+      "epoch": 18.04,
+      "grad_norm": 1.4815860986709595,
+      "learning_rate": 2e-05,
+      "loss": 0.04387635,
+      "step": 9020
+    },
+    {
+      "epoch": 18.042,
+      "grad_norm": 1.163259744644165,
+      "learning_rate": 2e-05,
+      "loss": 0.0341371,
+      "step": 9021
+    },
+    {
+      "epoch": 18.044,
+      "grad_norm": 1.398336410522461,
+      "learning_rate": 2e-05,
+      "loss": 0.03288767,
+      "step": 9022
+    },
+    {
+      "epoch": 18.046,
+      "grad_norm": 1.4146158695220947,
+      "learning_rate": 2e-05,
+      "loss": 0.03783367,
+      "step": 9023
+    },
+    {
+      "epoch": 18.048,
+      "grad_norm": 1.2341772317886353,
+      "learning_rate": 2e-05,
+      "loss": 0.0348506,
+      "step": 9024
+    },
+    {
+      "epoch": 18.05,
+      "grad_norm": 1.4742604494094849,
+      "learning_rate": 2e-05,
+      "loss": 0.04025548,
+      "step": 9025
+    },
+    {
+      "epoch": 18.052,
+      "grad_norm": 1.2109090089797974,
+      "learning_rate": 2e-05,
+      "loss": 0.02520112,
+      "step": 9026
+    },
+    {
+      "epoch": 18.054,
+      "grad_norm": 0.9972143173217773,
+      "learning_rate": 2e-05,
+      "loss": 0.03415427,
+      "step": 9027
+    },
+    {
+      "epoch": 18.056,
+      "grad_norm": 1.110215425491333,
+      "learning_rate": 2e-05,
+      "loss": 0.03701129,
+      "step": 9028
+    },
+    {
+      "epoch": 18.058,
+      "grad_norm": 1.2750699520111084,
+      "learning_rate": 2e-05,
+      "loss": 0.03478452,
+      "step": 9029
+    },
+    {
+      "epoch": 18.06,
+      "grad_norm": 0.906257152557373,
+      "learning_rate": 2e-05,
+      "loss": 0.02598533,
+      "step": 9030
+    },
+    {
+      "epoch": 18.062,
+      "grad_norm": 2.5207362174987793,
+      "learning_rate": 2e-05,
+      "loss": 0.04807776,
+      "step": 9031
+    },
+    {
+      "epoch": 18.064,
+      "grad_norm": 1.246948480606079,
+      "learning_rate": 2e-05,
+      "loss": 0.03488917,
+      "step": 9032
+    },
+    {
+      "epoch": 18.066,
+      "grad_norm": 2.9162485599517822,
+      "learning_rate": 2e-05,
+      "loss": 0.03648629,
+      "step": 9033
+    },
+    {
+      "epoch": 18.068,
+      "grad_norm": 1.169908046722412,
+      "learning_rate": 2e-05,
+      "loss": 0.02641434,
+      "step": 9034
+    },
+    {
+      "epoch": 18.07,
+      "grad_norm": 1.152937889099121,
+      "learning_rate": 2e-05,
+      "loss": 0.03226584,
+      "step": 9035
+    },
+    {
+      "epoch": 18.072,
+      "grad_norm": 1.7191685438156128,
+      "learning_rate": 2e-05,
+      "loss": 0.05071743,
+      "step": 9036
+    },
+    {
+      "epoch": 18.074,
+      "grad_norm": 1.6077672243118286,
+      "learning_rate": 2e-05,
+      "loss": 0.02946869,
+      "step": 9037
+    },
+    {
+      "epoch": 18.076,
+      "grad_norm": 1.787205457687378,
+      "learning_rate": 2e-05,
+      "loss": 0.02539303,
+      "step": 9038
+    },
+    {
+      "epoch": 18.078,
+      "grad_norm": 1.3960199356079102,
+      "learning_rate": 2e-05,
+      "loss": 0.03590056,
+      "step": 9039
+    },
+    {
+      "epoch": 18.08,
+      "grad_norm": 1.8743473291397095,
+      "learning_rate": 2e-05,
+      "loss": 0.03464347,
+      "step": 9040
+    },
+    {
+      "epoch": 18.082,
+      "grad_norm": 1.2113933563232422,
+      "learning_rate": 2e-05,
+      "loss": 0.03156437,
+      "step": 9041
+    },
+    {
+      "epoch": 18.084,
+      "grad_norm": 1.8189163208007812,
+      "learning_rate": 2e-05,
+      "loss": 0.04444864,
+      "step": 9042
+    },
+    {
+      "epoch": 18.086,
+      "grad_norm": 1.960848331451416,
+      "learning_rate": 2e-05,
+      "loss": 0.03387512,
+      "step": 9043
+    },
+    {
+      "epoch": 18.088,
+      "grad_norm": 1.3409546613693237,
+      "learning_rate": 2e-05,
+      "loss": 0.03057274,
+      "step": 9044
+    },
+    {
+      "epoch": 18.09,
+      "grad_norm": 1.4044934511184692,
+      "learning_rate": 2e-05,
+      "loss": 0.03606112,
+      "step": 9045
+    },
+    {
+      "epoch": 18.092,
+      "grad_norm": 1.1396024227142334,
+      "learning_rate": 2e-05,
+      "loss": 0.03115062,
+      "step": 9046
+    },
+    {
+      "epoch": 18.094,
+      "grad_norm": 1.4510269165039062,
+      "learning_rate": 2e-05,
+      "loss": 0.02674197,
+      "step": 9047
+    },
+    {
+      "epoch": 18.096,
+      "grad_norm": 1.1940547227859497,
+      "learning_rate": 2e-05,
+      "loss": 0.03157366,
+      "step": 9048
+    },
+    {
+      "epoch": 18.098,
+      "grad_norm": 1.4505969285964966,
+      "learning_rate": 2e-05,
+      "loss": 0.04264753,
+      "step": 9049
+    },
+    {
+      "epoch": 18.1,
+      "grad_norm": 1.8302192687988281,
+      "learning_rate": 2e-05,
+      "loss": 0.04496709,
+      "step": 9050
+    },
+    {
+      "epoch": 18.102,
+      "grad_norm": 1.3621551990509033,
+      "learning_rate": 2e-05,
+      "loss": 0.03655332,
+      "step": 9051
+    },
+    {
+      "epoch": 18.104,
+      "grad_norm": 0.982716977596283,
+      "learning_rate": 2e-05,
+      "loss": 0.03123835,
+      "step": 9052
+    },
+    {
+      "epoch": 18.106,
+      "grad_norm": 1.5590128898620605,
+      "learning_rate": 2e-05,
+      "loss": 0.04147382,
+      "step": 9053
+    },
+    {
+      "epoch": 18.108,
+      "grad_norm": 1.2122939825057983,
+      "learning_rate": 2e-05,
+      "loss": 0.03553297,
+      "step": 9054
+    },
+    {
+      "epoch": 18.11,
+      "grad_norm": 1.2034052610397339,
+      "learning_rate": 2e-05,
+      "loss": 0.04050352,
+      "step": 9055
+    },
+    {
+      "epoch": 18.112,
+      "grad_norm": 1.3900257349014282,
+      "learning_rate": 2e-05,
+      "loss": 0.05008593,
+      "step": 9056
+    },
+    {
+      "epoch": 18.114,
+      "grad_norm": 1.8129817247390747,
+      "learning_rate": 2e-05,
+      "loss": 0.03101974,
+      "step": 9057
+    },
+    {
+      "epoch": 18.116,
+      "grad_norm": 1.3509774208068848,
+      "learning_rate": 2e-05,
+      "loss": 0.03191446,
+      "step": 9058
+    },
+    {
+      "epoch": 18.118,
+      "grad_norm": 1.6840113401412964,
+      "learning_rate": 2e-05,
+      "loss": 0.03167528,
+      "step": 9059
+    },
+    {
+      "epoch": 18.12,
+      "grad_norm": 1.132699966430664,
+      "learning_rate": 2e-05,
+      "loss": 0.03625121,
+      "step": 9060
+    },
+    {
+      "epoch": 18.122,
+      "grad_norm": 1.4251673221588135,
+      "learning_rate": 2e-05,
+      "loss": 0.04312788,
+      "step": 9061
+    },
+    {
+      "epoch": 18.124,
+      "grad_norm": 1.160861611366272,
+      "learning_rate": 2e-05,
+      "loss": 0.03482513,
+      "step": 9062
+    },
+    {
+      "epoch": 18.126,
+      "grad_norm": 1.3063910007476807,
+      "learning_rate": 2e-05,
+      "loss": 0.03584398,
+      "step": 9063
+    },
+    {
+      "epoch": 18.128,
+      "grad_norm": 1.410151481628418,
+      "learning_rate": 2e-05,
+      "loss": 0.04368694,
+      "step": 9064
+    },
+    {
+      "epoch": 18.13,
+      "grad_norm": 2.592433452606201,
+      "learning_rate": 2e-05,
+      "loss": 0.03493256,
+      "step": 9065
+    },
+    {
+      "epoch": 18.132,
+      "grad_norm": 1.192308783531189,
+      "learning_rate": 2e-05,
+      "loss": 0.03010586,
+      "step": 9066
+    },
+    {
+      "epoch": 18.134,
+      "grad_norm": 1.1783413887023926,
+      "learning_rate": 2e-05,
+      "loss": 0.04549911,
+      "step": 9067
+    },
+    {
+      "epoch": 18.136,
+      "grad_norm": 0.9659681916236877,
+      "learning_rate": 2e-05,
+      "loss": 0.02878298,
+      "step": 9068
+    },
+    {
+      "epoch": 18.138,
+      "grad_norm": 1.4795477390289307,
+      "learning_rate": 2e-05,
+      "loss": 0.03338488,
+      "step": 9069
+    },
+    {
+      "epoch": 18.14,
+      "grad_norm": 1.3997915983200073,
+      "learning_rate": 2e-05,
+      "loss": 0.04527793,
+      "step": 9070
+    },
+    {
+      "epoch": 18.142,
+      "grad_norm": 0.974898099899292,
+      "learning_rate": 2e-05,
+      "loss": 0.02718613,
+      "step": 9071
+    },
+    {
+      "epoch": 18.144,
+      "grad_norm": 1.643044114112854,
+      "learning_rate": 2e-05,
+      "loss": 0.04503238,
+      "step": 9072
+    },
+    {
+      "epoch": 18.146,
+      "grad_norm": 2.4663209915161133,
+      "learning_rate": 2e-05,
+      "loss": 0.05017319,
+      "step": 9073
+    },
+    {
+      "epoch": 18.148,
+      "grad_norm": 1.3137898445129395,
+      "learning_rate": 2e-05,
+      "loss": 0.03679071,
+      "step": 9074
+    },
+    {
+      "epoch": 18.15,
+      "grad_norm": 1.2270176410675049,
+      "learning_rate": 2e-05,
+      "loss": 0.02902906,
+      "step": 9075
+    },
+    {
+      "epoch": 18.152,
+      "grad_norm": 1.5809822082519531,
+      "learning_rate": 2e-05,
+      "loss": 0.02031413,
+      "step": 9076
+    },
+    {
+      "epoch": 18.154,
+      "grad_norm": 1.235666275024414,
+      "learning_rate": 2e-05,
+      "loss": 0.0429434,
+      "step": 9077
+    },
+    {
+      "epoch": 18.156,
+      "grad_norm": 1.3485252857208252,
+      "learning_rate": 2e-05,
+      "loss": 0.04158372,
+      "step": 9078
+    },
+    {
+      "epoch": 18.158,
+      "grad_norm": 1.5450528860092163,
+      "learning_rate": 2e-05,
+      "loss": 0.03034053,
+      "step": 9079
+    },
+    {
+      "epoch": 18.16,
+      "grad_norm": 1.102346420288086,
+      "learning_rate": 2e-05,
+      "loss": 0.03850158,
+      "step": 9080
+    },
+    {
+      "epoch": 18.162,
+      "grad_norm": 1.7495108842849731,
+      "learning_rate": 2e-05,
+      "loss": 0.04332937,
+      "step": 9081
+    },
+    {
+      "epoch": 18.164,
+      "grad_norm": 1.059772253036499,
+      "learning_rate": 2e-05,
+      "loss": 0.04180691,
+      "step": 9082
+    },
+    {
+      "epoch": 18.166,
+      "grad_norm": 2.6385281085968018,
+      "learning_rate": 2e-05,
+      "loss": 0.04071419,
+      "step": 9083
+    },
+    {
+      "epoch": 18.168,
+      "grad_norm": 1.432405948638916,
+      "learning_rate": 2e-05,
+      "loss": 0.02765241,
+      "step": 9084
+    },
+    {
+      "epoch": 18.17,
+      "grad_norm": 1.8066904544830322,
+      "learning_rate": 2e-05,
+      "loss": 0.04745378,
+      "step": 9085
+    },
+    {
+      "epoch": 18.172,
+      "grad_norm": 0.885301411151886,
+      "learning_rate": 2e-05,
+      "loss": 0.02941084,
+      "step": 9086
+    },
+    {
+      "epoch": 18.174,
+      "grad_norm": 1.0743720531463623,
+      "learning_rate": 2e-05,
+      "loss": 0.03165472,
+      "step": 9087
+    },
+    {
+      "epoch": 18.176,
+      "grad_norm": 1.8421827554702759,
+      "learning_rate": 2e-05,
+      "loss": 0.03389416,
+      "step": 9088
+    },
+    {
+      "epoch": 18.178,
+      "grad_norm": 1.3630011081695557,
+      "learning_rate": 2e-05,
+      "loss": 0.03734118,
+      "step": 9089
+    },
+    {
+      "epoch": 18.18,
+      "grad_norm": 1.5092341899871826,
+      "learning_rate": 2e-05,
+      "loss": 0.04258218,
+      "step": 9090
+    },
+    {
+      "epoch": 18.182,
+      "grad_norm": 2.4805521965026855,
+      "learning_rate": 2e-05,
+      "loss": 0.05131678,
+      "step": 9091
+    },
+    {
+      "epoch": 18.184,
+      "grad_norm": 1.2178826332092285,
+      "learning_rate": 2e-05,
+      "loss": 0.03363734,
+      "step": 9092
+    },
+    {
+      "epoch": 18.186,
+      "grad_norm": 1.2770280838012695,
+      "learning_rate": 2e-05,
+      "loss": 0.04184774,
+      "step": 9093
+    },
+    {
+      "epoch": 18.188,
+      "grad_norm": 2.7290360927581787,
+      "learning_rate": 2e-05,
+      "loss": 0.04432489,
+      "step": 9094
+    },
+    {
+      "epoch": 18.19,
+      "grad_norm": 1.8310285806655884,
+      "learning_rate": 2e-05,
+      "loss": 0.03530759,
+      "step": 9095
+    },
+    {
+      "epoch": 18.192,
+      "grad_norm": 1.0748549699783325,
+      "learning_rate": 2e-05,
+      "loss": 0.03645511,
+      "step": 9096
+    },
+    {
+      "epoch": 18.194,
+      "grad_norm": 1.8210999965667725,
+      "learning_rate": 2e-05,
+      "loss": 0.04495174,
+      "step": 9097
+    },
+    {
+      "epoch": 18.196,
+      "grad_norm": 1.0958075523376465,
+      "learning_rate": 2e-05,
+      "loss": 0.02407749,
+      "step": 9098
+    },
+    {
+      "epoch": 18.198,
+      "grad_norm": 1.8292564153671265,
+      "learning_rate": 2e-05,
+      "loss": 0.0362148,
+      "step": 9099
+    },
+    {
+      "epoch": 18.2,
+      "grad_norm": 1.0899569988250732,
+      "learning_rate": 2e-05,
+      "loss": 0.03374222,
+      "step": 9100
+    },
+    {
+      "epoch": 18.202,
+      "grad_norm": 1.0096633434295654,
+      "learning_rate": 2e-05,
+      "loss": 0.03209589,
+      "step": 9101
+    },
+    {
+      "epoch": 18.204,
+      "grad_norm": 1.1050734519958496,
+      "learning_rate": 2e-05,
+      "loss": 0.03580528,
+      "step": 9102
+    },
+    {
+      "epoch": 18.206,
+      "grad_norm": 2.117532253265381,
+      "learning_rate": 2e-05,
+      "loss": 0.04968563,
+      "step": 9103
+    },
+    {
+      "epoch": 18.208,
+      "grad_norm": 2.2677221298217773,
+      "learning_rate": 2e-05,
+      "loss": 0.06845517,
+      "step": 9104
+    },
+    {
+      "epoch": 18.21,
+      "grad_norm": 0.8368435502052307,
+      "learning_rate": 2e-05,
+      "loss": 0.02537123,
+      "step": 9105
+    },
+    {
+      "epoch": 18.212,
+      "grad_norm": 1.149839162826538,
+      "learning_rate": 2e-05,
+      "loss": 0.03183338,
+      "step": 9106
+    },
+    {
+      "epoch": 18.214,
+      "grad_norm": 1.2164126634597778,
+      "learning_rate": 2e-05,
+      "loss": 0.0268788,
+      "step": 9107
+    },
+    {
+      "epoch": 18.216,
+      "grad_norm": 1.1083381175994873,
+      "learning_rate": 2e-05,
+      "loss": 0.0347072,
+      "step": 9108
+    },
+    {
+      "epoch": 18.218,
+      "grad_norm": 1.0997363328933716,
+      "learning_rate": 2e-05,
+      "loss": 0.03136883,
+      "step": 9109
+    },
+    {
+      "epoch": 18.22,
+      "grad_norm": 1.3243005275726318,
+      "learning_rate": 2e-05,
+      "loss": 0.04105141,
+      "step": 9110
+    },
+    {
+      "epoch": 18.222,
+      "grad_norm": 1.123883843421936,
+      "learning_rate": 2e-05,
+      "loss": 0.04037739,
+      "step": 9111
+    },
+    {
+      "epoch": 18.224,
+      "grad_norm": 1.3631707429885864,
+      "learning_rate": 2e-05,
+      "loss": 0.02595056,
+      "step": 9112
+    },
+    {
+      "epoch": 18.226,
+      "grad_norm": 1.5019488334655762,
+      "learning_rate": 2e-05,
+      "loss": 0.03916592,
+      "step": 9113
+    },
+    {
+      "epoch": 18.228,
+      "grad_norm": 1.153467059135437,
+      "learning_rate": 2e-05,
+      "loss": 0.03865172,
+      "step": 9114
+    },
+    {
+      "epoch": 18.23,
+      "grad_norm": 1.4573817253112793,
+      "learning_rate": 2e-05,
+      "loss": 0.02908707,
+      "step": 9115
+    },
+    {
+      "epoch": 18.232,
+      "grad_norm": 1.7540217638015747,
+      "learning_rate": 2e-05,
+      "loss": 0.03963351,
+      "step": 9116
+    },
+    {
+      "epoch": 18.234,
+      "grad_norm": 2.5012366771698,
+      "learning_rate": 2e-05,
+      "loss": 0.04361957,
+      "step": 9117
+    },
+    {
+      "epoch": 18.236,
+      "grad_norm": 1.4801292419433594,
+      "learning_rate": 2e-05,
+      "loss": 0.03021835,
+      "step": 9118
+    },
+    {
+      "epoch": 18.238,
+      "grad_norm": 1.3169306516647339,
+      "learning_rate": 2e-05,
+      "loss": 0.03268102,
+      "step": 9119
+    },
+    {
+      "epoch": 18.24,
+      "grad_norm": 1.0948139429092407,
+      "learning_rate": 2e-05,
+      "loss": 0.02789821,
+      "step": 9120
+    },
+    {
+      "epoch": 18.242,
+      "grad_norm": 1.2871372699737549,
+      "learning_rate": 2e-05,
+      "loss": 0.03492365,
+      "step": 9121
+    },
+    {
+      "epoch": 18.244,
+      "grad_norm": 1.7145776748657227,
+      "learning_rate": 2e-05,
+      "loss": 0.03364137,
+      "step": 9122
+    },
+    {
+      "epoch": 18.246,
+      "grad_norm": 1.5056641101837158,
+      "learning_rate": 2e-05,
+      "loss": 0.03888101,
+      "step": 9123
+    },
+    {
+      "epoch": 18.248,
+      "grad_norm": 2.125732660293579,
+      "learning_rate": 2e-05,
+      "loss": 0.04138645,
+      "step": 9124
+    },
+    {
+      "epoch": 18.25,
+      "grad_norm": 2.0914463996887207,
+      "learning_rate": 2e-05,
+      "loss": 0.04323914,
+      "step": 9125
+    },
+    {
+      "epoch": 18.252,
+      "grad_norm": 1.3840261697769165,
+      "learning_rate": 2e-05,
+      "loss": 0.04744683,
+      "step": 9126
+    },
+    {
+      "epoch": 18.254,
+      "grad_norm": 1.3877594470977783,
+      "learning_rate": 2e-05,
+      "loss": 0.03428207,
+      "step": 9127
+    },
+    {
+      "epoch": 18.256,
+      "grad_norm": 1.9282671213150024,
+      "learning_rate": 2e-05,
+      "loss": 0.03370817,
+      "step": 9128
+    },
+    {
+      "epoch": 18.258,
+      "grad_norm": 1.2625566720962524,
+      "learning_rate": 2e-05,
+      "loss": 0.03845701,
+      "step": 9129
+    },
+    {
+      "epoch": 18.26,
+      "grad_norm": 1.497373104095459,
+      "learning_rate": 2e-05,
+      "loss": 0.02765183,
+      "step": 9130
+    },
+    {
+      "epoch": 18.262,
+      "grad_norm": 0.938334584236145,
+      "learning_rate": 2e-05,
+      "loss": 0.0283683,
+      "step": 9131
+    },
+    {
+      "epoch": 18.264,
+      "grad_norm": 0.8075963258743286,
+      "learning_rate": 2e-05,
+      "loss": 0.0257271,
+      "step": 9132
+    },
+    {
+      "epoch": 18.266,
+      "grad_norm": 2.6727170944213867,
+      "learning_rate": 2e-05,
+      "loss": 0.03528313,
+      "step": 9133
+    },
+    {
+      "epoch": 18.268,
+      "grad_norm": 1.1714603900909424,
+      "learning_rate": 2e-05,
+      "loss": 0.03839287,
+      "step": 9134
+    },
+    {
+      "epoch": 18.27,
+      "grad_norm": 1.3799819946289062,
+      "learning_rate": 2e-05,
+      "loss": 0.04040816,
+      "step": 9135
+    },
+    {
+      "epoch": 18.272,
+      "grad_norm": 0.8756110668182373,
+      "learning_rate": 2e-05,
+      "loss": 0.02800023,
+      "step": 9136
+    },
+    {
+      "epoch": 18.274,
+      "grad_norm": 1.278294563293457,
+      "learning_rate": 2e-05,
+      "loss": 0.03553412,
+      "step": 9137
+    },
+    {
+      "epoch": 18.276,
+      "grad_norm": 1.649017333984375,
+      "learning_rate": 2e-05,
+      "loss": 0.0240197,
+      "step": 9138
+    },
+    {
+      "epoch": 18.278,
+      "grad_norm": 1.2307323217391968,
+      "learning_rate": 2e-05,
+      "loss": 0.03196525,
+      "step": 9139
+    },
+    {
+      "epoch": 18.28,
+      "grad_norm": 1.2029914855957031,
+      "learning_rate": 2e-05,
+      "loss": 0.04263403,
+      "step": 9140
+    },
+    {
+      "epoch": 18.282,
+      "grad_norm": 1.1589373350143433,
+      "learning_rate": 2e-05,
+      "loss": 0.04057025,
+      "step": 9141
+    },
+    {
+      "epoch": 18.284,
+      "grad_norm": 1.0348948240280151,
+      "learning_rate": 2e-05,
+      "loss": 0.04266765,
+      "step": 9142
+    },
+    {
+      "epoch": 18.286,
+      "grad_norm": 0.8373733162879944,
+      "learning_rate": 2e-05,
+      "loss": 0.01983244,
+      "step": 9143
+    },
+    {
+      "epoch": 18.288,
+      "grad_norm": 2.752629041671753,
+      "learning_rate": 2e-05,
+      "loss": 0.04166656,
+      "step": 9144
+    },
+    {
+      "epoch": 18.29,
+      "grad_norm": 1.1935979127883911,
+      "learning_rate": 2e-05,
+      "loss": 0.04558745,
+      "step": 9145
+    },
+    {
+      "epoch": 18.292,
+      "grad_norm": 0.9616577625274658,
+      "learning_rate": 2e-05,
+      "loss": 0.02630688,
+      "step": 9146
+    },
+    {
+      "epoch": 18.294,
+      "grad_norm": 1.3004382848739624,
+      "learning_rate": 2e-05,
+      "loss": 0.04482454,
+      "step": 9147
+    },
+    {
+      "epoch": 18.296,
+      "grad_norm": 2.1716208457946777,
+      "learning_rate": 2e-05,
+      "loss": 0.03779503,
+      "step": 9148
+    },
+    {
+      "epoch": 18.298,
+      "grad_norm": 1.1647388935089111,
+      "learning_rate": 2e-05,
+      "loss": 0.03616937,
+      "step": 9149
+    },
+    {
+      "epoch": 18.3,
+      "grad_norm": 1.6994318962097168,
+      "learning_rate": 2e-05,
+      "loss": 0.03434645,
+      "step": 9150
+    },
+    {
+      "epoch": 18.302,
+      "grad_norm": 1.055916666984558,
+      "learning_rate": 2e-05,
+      "loss": 0.04265877,
+      "step": 9151
+    },
+    {
+      "epoch": 18.304,
+      "grad_norm": 1.1713130474090576,
+      "learning_rate": 2e-05,
+      "loss": 0.03589529,
+      "step": 9152
+    },
+    {
+      "epoch": 18.306,
+      "grad_norm": 1.2549328804016113,
+      "learning_rate": 2e-05,
+      "loss": 0.03371196,
+      "step": 9153
+    },
+    {
+      "epoch": 18.308,
+      "grad_norm": 0.9361233115196228,
+      "learning_rate": 2e-05,
+      "loss": 0.02483992,
+      "step": 9154
+    },
+    {
+      "epoch": 18.31,
+      "grad_norm": 1.0725373029708862,
+      "learning_rate": 2e-05,
+      "loss": 0.03467751,
+      "step": 9155
+    },
+    {
+      "epoch": 18.312,
+      "grad_norm": 1.8279783725738525,
+      "learning_rate": 2e-05,
+      "loss": 0.05167603,
+      "step": 9156
+    },
+    {
+      "epoch": 18.314,
+      "grad_norm": 0.995499312877655,
+      "learning_rate": 2e-05,
+      "loss": 0.02847271,
+      "step": 9157
+    },
+    {
+      "epoch": 18.316,
+      "grad_norm": 1.0232125520706177,
+      "learning_rate": 2e-05,
+      "loss": 0.0327347,
+      "step": 9158
+    },
+    {
+      "epoch": 18.318,
+      "grad_norm": 1.25034761428833,
+      "learning_rate": 2e-05,
+      "loss": 0.04012694,
+      "step": 9159
+    },
+    {
+      "epoch": 18.32,
+      "grad_norm": 2.3386640548706055,
+      "learning_rate": 2e-05,
+      "loss": 0.03319935,
+      "step": 9160
+    },
+    {
+      "epoch": 18.322,
+      "grad_norm": 1.7142624855041504,
+      "learning_rate": 2e-05,
+      "loss": 0.03917816,
+      "step": 9161
+    },
+    {
+      "epoch": 18.324,
+      "grad_norm": 1.209825038909912,
+      "learning_rate": 2e-05,
+      "loss": 0.03275396,
+      "step": 9162
+    },
+    {
+      "epoch": 18.326,
+      "grad_norm": 1.3008837699890137,
+      "learning_rate": 2e-05,
+      "loss": 0.05405723,
+      "step": 9163
+    },
+    {
+      "epoch": 18.328,
+      "grad_norm": 1.4746818542480469,
+      "learning_rate": 2e-05,
+      "loss": 0.05087238,
+      "step": 9164
+    },
+    {
+      "epoch": 18.33,
+      "grad_norm": 1.1503615379333496,
+      "learning_rate": 2e-05,
+      "loss": 0.03378589,
+      "step": 9165
+    },
+    {
+      "epoch": 18.332,
+      "grad_norm": 1.8128838539123535,
+      "learning_rate": 2e-05,
+      "loss": 0.0437137,
+      "step": 9166
+    },
+    {
+      "epoch": 18.334,
+      "grad_norm": 0.9782925248146057,
+      "learning_rate": 2e-05,
+      "loss": 0.02864728,
+      "step": 9167
+    },
+    {
+      "epoch": 18.336,
+      "grad_norm": 1.0823737382888794,
+      "learning_rate": 2e-05,
+      "loss": 0.03686111,
+      "step": 9168
+    },
+    {
+      "epoch": 18.338,
+      "grad_norm": 1.7901039123535156,
+      "learning_rate": 2e-05,
+      "loss": 0.05045727,
+      "step": 9169
+    },
+    {
+      "epoch": 18.34,
+      "grad_norm": 1.2802319526672363,
+      "learning_rate": 2e-05,
+      "loss": 0.031836,
+      "step": 9170
+    },
+    {
+      "epoch": 18.342,
+      "grad_norm": 1.067213535308838,
+      "learning_rate": 2e-05,
+      "loss": 0.04307377,
+      "step": 9171
+    },
+    {
+      "epoch": 18.344,
+      "grad_norm": 2.6261112689971924,
+      "learning_rate": 2e-05,
+      "loss": 0.03852256,
+      "step": 9172
+    },
+    {
+      "epoch": 18.346,
+      "grad_norm": 2.581172466278076,
+      "learning_rate": 2e-05,
+      "loss": 0.04666492,
+      "step": 9173
+    },
+    {
+      "epoch": 18.348,
+      "grad_norm": 1.2471551895141602,
+      "learning_rate": 2e-05,
+      "loss": 0.03403299,
+      "step": 9174
+    },
+    {
+      "epoch": 18.35,
+      "grad_norm": 0.8967499732971191,
+      "learning_rate": 2e-05,
+      "loss": 0.02881518,
+      "step": 9175
+    },
+    {
+      "epoch": 18.352,
+      "grad_norm": 1.7108482122421265,
+      "learning_rate": 2e-05,
+      "loss": 0.03858823,
+      "step": 9176
+    },
+    {
+      "epoch": 18.354,
+      "grad_norm": 1.9164936542510986,
+      "learning_rate": 2e-05,
+      "loss": 0.02809501,
+      "step": 9177
+    },
+    {
+      "epoch": 18.356,
+      "grad_norm": 1.0232477188110352,
+      "learning_rate": 2e-05,
+      "loss": 0.02891929,
+      "step": 9178
+    },
+    {
+      "epoch": 18.358,
+      "grad_norm": 1.1295270919799805,
+      "learning_rate": 2e-05,
+      "loss": 0.0412226,
+      "step": 9179
+    },
+    {
+      "epoch": 18.36,
+      "grad_norm": 1.951155662536621,
+      "learning_rate": 2e-05,
+      "loss": 0.05396903,
+      "step": 9180
+    },
+    {
+      "epoch": 18.362,
+      "grad_norm": 1.2782940864562988,
+      "learning_rate": 2e-05,
+      "loss": 0.03953922,
+      "step": 9181
+    },
+    {
+      "epoch": 18.364,
+      "grad_norm": 1.5052493810653687,
+      "learning_rate": 2e-05,
+      "loss": 0.04199969,
+      "step": 9182
+    },
+    {
+      "epoch": 18.366,
+      "grad_norm": 2.334589958190918,
+      "learning_rate": 2e-05,
+      "loss": 0.03334442,
+      "step": 9183
+    },
+    {
+      "epoch": 18.368,
+      "grad_norm": 1.534926176071167,
+      "learning_rate": 2e-05,
+      "loss": 0.03135791,
+      "step": 9184
+    },
+    {
+      "epoch": 18.37,
+      "grad_norm": 2.0098283290863037,
+      "learning_rate": 2e-05,
+      "loss": 0.04160352,
+      "step": 9185
+    },
+    {
+      "epoch": 18.372,
+      "grad_norm": 0.8825958967208862,
+      "learning_rate": 2e-05,
+      "loss": 0.02310374,
+      "step": 9186
+    },
+    {
+      "epoch": 18.374,
+      "grad_norm": 1.5516126155853271,
+      "learning_rate": 2e-05,
+      "loss": 0.03449169,
+      "step": 9187
+    },
+    {
+      "epoch": 18.376,
+      "grad_norm": 1.0211329460144043,
+      "learning_rate": 2e-05,
+      "loss": 0.03518451,
+      "step": 9188
+    },
+    {
+      "epoch": 18.378,
+      "grad_norm": 1.4555100202560425,
+      "learning_rate": 2e-05,
+      "loss": 0.03537372,
+      "step": 9189
+    },
+    {
+      "epoch": 18.38,
+      "grad_norm": 1.6372886896133423,
+      "learning_rate": 2e-05,
+      "loss": 0.03058396,
+      "step": 9190
+    },
+    {
+      "epoch": 18.382,
+      "grad_norm": 1.5627180337905884,
+      "learning_rate": 2e-05,
+      "loss": 0.02930931,
+      "step": 9191
+    },
+    {
+      "epoch": 18.384,
+      "grad_norm": 1.7449994087219238,
+      "learning_rate": 2e-05,
+      "loss": 0.03934868,
+      "step": 9192
+    },
+    {
+      "epoch": 18.386,
+      "grad_norm": 0.9360613226890564,
+      "learning_rate": 2e-05,
+      "loss": 0.02767797,
+      "step": 9193
+    },
+    {
+      "epoch": 18.388,
+      "grad_norm": 1.376953125,
+      "learning_rate": 2e-05,
+      "loss": 0.02869485,
+      "step": 9194
+    },
+    {
+      "epoch": 18.39,
+      "grad_norm": 0.9833387732505798,
+      "learning_rate": 2e-05,
+      "loss": 0.02881843,
+      "step": 9195
+    },
+    {
+      "epoch": 18.392,
+      "grad_norm": 1.3985074758529663,
+      "learning_rate": 2e-05,
+      "loss": 0.03320839,
+      "step": 9196
+    },
+    {
+      "epoch": 18.394,
+      "grad_norm": 1.116469144821167,
+      "learning_rate": 2e-05,
+      "loss": 0.031327,
+      "step": 9197
+    },
+    {
+      "epoch": 18.396,
+      "grad_norm": 1.913405418395996,
+      "learning_rate": 2e-05,
+      "loss": 0.03109054,
+      "step": 9198
+    },
+    {
+      "epoch": 18.398,
+      "grad_norm": 0.9129771590232849,
+      "learning_rate": 2e-05,
+      "loss": 0.02565073,
+      "step": 9199
+    },
+    {
+      "epoch": 18.4,
+      "grad_norm": 0.8751816749572754,
+      "learning_rate": 2e-05,
+      "loss": 0.02509234,
+      "step": 9200
+    },
+    {
+      "epoch": 18.402,
+      "grad_norm": 1.0799328088760376,
+      "learning_rate": 2e-05,
+      "loss": 0.02929278,
+      "step": 9201
+    },
+    {
+      "epoch": 18.404,
+      "grad_norm": 1.4381532669067383,
+      "learning_rate": 2e-05,
+      "loss": 0.03288494,
+      "step": 9202
+    },
+    {
+      "epoch": 18.406,
+      "grad_norm": 1.2862842082977295,
+      "learning_rate": 2e-05,
+      "loss": 0.04411345,
+      "step": 9203
+    },
+    {
+      "epoch": 18.408,
+      "grad_norm": 1.2750723361968994,
+      "learning_rate": 2e-05,
+      "loss": 0.04329745,
+      "step": 9204
+    },
+    {
+      "epoch": 18.41,
+      "grad_norm": 1.3029229640960693,
+      "learning_rate": 2e-05,
+      "loss": 0.03379948,
+      "step": 9205
+    },
+    {
+      "epoch": 18.412,
+      "grad_norm": 1.3032692670822144,
+      "learning_rate": 2e-05,
+      "loss": 0.04243733,
+      "step": 9206
+    },
+    {
+      "epoch": 18.414,
+      "grad_norm": 1.4686411619186401,
+      "learning_rate": 2e-05,
+      "loss": 0.04486408,
+      "step": 9207
+    },
+    {
+      "epoch": 18.416,
+      "grad_norm": 2.9766223430633545,
+      "learning_rate": 2e-05,
+      "loss": 0.04959375,
+      "step": 9208
+    },
+    {
+      "epoch": 18.418,
+      "grad_norm": 1.8033760786056519,
+      "learning_rate": 2e-05,
+      "loss": 0.04121406,
+      "step": 9209
+    },
+    {
+      "epoch": 18.42,
+      "grad_norm": 1.320888876914978,
+      "learning_rate": 2e-05,
+      "loss": 0.03473954,
+      "step": 9210
+    },
+    {
+      "epoch": 18.422,
+      "grad_norm": 1.6498913764953613,
+      "learning_rate": 2e-05,
+      "loss": 0.04184615,
+      "step": 9211
+    },
+    {
+      "epoch": 18.424,
+      "grad_norm": 1.0657768249511719,
+      "learning_rate": 2e-05,
+      "loss": 0.03823342,
+      "step": 9212
+    },
+    {
+      "epoch": 18.426,
+      "grad_norm": 1.5349171161651611,
+      "learning_rate": 2e-05,
+      "loss": 0.04052274,
+      "step": 9213
+    },
+    {
+      "epoch": 18.428,
+      "grad_norm": 1.3626680374145508,
+      "learning_rate": 2e-05,
+      "loss": 0.03621456,
+      "step": 9214
+    },
+    {
+      "epoch": 18.43,
+      "grad_norm": 2.154515027999878,
+      "learning_rate": 2e-05,
+      "loss": 0.05104381,
+      "step": 9215
+    },
+    {
+      "epoch": 18.432,
+      "grad_norm": 2.7905545234680176,
+      "learning_rate": 2e-05,
+      "loss": 0.04180931,
+      "step": 9216
+    },
+    {
+      "epoch": 18.434,
+      "grad_norm": 1.2199686765670776,
+      "learning_rate": 2e-05,
+      "loss": 0.0278045,
+      "step": 9217
+    },
+    {
+      "epoch": 18.436,
+      "grad_norm": 1.7791190147399902,
+      "learning_rate": 2e-05,
+      "loss": 0.03621379,
+      "step": 9218
+    },
+    {
+      "epoch": 18.438,
+      "grad_norm": 2.6824915409088135,
+      "learning_rate": 2e-05,
+      "loss": 0.04269816,
+      "step": 9219
+    },
+    {
+      "epoch": 18.44,
+      "grad_norm": 1.4576973915100098,
+      "learning_rate": 2e-05,
+      "loss": 0.04696831,
+      "step": 9220
+    },
+    {
+      "epoch": 18.442,
+      "grad_norm": 1.0133570432662964,
+      "learning_rate": 2e-05,
+      "loss": 0.03169259,
+      "step": 9221
+    },
+    {
+      "epoch": 18.444,
+      "grad_norm": 1.0140568017959595,
+      "learning_rate": 2e-05,
+      "loss": 0.02977751,
+      "step": 9222
+    },
+    {
+      "epoch": 18.446,
+      "grad_norm": 1.2628326416015625,
+      "learning_rate": 2e-05,
+      "loss": 0.0529662,
+      "step": 9223
+    },
+    {
+      "epoch": 18.448,
+      "grad_norm": 1.7104601860046387,
+      "learning_rate": 2e-05,
+      "loss": 0.05581909,
+      "step": 9224
+    },
+    {
+      "epoch": 18.45,
+      "grad_norm": 0.982793390750885,
+      "learning_rate": 2e-05,
+      "loss": 0.03106131,
+      "step": 9225
+    },
+    {
+      "epoch": 18.452,
+      "grad_norm": 1.2446315288543701,
+      "learning_rate": 2e-05,
+      "loss": 0.04647019,
+      "step": 9226
+    },
+    {
+      "epoch": 18.454,
+      "grad_norm": 1.1005967855453491,
+      "learning_rate": 2e-05,
+      "loss": 0.02941687,
+      "step": 9227
+    },
+    {
+      "epoch": 18.456,
+      "grad_norm": 1.3427083492279053,
+      "learning_rate": 2e-05,
+      "loss": 0.03301994,
+      "step": 9228
+    },
+    {
+      "epoch": 18.458,
+      "grad_norm": 0.9936813712120056,
+      "learning_rate": 2e-05,
+      "loss": 0.03047378,
+      "step": 9229
+    },
+    {
+      "epoch": 18.46,
+      "grad_norm": 1.9145174026489258,
+      "learning_rate": 2e-05,
+      "loss": 0.03703724,
+      "step": 9230
+    },
+    {
+      "epoch": 18.462,
+      "grad_norm": 1.308974027633667,
+      "learning_rate": 2e-05,
+      "loss": 0.04389992,
+      "step": 9231
+    },
+    {
+      "epoch": 18.464,
+      "grad_norm": 0.7866896390914917,
+      "learning_rate": 2e-05,
+      "loss": 0.0197884,
+      "step": 9232
+    },
+    {
+      "epoch": 18.466,
+      "grad_norm": 1.48411226272583,
+      "learning_rate": 2e-05,
+      "loss": 0.03440548,
+      "step": 9233
+    },
+    {
+      "epoch": 18.468,
+      "grad_norm": 1.2252159118652344,
+      "learning_rate": 2e-05,
+      "loss": 0.03730087,
+      "step": 9234
+    },
+    {
+      "epoch": 18.47,
+      "grad_norm": 2.652885675430298,
+      "learning_rate": 2e-05,
+      "loss": 0.04888747,
+      "step": 9235
+    },
+    {
+      "epoch": 18.472,
+      "grad_norm": 0.9633825421333313,
+      "learning_rate": 2e-05,
+      "loss": 0.03162334,
+      "step": 9236
+    },
+    {
+      "epoch": 18.474,
+      "grad_norm": 1.207777500152588,
+      "learning_rate": 2e-05,
+      "loss": 0.03481243,
+      "step": 9237
+    },
+    {
+      "epoch": 18.476,
+      "grad_norm": 2.7908196449279785,
+      "learning_rate": 2e-05,
+      "loss": 0.06109419,
+      "step": 9238
+    },
+    {
+      "epoch": 18.478,
+      "grad_norm": 1.291577935218811,
+      "learning_rate": 2e-05,
+      "loss": 0.04738308,
+      "step": 9239
+    },
+    {
+      "epoch": 18.48,
+      "grad_norm": 1.2779804468154907,
+      "learning_rate": 2e-05,
+      "loss": 0.03293496,
+      "step": 9240
+    },
+    {
+      "epoch": 18.482,
+      "grad_norm": 1.388176679611206,
+      "learning_rate": 2e-05,
+      "loss": 0.03266312,
+      "step": 9241
+    },
+    {
+      "epoch": 18.484,
+      "grad_norm": 1.3886147737503052,
+      "learning_rate": 2e-05,
+      "loss": 0.02937507,
+      "step": 9242
+    },
+    {
+      "epoch": 18.486,
+      "grad_norm": 1.400505781173706,
+      "learning_rate": 2e-05,
+      "loss": 0.04522613,
+      "step": 9243
+    },
+    {
+      "epoch": 18.488,
+      "grad_norm": 1.2880887985229492,
+      "learning_rate": 2e-05,
+      "loss": 0.03753714,
+      "step": 9244
+    },
+    {
+      "epoch": 18.49,
+      "grad_norm": 1.0358575582504272,
+      "learning_rate": 2e-05,
+      "loss": 0.03539129,
+      "step": 9245
+    },
+    {
+      "epoch": 18.492,
+      "grad_norm": 3.62886905670166,
+      "learning_rate": 2e-05,
+      "loss": 0.03471145,
+      "step": 9246
+    },
+    {
+      "epoch": 18.494,
+      "grad_norm": 1.1280561685562134,
+      "learning_rate": 2e-05,
+      "loss": 0.03343772,
+      "step": 9247
+    },
+    {
+      "epoch": 18.496,
+      "grad_norm": 1.5975375175476074,
+      "learning_rate": 2e-05,
+      "loss": 0.0507987,
+      "step": 9248
+    },
+    {
+      "epoch": 18.498,
+      "grad_norm": 1.5075079202651978,
+      "learning_rate": 2e-05,
+      "loss": 0.05328429,
+      "step": 9249
+    },
+    {
+      "epoch": 18.5,
+      "grad_norm": 0.8867833614349365,
+      "learning_rate": 2e-05,
+      "loss": 0.02282133,
+      "step": 9250
+    },
+    {
+      "epoch": 18.502,
+      "grad_norm": 0.9173626899719238,
+      "learning_rate": 2e-05,
+      "loss": 0.02405676,
+      "step": 9251
+    },
+    {
+      "epoch": 18.504,
+      "grad_norm": 0.9300764799118042,
+      "learning_rate": 2e-05,
+      "loss": 0.03014572,
+      "step": 9252
+    },
+    {
+      "epoch": 18.506,
+      "grad_norm": 0.6862570643424988,
+      "learning_rate": 2e-05,
+      "loss": 0.01409816,
+      "step": 9253
+    },
+    {
+      "epoch": 18.508,
+      "grad_norm": 1.5550457239151,
+      "learning_rate": 2e-05,
+      "loss": 0.04886139,
+      "step": 9254
+    },
+    {
+      "epoch": 18.51,
+      "grad_norm": 1.2698708772659302,
+      "learning_rate": 2e-05,
+      "loss": 0.03988787,
+      "step": 9255
+    },
+    {
+      "epoch": 18.512,
+      "grad_norm": 1.4811738729476929,
+      "learning_rate": 2e-05,
+      "loss": 0.03565231,
+      "step": 9256
+    },
+    {
+      "epoch": 18.514,
+      "grad_norm": 2.2947709560394287,
+      "learning_rate": 2e-05,
+      "loss": 0.03940463,
+      "step": 9257
+    },
+    {
+      "epoch": 18.516,
+      "grad_norm": 1.0649924278259277,
+      "learning_rate": 2e-05,
+      "loss": 0.02794782,
+      "step": 9258
+    },
+    {
+      "epoch": 18.518,
+      "grad_norm": 1.2015745639801025,
+      "learning_rate": 2e-05,
+      "loss": 0.03615477,
+      "step": 9259
+    },
+    {
+      "epoch": 18.52,
+      "grad_norm": 1.1461511850357056,
+      "learning_rate": 2e-05,
+      "loss": 0.03559923,
+      "step": 9260
+    },
+    {
+      "epoch": 18.522,
+      "grad_norm": 1.8255705833435059,
+      "learning_rate": 2e-05,
+      "loss": 0.04908207,
+      "step": 9261
+    },
+    {
+      "epoch": 18.524,
+      "grad_norm": 1.0880002975463867,
+      "learning_rate": 2e-05,
+      "loss": 0.02855574,
+      "step": 9262
+    },
+    {
+      "epoch": 18.526,
+      "grad_norm": 2.4838008880615234,
+      "learning_rate": 2e-05,
+      "loss": 0.03444799,
+      "step": 9263
+    },
+    {
+      "epoch": 18.528,
+      "grad_norm": 1.1246633529663086,
+      "learning_rate": 2e-05,
+      "loss": 0.02719912,
+      "step": 9264
+    },
+    {
+      "epoch": 18.53,
+      "grad_norm": 0.9572923183441162,
+      "learning_rate": 2e-05,
+      "loss": 0.02941136,
+      "step": 9265
+    },
+    {
+      "epoch": 18.532,
+      "grad_norm": 1.3162039518356323,
+      "learning_rate": 2e-05,
+      "loss": 0.0248658,
+      "step": 9266
+    },
+    {
+      "epoch": 18.534,
+      "grad_norm": 1.1983442306518555,
+      "learning_rate": 2e-05,
+      "loss": 0.04123002,
+      "step": 9267
+    },
+    {
+      "epoch": 18.536,
+      "grad_norm": 1.4255578517913818,
+      "learning_rate": 2e-05,
+      "loss": 0.03819581,
+      "step": 9268
+    },
+    {
+      "epoch": 18.538,
+      "grad_norm": 1.2587826251983643,
+      "learning_rate": 2e-05,
+      "loss": 0.03482603,
+      "step": 9269
+    },
+    {
+      "epoch": 18.54,
+      "grad_norm": 1.2358354330062866,
+      "learning_rate": 2e-05,
+      "loss": 0.03826098,
+      "step": 9270
+    },
+    {
+      "epoch": 18.542,
+      "grad_norm": 1.1539580821990967,
+      "learning_rate": 2e-05,
+      "loss": 0.03745293,
+      "step": 9271
+    },
+    {
+      "epoch": 18.544,
+      "grad_norm": 1.8408349752426147,
+      "learning_rate": 2e-05,
+      "loss": 0.04294346,
+      "step": 9272
+    },
+    {
+      "epoch": 18.546,
+      "grad_norm": 2.7005791664123535,
+      "learning_rate": 2e-05,
+      "loss": 0.03372211,
+      "step": 9273
+    },
+    {
+      "epoch": 18.548000000000002,
+      "grad_norm": 2.7007863521575928,
+      "learning_rate": 2e-05,
+      "loss": 0.0421975,
+      "step": 9274
+    },
+    {
+      "epoch": 18.55,
+      "grad_norm": 1.045002818107605,
+      "learning_rate": 2e-05,
+      "loss": 0.03461481,
+      "step": 9275
+    },
+    {
+      "epoch": 18.552,
+      "grad_norm": 1.1427747011184692,
+      "learning_rate": 2e-05,
+      "loss": 0.0413534,
+      "step": 9276
+    },
+    {
+      "epoch": 18.554,
+      "grad_norm": 1.6876646280288696,
+      "learning_rate": 2e-05,
+      "loss": 0.04642463,
+      "step": 9277
+    },
+    {
+      "epoch": 18.556,
+      "grad_norm": 0.9396681189537048,
+      "learning_rate": 2e-05,
+      "loss": 0.02529829,
+      "step": 9278
+    },
+    {
+      "epoch": 18.558,
+      "grad_norm": 1.2221046686172485,
+      "learning_rate": 2e-05,
+      "loss": 0.03684171,
+      "step": 9279
+    },
+    {
+      "epoch": 18.56,
+      "grad_norm": 1.505953073501587,
+      "learning_rate": 2e-05,
+      "loss": 0.04012023,
+      "step": 9280
+    },
+    {
+      "epoch": 18.562,
+      "grad_norm": 1.1318151950836182,
+      "learning_rate": 2e-05,
+      "loss": 0.03027987,
+      "step": 9281
+    },
+    {
+      "epoch": 18.564,
+      "grad_norm": 2.6939613819122314,
+      "learning_rate": 2e-05,
+      "loss": 0.0387833,
+      "step": 9282
+    },
+    {
+      "epoch": 18.566,
+      "grad_norm": 1.3286248445510864,
+      "learning_rate": 2e-05,
+      "loss": 0.03856936,
+      "step": 9283
+    },
+    {
+      "epoch": 18.568,
+      "grad_norm": 2.2467446327209473,
+      "learning_rate": 2e-05,
+      "loss": 0.03416197,
+      "step": 9284
+    },
+    {
+      "epoch": 18.57,
+      "grad_norm": 0.8738142848014832,
+      "learning_rate": 2e-05,
+      "loss": 0.030899,
+      "step": 9285
+    },
+    {
+      "epoch": 18.572,
+      "grad_norm": 1.4438823461532593,
+      "learning_rate": 2e-05,
+      "loss": 0.03023734,
+      "step": 9286
+    },
+    {
+      "epoch": 18.574,
+      "grad_norm": 1.0707905292510986,
+      "learning_rate": 2e-05,
+      "loss": 0.03663265,
+      "step": 9287
+    },
+    {
+      "epoch": 18.576,
+      "grad_norm": 3.671177864074707,
+      "learning_rate": 2e-05,
+      "loss": 0.03615709,
+      "step": 9288
+    },
+    {
+      "epoch": 18.578,
+      "grad_norm": 1.270842432975769,
+      "learning_rate": 2e-05,
+      "loss": 0.03984865,
+      "step": 9289
+    },
+    {
+      "epoch": 18.58,
+      "grad_norm": 1.2456284761428833,
+      "learning_rate": 2e-05,
+      "loss": 0.03165886,
+      "step": 9290
+    },
+    {
+      "epoch": 18.582,
+      "grad_norm": 1.7004910707473755,
+      "learning_rate": 2e-05,
+      "loss": 0.0358137,
+      "step": 9291
+    },
+    {
+      "epoch": 18.584,
+      "grad_norm": 1.0085270404815674,
+      "learning_rate": 2e-05,
+      "loss": 0.03215361,
+      "step": 9292
+    },
+    {
+      "epoch": 18.586,
+      "grad_norm": 1.081712245941162,
+      "learning_rate": 2e-05,
+      "loss": 0.02386444,
+      "step": 9293
+    },
+    {
+      "epoch": 18.588,
+      "grad_norm": 1.1721680164337158,
+      "learning_rate": 2e-05,
+      "loss": 0.02527738,
+      "step": 9294
+    },
+    {
+      "epoch": 18.59,
+      "grad_norm": 0.7640558481216431,
+      "learning_rate": 2e-05,
+      "loss": 0.02131698,
+      "step": 9295
+    },
+    {
+      "epoch": 18.592,
+      "grad_norm": 1.278498888015747,
+      "learning_rate": 2e-05,
+      "loss": 0.03705659,
+      "step": 9296
+    },
+    {
+      "epoch": 18.594,
+      "grad_norm": 1.4128645658493042,
+      "learning_rate": 2e-05,
+      "loss": 0.02741608,
+      "step": 9297
+    },
+    {
+      "epoch": 18.596,
+      "grad_norm": 0.9089370369911194,
+      "learning_rate": 2e-05,
+      "loss": 0.02473187,
+      "step": 9298
+    },
+    {
+      "epoch": 18.598,
+      "grad_norm": 1.2044442892074585,
+      "learning_rate": 2e-05,
+      "loss": 0.03941024,
+      "step": 9299
+    },
+    {
+      "epoch": 18.6,
+      "grad_norm": 1.4648476839065552,
+      "learning_rate": 2e-05,
+      "loss": 0.02977973,
+      "step": 9300
+    },
+    {
+      "epoch": 18.602,
+      "grad_norm": 2.9288134574890137,
+      "learning_rate": 2e-05,
+      "loss": 0.05101031,
+      "step": 9301
+    },
+    {
+      "epoch": 18.604,
+      "grad_norm": 1.2922507524490356,
+      "learning_rate": 2e-05,
+      "loss": 0.03375695,
+      "step": 9302
+    },
+    {
+      "epoch": 18.606,
+      "grad_norm": 3.628523588180542,
+      "learning_rate": 2e-05,
+      "loss": 0.03046746,
+      "step": 9303
+    },
+    {
+      "epoch": 18.608,
+      "grad_norm": 1.5554486513137817,
+      "learning_rate": 2e-05,
+      "loss": 0.03474073,
+      "step": 9304
+    },
+    {
+      "epoch": 18.61,
+      "grad_norm": 1.2011704444885254,
+      "learning_rate": 2e-05,
+      "loss": 0.03348228,
+      "step": 9305
+    },
+    {
+      "epoch": 18.612,
+      "grad_norm": 1.0347377061843872,
+      "learning_rate": 2e-05,
+      "loss": 0.03440222,
+      "step": 9306
+    },
+    {
+      "epoch": 18.614,
+      "grad_norm": 1.410341739654541,
+      "learning_rate": 2e-05,
+      "loss": 0.04505805,
+      "step": 9307
+    },
+    {
+      "epoch": 18.616,
+      "grad_norm": 1.230013370513916,
+      "learning_rate": 2e-05,
+      "loss": 0.04353497,
+      "step": 9308
+    },
+    {
+      "epoch": 18.618,
+      "grad_norm": 0.8737069964408875,
+      "learning_rate": 2e-05,
+      "loss": 0.02157835,
+      "step": 9309
+    },
+    {
+      "epoch": 18.62,
+      "grad_norm": 1.7111151218414307,
+      "learning_rate": 2e-05,
+      "loss": 0.0389607,
+      "step": 9310
+    },
+    {
+      "epoch": 18.622,
+      "grad_norm": 1.636101245880127,
+      "learning_rate": 2e-05,
+      "loss": 0.04191142,
+      "step": 9311
+    },
+    {
+      "epoch": 18.624,
+      "grad_norm": 2.3487284183502197,
+      "learning_rate": 2e-05,
+      "loss": 0.04457171,
+      "step": 9312
+    },
+    {
+      "epoch": 18.626,
+      "grad_norm": 1.1017884016036987,
+      "learning_rate": 2e-05,
+      "loss": 0.03230072,
+      "step": 9313
+    },
+    {
+      "epoch": 18.628,
+      "grad_norm": 1.428139567375183,
+      "learning_rate": 2e-05,
+      "loss": 0.03711209,
+      "step": 9314
+    },
+    {
+      "epoch": 18.63,
+      "grad_norm": 1.0108582973480225,
+      "learning_rate": 2e-05,
+      "loss": 0.03145947,
+      "step": 9315
+    },
+    {
+      "epoch": 18.632,
+      "grad_norm": 1.5939596891403198,
+      "learning_rate": 2e-05,
+      "loss": 0.03615252,
+      "step": 9316
+    },
+    {
+      "epoch": 18.634,
+      "grad_norm": 1.2237117290496826,
+      "learning_rate": 2e-05,
+      "loss": 0.03127633,
+      "step": 9317
+    },
+    {
+      "epoch": 18.636,
+      "grad_norm": 0.8785711526870728,
+      "learning_rate": 2e-05,
+      "loss": 0.01814137,
+      "step": 9318
+    },
+    {
+      "epoch": 18.638,
+      "grad_norm": 1.659043550491333,
+      "learning_rate": 2e-05,
+      "loss": 0.03371783,
+      "step": 9319
+    },
+    {
+      "epoch": 18.64,
+      "grad_norm": 1.5230258703231812,
+      "learning_rate": 2e-05,
+      "loss": 0.04253473,
+      "step": 9320
+    },
+    {
+      "epoch": 18.642,
+      "grad_norm": 1.0703048706054688,
+      "learning_rate": 2e-05,
+      "loss": 0.02780952,
+      "step": 9321
+    },
+    {
+      "epoch": 18.644,
+      "grad_norm": 2.55739688873291,
+      "learning_rate": 2e-05,
+      "loss": 0.05347919,
+      "step": 9322
+    },
+    {
+      "epoch": 18.646,
+      "grad_norm": 1.2097032070159912,
+      "learning_rate": 2e-05,
+      "loss": 0.03133424,
+      "step": 9323
+    },
+    {
+      "epoch": 18.648,
+      "grad_norm": 1.7412601709365845,
+      "learning_rate": 2e-05,
+      "loss": 0.04346561,
+      "step": 9324
+    },
+    {
+      "epoch": 18.65,
+      "grad_norm": 0.9796632528305054,
+      "learning_rate": 2e-05,
+      "loss": 0.02903066,
+      "step": 9325
+    },
+    {
+      "epoch": 18.652,
+      "grad_norm": 1.5429102182388306,
+      "learning_rate": 2e-05,
+      "loss": 0.03083556,
+      "step": 9326
+    },
+    {
+      "epoch": 18.654,
+      "grad_norm": 3.025692939758301,
+      "learning_rate": 2e-05,
+      "loss": 0.04267042,
+      "step": 9327
+    },
+    {
+      "epoch": 18.656,
+      "grad_norm": 1.8448047637939453,
+      "learning_rate": 2e-05,
+      "loss": 0.04758323,
+      "step": 9328
+    },
+    {
+      "epoch": 18.658,
+      "grad_norm": 1.5989882946014404,
+      "learning_rate": 2e-05,
+      "loss": 0.04394848,
+      "step": 9329
+    },
+    {
+      "epoch": 18.66,
+      "grad_norm": 1.270090103149414,
+      "learning_rate": 2e-05,
+      "loss": 0.03506234,
+      "step": 9330
+    },
+    {
+      "epoch": 18.662,
+      "grad_norm": 1.0124268531799316,
+      "learning_rate": 2e-05,
+      "loss": 0.02979084,
+      "step": 9331
+    },
+    {
+      "epoch": 18.664,
+      "grad_norm": 1.4778650999069214,
+      "learning_rate": 2e-05,
+      "loss": 0.03793935,
+      "step": 9332
+    },
+    {
+      "epoch": 18.666,
+      "grad_norm": 1.5023845434188843,
+      "learning_rate": 2e-05,
+      "loss": 0.0416206,
+      "step": 9333
+    },
+    {
+      "epoch": 18.668,
+      "grad_norm": 0.8940832614898682,
+      "learning_rate": 2e-05,
+      "loss": 0.01142349,
+      "step": 9334
+    },
+    {
+      "epoch": 18.67,
+      "grad_norm": 1.727463960647583,
+      "learning_rate": 2e-05,
+      "loss": 0.03897662,
+      "step": 9335
+    },
+    {
+      "epoch": 18.672,
+      "grad_norm": 1.1958836317062378,
+      "learning_rate": 2e-05,
+      "loss": 0.04261857,
+      "step": 9336
+    },
+    {
+      "epoch": 18.674,
+      "grad_norm": 1.020642638206482,
+      "learning_rate": 2e-05,
+      "loss": 0.02642402,
+      "step": 9337
+    },
+    {
+      "epoch": 18.676,
+      "grad_norm": 2.0219318866729736,
+      "learning_rate": 2e-05,
+      "loss": 0.0680473,
+      "step": 9338
+    },
+    {
+      "epoch": 18.678,
+      "grad_norm": 1.318962574005127,
+      "learning_rate": 2e-05,
+      "loss": 0.02908128,
+      "step": 9339
+    },
+    {
+      "epoch": 18.68,
+      "grad_norm": 1.432193636894226,
+      "learning_rate": 2e-05,
+      "loss": 0.03741326,
+      "step": 9340
+    },
+    {
+      "epoch": 18.682,
+      "grad_norm": 1.0566816329956055,
+      "learning_rate": 2e-05,
+      "loss": 0.03124825,
+      "step": 9341
+    },
+    {
+      "epoch": 18.684,
+      "grad_norm": 1.3076931238174438,
+      "learning_rate": 2e-05,
+      "loss": 0.04089387,
+      "step": 9342
+    },
+    {
+      "epoch": 18.686,
+      "grad_norm": 1.0010727643966675,
+      "learning_rate": 2e-05,
+      "loss": 0.03195993,
+      "step": 9343
+    },
+    {
+      "epoch": 18.688,
+      "grad_norm": 1.1595393419265747,
+      "learning_rate": 2e-05,
+      "loss": 0.03774239,
+      "step": 9344
+    },
+    {
+      "epoch": 18.69,
+      "grad_norm": 0.8064015507698059,
+      "learning_rate": 2e-05,
+      "loss": 0.02562628,
+      "step": 9345
+    },
+    {
+      "epoch": 18.692,
+      "grad_norm": 1.1677731275558472,
+      "learning_rate": 2e-05,
+      "loss": 0.02810512,
+      "step": 9346
+    },
+    {
+      "epoch": 18.694,
+      "grad_norm": 2.073591709136963,
+      "learning_rate": 2e-05,
+      "loss": 0.04026121,
+      "step": 9347
+    },
+    {
+      "epoch": 18.696,
+      "grad_norm": 0.8951675891876221,
+      "learning_rate": 2e-05,
+      "loss": 0.02346078,
+      "step": 9348
+    },
+    {
+      "epoch": 18.698,
+      "grad_norm": 1.4167735576629639,
+      "learning_rate": 2e-05,
+      "loss": 0.04481205,
+      "step": 9349
+    },
+    {
+      "epoch": 18.7,
+      "grad_norm": 1.254514217376709,
+      "learning_rate": 2e-05,
+      "loss": 0.02620357,
+      "step": 9350
+    },
+    {
+      "epoch": 18.701999999999998,
+      "grad_norm": 2.4272637367248535,
+      "learning_rate": 2e-05,
+      "loss": 0.04456335,
+      "step": 9351
+    },
+    {
+      "epoch": 18.704,
+      "grad_norm": 1.1000620126724243,
+      "learning_rate": 2e-05,
+      "loss": 0.04522509,
+      "step": 9352
+    },
+    {
+      "epoch": 18.706,
+      "grad_norm": 1.0736738443374634,
+      "learning_rate": 2e-05,
+      "loss": 0.04045687,
+      "step": 9353
+    },
+    {
+      "epoch": 18.708,
+      "grad_norm": 2.3483619689941406,
+      "learning_rate": 2e-05,
+      "loss": 0.05076428,
+      "step": 9354
+    },
+    {
+      "epoch": 18.71,
+      "grad_norm": 0.972253680229187,
+      "learning_rate": 2e-05,
+      "loss": 0.02645556,
+      "step": 9355
+    },
+    {
+      "epoch": 18.712,
+      "grad_norm": 0.9229649305343628,
+      "learning_rate": 2e-05,
+      "loss": 0.02200065,
+      "step": 9356
+    },
+    {
+      "epoch": 18.714,
+      "grad_norm": 1.2092173099517822,
+      "learning_rate": 2e-05,
+      "loss": 0.03871221,
+      "step": 9357
+    },
+    {
+      "epoch": 18.716,
+      "grad_norm": 1.9008631706237793,
+      "learning_rate": 2e-05,
+      "loss": 0.04458017,
+      "step": 9358
+    },
+    {
+      "epoch": 18.718,
+      "grad_norm": 1.0636965036392212,
+      "learning_rate": 2e-05,
+      "loss": 0.02840264,
+      "step": 9359
+    },
+    {
+      "epoch": 18.72,
+      "grad_norm": 1.2430663108825684,
+      "learning_rate": 2e-05,
+      "loss": 0.03148837,
+      "step": 9360
+    },
+    {
+      "epoch": 18.722,
+      "grad_norm": 1.123664140701294,
+      "learning_rate": 2e-05,
+      "loss": 0.03394098,
+      "step": 9361
+    },
+    {
+      "epoch": 18.724,
+      "grad_norm": 1.0054221153259277,
+      "learning_rate": 2e-05,
+      "loss": 0.03370092,
+      "step": 9362
+    },
+    {
+      "epoch": 18.726,
+      "grad_norm": 1.5903475284576416,
+      "learning_rate": 2e-05,
+      "loss": 0.04029074,
+      "step": 9363
+    },
+    {
+      "epoch": 18.728,
+      "grad_norm": 1.9477814435958862,
+      "learning_rate": 2e-05,
+      "loss": 0.06019651,
+      "step": 9364
+    },
+    {
+      "epoch": 18.73,
+      "grad_norm": 2.084622383117676,
+      "learning_rate": 2e-05,
+      "loss": 0.02988496,
+      "step": 9365
+    },
+    {
+      "epoch": 18.732,
+      "grad_norm": 1.7150975465774536,
+      "learning_rate": 2e-05,
+      "loss": 0.05392132,
+      "step": 9366
+    },
+    {
+      "epoch": 18.734,
+      "grad_norm": 0.9419837594032288,
+      "learning_rate": 2e-05,
+      "loss": 0.02924297,
+      "step": 9367
+    },
+    {
+      "epoch": 18.736,
+      "grad_norm": 1.4422987699508667,
+      "learning_rate": 2e-05,
+      "loss": 0.03685901,
+      "step": 9368
+    },
+    {
+      "epoch": 18.738,
+      "grad_norm": 0.9840115904808044,
+      "learning_rate": 2e-05,
+      "loss": 0.02506851,
+      "step": 9369
+    },
+    {
+      "epoch": 18.74,
+      "grad_norm": 1.052154302597046,
+      "learning_rate": 2e-05,
+      "loss": 0.03087404,
+      "step": 9370
+    },
+    {
+      "epoch": 18.742,
+      "grad_norm": 1.2569382190704346,
+      "learning_rate": 2e-05,
+      "loss": 0.04886915,
+      "step": 9371
+    },
+    {
+      "epoch": 18.744,
+      "grad_norm": 1.5744709968566895,
+      "learning_rate": 2e-05,
+      "loss": 0.04542425,
+      "step": 9372
+    },
+    {
+      "epoch": 18.746,
+      "grad_norm": 2.963470458984375,
+      "learning_rate": 2e-05,
+      "loss": 0.03224288,
+      "step": 9373
+    },
+    {
+      "epoch": 18.748,
+      "grad_norm": 1.8991296291351318,
+      "learning_rate": 2e-05,
+      "loss": 0.046722,
+      "step": 9374
+    },
+    {
+      "epoch": 18.75,
+      "grad_norm": 0.9656346440315247,
+      "learning_rate": 2e-05,
+      "loss": 0.03440123,
+      "step": 9375
+    },
+    {
+      "epoch": 18.752,
+      "grad_norm": 1.0450048446655273,
+      "learning_rate": 2e-05,
+      "loss": 0.03775831,
+      "step": 9376
+    },
+    {
+      "epoch": 18.754,
+      "grad_norm": 1.7124624252319336,
+      "learning_rate": 2e-05,
+      "loss": 0.05066157,
+      "step": 9377
+    },
+    {
+      "epoch": 18.756,
+      "grad_norm": 1.501320242881775,
+      "learning_rate": 2e-05,
+      "loss": 0.03876662,
+      "step": 9378
+    },
+    {
+      "epoch": 18.758,
+      "grad_norm": 1.2065678834915161,
+      "learning_rate": 2e-05,
+      "loss": 0.03976551,
+      "step": 9379
+    },
+    {
+      "epoch": 18.76,
+      "grad_norm": 1.4827100038528442,
+      "learning_rate": 2e-05,
+      "loss": 0.03961009,
+      "step": 9380
+    },
+    {
+      "epoch": 18.762,
+      "grad_norm": 1.3919841051101685,
+      "learning_rate": 2e-05,
+      "loss": 0.0386041,
+      "step": 9381
+    },
+    {
+      "epoch": 18.764,
+      "grad_norm": 1.1867378950119019,
+      "learning_rate": 2e-05,
+      "loss": 0.03478377,
+      "step": 9382
+    },
+    {
+      "epoch": 18.766,
+      "grad_norm": 1.0938994884490967,
+      "learning_rate": 2e-05,
+      "loss": 0.03122077,
+      "step": 9383
+    },
+    {
+      "epoch": 18.768,
+      "grad_norm": 2.3887717723846436,
+      "learning_rate": 2e-05,
+      "loss": 0.04409286,
+      "step": 9384
+    },
+    {
+      "epoch": 18.77,
+      "grad_norm": 1.0273598432540894,
+      "learning_rate": 2e-05,
+      "loss": 0.02485752,
+      "step": 9385
+    },
+    {
+      "epoch": 18.772,
+      "grad_norm": 1.7304757833480835,
+      "learning_rate": 2e-05,
+      "loss": 0.03049651,
+      "step": 9386
+    },
+    {
+      "epoch": 18.774,
+      "grad_norm": 1.9166194200515747,
+      "learning_rate": 2e-05,
+      "loss": 0.04044746,
+      "step": 9387
+    },
+    {
+      "epoch": 18.776,
+      "grad_norm": 1.7198108434677124,
+      "learning_rate": 2e-05,
+      "loss": 0.0405063,
+      "step": 9388
+    },
+    {
+      "epoch": 18.778,
+      "grad_norm": 1.8082114458084106,
+      "learning_rate": 2e-05,
+      "loss": 0.03769893,
+      "step": 9389
+    },
+    {
+      "epoch": 18.78,
+      "grad_norm": 0.9580773115158081,
+      "learning_rate": 2e-05,
+      "loss": 0.03357112,
+      "step": 9390
+    },
+    {
+      "epoch": 18.782,
+      "grad_norm": 1.059820532798767,
+      "learning_rate": 2e-05,
+      "loss": 0.02640706,
+      "step": 9391
+    },
+    {
+      "epoch": 18.784,
+      "grad_norm": 4.033216953277588,
+      "learning_rate": 2e-05,
+      "loss": 0.03864349,
+      "step": 9392
+    },
+    {
+      "epoch": 18.786,
+      "grad_norm": 1.1333658695220947,
+      "learning_rate": 2e-05,
+      "loss": 0.03008582,
+      "step": 9393
+    },
+    {
+      "epoch": 18.788,
+      "grad_norm": 2.1095573902130127,
+      "learning_rate": 2e-05,
+      "loss": 0.03836686,
+      "step": 9394
+    },
+    {
+      "epoch": 18.79,
+      "grad_norm": 0.7854803204536438,
+      "learning_rate": 2e-05,
+      "loss": 0.01916845,
+      "step": 9395
+    },
+    {
+      "epoch": 18.792,
+      "grad_norm": 1.424535870552063,
+      "learning_rate": 2e-05,
+      "loss": 0.03236981,
+      "step": 9396
+    },
+    {
+      "epoch": 18.794,
+      "grad_norm": 1.234863042831421,
+      "learning_rate": 2e-05,
+      "loss": 0.0439956,
+      "step": 9397
+    },
+    {
+      "epoch": 18.796,
+      "grad_norm": 1.0828542709350586,
+      "learning_rate": 2e-05,
+      "loss": 0.03070717,
+      "step": 9398
+    },
+    {
+      "epoch": 18.798000000000002,
+      "grad_norm": 1.125207781791687,
+      "learning_rate": 2e-05,
+      "loss": 0.03891785,
+      "step": 9399
+    },
+    {
+      "epoch": 18.8,
+      "grad_norm": 1.4716780185699463,
+      "learning_rate": 2e-05,
+      "loss": 0.04283679,
+      "step": 9400
+    },
+    {
+      "epoch": 18.802,
+      "grad_norm": 2.2038402557373047,
+      "learning_rate": 2e-05,
+      "loss": 0.04184093,
+      "step": 9401
+    },
+    {
+      "epoch": 18.804,
+      "grad_norm": 2.594099283218384,
+      "learning_rate": 2e-05,
+      "loss": 0.0254508,
+      "step": 9402
+    },
+    {
+      "epoch": 18.806,
+      "grad_norm": 0.977802038192749,
+      "learning_rate": 2e-05,
+      "loss": 0.02622373,
+      "step": 9403
+    },
+    {
+      "epoch": 18.808,
+      "grad_norm": 1.5012508630752563,
+      "learning_rate": 2e-05,
+      "loss": 0.05293182,
+      "step": 9404
+    },
+    {
+      "epoch": 18.81,
+      "grad_norm": 1.1779042482376099,
+      "learning_rate": 2e-05,
+      "loss": 0.03067445,
+      "step": 9405
+    },
+    {
+      "epoch": 18.812,
+      "grad_norm": 1.9061391353607178,
+      "learning_rate": 2e-05,
+      "loss": 0.04364157,
+      "step": 9406
+    },
+    {
+      "epoch": 18.814,
+      "grad_norm": 1.0848644971847534,
+      "learning_rate": 2e-05,
+      "loss": 0.03666593,
+      "step": 9407
+    },
+    {
+      "epoch": 18.816,
+      "grad_norm": 4.1378960609436035,
+      "learning_rate": 2e-05,
+      "loss": 0.03240601,
+      "step": 9408
+    },
+    {
+      "epoch": 18.818,
+      "grad_norm": 1.168457269668579,
+      "learning_rate": 2e-05,
+      "loss": 0.03614372,
+      "step": 9409
+    },
+    {
+      "epoch": 18.82,
+      "grad_norm": 1.1519936323165894,
+      "learning_rate": 2e-05,
+      "loss": 0.04695103,
+      "step": 9410
+    },
+    {
+      "epoch": 18.822,
+      "grad_norm": 1.9238035678863525,
+      "learning_rate": 2e-05,
+      "loss": 0.04975787,
+      "step": 9411
+    },
+    {
+      "epoch": 18.824,
+      "grad_norm": 1.304675579071045,
+      "learning_rate": 2e-05,
+      "loss": 0.04074424,
+      "step": 9412
+    },
+    {
+      "epoch": 18.826,
+      "grad_norm": 2.158198833465576,
+      "learning_rate": 2e-05,
+      "loss": 0.0502829,
+      "step": 9413
+    },
+    {
+      "epoch": 18.828,
+      "grad_norm": 1.3099114894866943,
+      "learning_rate": 2e-05,
+      "loss": 0.04483989,
+      "step": 9414
+    },
+    {
+      "epoch": 18.83,
+      "grad_norm": 1.9079896211624146,
+      "learning_rate": 2e-05,
+      "loss": 0.04286634,
+      "step": 9415
+    },
+    {
+      "epoch": 18.832,
+      "grad_norm": 1.7033705711364746,
+      "learning_rate": 2e-05,
+      "loss": 0.03799369,
+      "step": 9416
+    },
+    {
+      "epoch": 18.834,
+      "grad_norm": 0.991951584815979,
+      "learning_rate": 2e-05,
+      "loss": 0.03253823,
+      "step": 9417
+    },
+    {
+      "epoch": 18.836,
+      "grad_norm": 1.4874738454818726,
+      "learning_rate": 2e-05,
+      "loss": 0.05053546,
+      "step": 9418
+    },
+    {
+      "epoch": 18.838,
+      "grad_norm": 1.4518730640411377,
+      "learning_rate": 2e-05,
+      "loss": 0.03553774,
+      "step": 9419
+    },
+    {
+      "epoch": 18.84,
+      "grad_norm": 1.0209287405014038,
+      "learning_rate": 2e-05,
+      "loss": 0.03443102,
+      "step": 9420
+    },
+    {
+      "epoch": 18.842,
+      "grad_norm": 1.0967309474945068,
+      "learning_rate": 2e-05,
+      "loss": 0.02932144,
+      "step": 9421
+    },
+    {
+      "epoch": 18.844,
+      "grad_norm": 2.3560574054718018,
+      "learning_rate": 2e-05,
+      "loss": 0.04702324,
+      "step": 9422
+    },
+    {
+      "epoch": 18.846,
+      "grad_norm": 1.2899200916290283,
+      "learning_rate": 2e-05,
+      "loss": 0.0305502,
+      "step": 9423
+    },
+    {
+      "epoch": 18.848,
+      "grad_norm": 1.8223811388015747,
+      "learning_rate": 2e-05,
+      "loss": 0.05424388,
+      "step": 9424
+    },
+    {
+      "epoch": 18.85,
+      "grad_norm": 1.6738855838775635,
+      "learning_rate": 2e-05,
+      "loss": 0.04101273,
+      "step": 9425
+    },
+    {
+      "epoch": 18.852,
+      "grad_norm": 1.2631632089614868,
+      "learning_rate": 2e-05,
+      "loss": 0.03289703,
+      "step": 9426
+    },
+    {
+      "epoch": 18.854,
+      "grad_norm": 1.1822715997695923,
+      "learning_rate": 2e-05,
+      "loss": 0.03998087,
+      "step": 9427
+    },
+    {
+      "epoch": 18.856,
+      "grad_norm": 1.7935429811477661,
+      "learning_rate": 2e-05,
+      "loss": 0.04708499,
+      "step": 9428
+    },
+    {
+      "epoch": 18.858,
+      "grad_norm": 1.6989531517028809,
+      "learning_rate": 2e-05,
+      "loss": 0.03702974,
+      "step": 9429
+    },
+    {
+      "epoch": 18.86,
+      "grad_norm": 1.411005973815918,
+      "learning_rate": 2e-05,
+      "loss": 0.03301138,
+      "step": 9430
+    },
+    {
+      "epoch": 18.862,
+      "grad_norm": 1.575526475906372,
+      "learning_rate": 2e-05,
+      "loss": 0.0518712,
+      "step": 9431
+    },
+    {
+      "epoch": 18.864,
+      "grad_norm": 0.8966975808143616,
+      "learning_rate": 2e-05,
+      "loss": 0.02268358,
+      "step": 9432
+    },
+    {
+      "epoch": 18.866,
+      "grad_norm": 1.3550505638122559,
+      "learning_rate": 2e-05,
+      "loss": 0.03749962,
+      "step": 9433
+    },
+    {
+      "epoch": 18.868,
+      "grad_norm": 1.207885980606079,
+      "learning_rate": 2e-05,
+      "loss": 0.0400594,
+      "step": 9434
+    },
+    {
+      "epoch": 18.87,
+      "grad_norm": 1.0403681993484497,
+      "learning_rate": 2e-05,
+      "loss": 0.0228872,
+      "step": 9435
+    },
+    {
+      "epoch": 18.872,
+      "grad_norm": 2.362938404083252,
+      "learning_rate": 2e-05,
+      "loss": 0.06088825,
+      "step": 9436
+    },
+    {
+      "epoch": 18.874,
+      "grad_norm": 0.9768615365028381,
+      "learning_rate": 2e-05,
+      "loss": 0.03623533,
+      "step": 9437
+    },
+    {
+      "epoch": 18.876,
+      "grad_norm": 0.7810379862785339,
+      "learning_rate": 2e-05,
+      "loss": 0.01865301,
+      "step": 9438
+    },
+    {
+      "epoch": 18.878,
+      "grad_norm": 1.4565064907073975,
+      "learning_rate": 2e-05,
+      "loss": 0.04857381,
+      "step": 9439
+    },
+    {
+      "epoch": 18.88,
+      "grad_norm": 1.5316739082336426,
+      "learning_rate": 2e-05,
+      "loss": 0.04568139,
+      "step": 9440
+    },
+    {
+      "epoch": 18.882,
+      "grad_norm": 1.5420665740966797,
+      "learning_rate": 2e-05,
+      "loss": 0.03343572,
+      "step": 9441
+    },
+    {
+      "epoch": 18.884,
+      "grad_norm": 1.7624716758728027,
+      "learning_rate": 2e-05,
+      "loss": 0.03562377,
+      "step": 9442
+    },
+    {
+      "epoch": 18.886,
+      "grad_norm": 1.3358241319656372,
+      "learning_rate": 2e-05,
+      "loss": 0.06009006,
+      "step": 9443
+    },
+    {
+      "epoch": 18.888,
+      "grad_norm": 1.3039106130599976,
+      "learning_rate": 2e-05,
+      "loss": 0.04013005,
+      "step": 9444
+    },
+    {
+      "epoch": 18.89,
+      "grad_norm": 0.8214094638824463,
+      "learning_rate": 2e-05,
+      "loss": 0.02453827,
+      "step": 9445
+    },
+    {
+      "epoch": 18.892,
+      "grad_norm": 2.3487486839294434,
+      "learning_rate": 2e-05,
+      "loss": 0.04817913,
+      "step": 9446
+    },
+    {
+      "epoch": 18.894,
+      "grad_norm": 2.1647088527679443,
+      "learning_rate": 2e-05,
+      "loss": 0.03035383,
+      "step": 9447
+    },
+    {
+      "epoch": 18.896,
+      "grad_norm": 1.254131555557251,
+      "learning_rate": 2e-05,
+      "loss": 0.03613605,
+      "step": 9448
+    },
+    {
+      "epoch": 18.898,
+      "grad_norm": 1.1320441961288452,
+      "learning_rate": 2e-05,
+      "loss": 0.03620258,
+      "step": 9449
+    },
+    {
+      "epoch": 18.9,
+      "grad_norm": 1.6133835315704346,
+      "learning_rate": 2e-05,
+      "loss": 0.03855583,
+      "step": 9450
+    },
+    {
+      "epoch": 18.902,
+      "grad_norm": 0.9080222845077515,
+      "learning_rate": 2e-05,
+      "loss": 0.02581986,
+      "step": 9451
+    },
+    {
+      "epoch": 18.904,
+      "grad_norm": 1.1648834943771362,
+      "learning_rate": 2e-05,
+      "loss": 0.04252192,
+      "step": 9452
+    },
+    {
+      "epoch": 18.906,
+      "grad_norm": 1.1894398927688599,
+      "learning_rate": 2e-05,
+      "loss": 0.03881192,
+      "step": 9453
+    },
+    {
+      "epoch": 18.908,
+      "grad_norm": 2.000978469848633,
+      "learning_rate": 2e-05,
+      "loss": 0.03810741,
+      "step": 9454
+    },
+    {
+      "epoch": 18.91,
+      "grad_norm": 1.288516879081726,
+      "learning_rate": 2e-05,
+      "loss": 0.0437531,
+      "step": 9455
+    },
+    {
+      "epoch": 18.912,
+      "grad_norm": 1.3083631992340088,
+      "learning_rate": 2e-05,
+      "loss": 0.03637118,
+      "step": 9456
+    },
+    {
+      "epoch": 18.914,
+      "grad_norm": 1.1857115030288696,
+      "learning_rate": 2e-05,
+      "loss": 0.04493545,
+      "step": 9457
+    },
+    {
+      "epoch": 18.916,
+      "grad_norm": 1.9343324899673462,
+      "learning_rate": 2e-05,
+      "loss": 0.04948426,
+      "step": 9458
+    },
+    {
+      "epoch": 18.918,
+      "grad_norm": 1.187826156616211,
+      "learning_rate": 2e-05,
+      "loss": 0.03746122,
+      "step": 9459
+    },
+    {
+      "epoch": 18.92,
+      "grad_norm": 2.1284165382385254,
+      "learning_rate": 2e-05,
+      "loss": 0.04689957,
+      "step": 9460
+    },
+    {
+      "epoch": 18.922,
+      "grad_norm": 1.2563556432724,
+      "learning_rate": 2e-05,
+      "loss": 0.03775541,
+      "step": 9461
+    },
+    {
+      "epoch": 18.924,
+      "grad_norm": 1.2920955419540405,
+      "learning_rate": 2e-05,
+      "loss": 0.02915121,
+      "step": 9462
+    },
+    {
+      "epoch": 18.926,
+      "grad_norm": 1.910418152809143,
+      "learning_rate": 2e-05,
+      "loss": 0.04141461,
+      "step": 9463
+    },
+    {
+      "epoch": 18.928,
+      "grad_norm": 0.9791347980499268,
+      "learning_rate": 2e-05,
+      "loss": 0.03308304,
+      "step": 9464
+    },
+    {
+      "epoch": 18.93,
+      "grad_norm": 2.0668258666992188,
+      "learning_rate": 2e-05,
+      "loss": 0.04485352,
+      "step": 9465
+    },
+    {
+      "epoch": 18.932,
+      "grad_norm": 0.8224574327468872,
+      "learning_rate": 2e-05,
+      "loss": 0.02453919,
+      "step": 9466
+    },
+    {
+      "epoch": 18.934,
+      "grad_norm": 0.975167453289032,
+      "learning_rate": 2e-05,
+      "loss": 0.03143321,
+      "step": 9467
+    },
+    {
+      "epoch": 18.936,
+      "grad_norm": 1.0642677545547485,
+      "learning_rate": 2e-05,
+      "loss": 0.03415136,
+      "step": 9468
+    },
+    {
+      "epoch": 18.938,
+      "grad_norm": 0.8453977704048157,
+      "learning_rate": 2e-05,
+      "loss": 0.02789571,
+      "step": 9469
+    },
+    {
+      "epoch": 18.94,
+      "grad_norm": 2.6224365234375,
+      "learning_rate": 2e-05,
+      "loss": 0.03373963,
+      "step": 9470
+    },
+    {
+      "epoch": 18.942,
+      "grad_norm": 1.6013566255569458,
+      "learning_rate": 2e-05,
+      "loss": 0.02791347,
+      "step": 9471
+    },
+    {
+      "epoch": 18.944,
+      "grad_norm": 1.6802016496658325,
+      "learning_rate": 2e-05,
+      "loss": 0.04182333,
+      "step": 9472
+    },
+    {
+      "epoch": 18.946,
+      "grad_norm": 1.069710373878479,
+      "learning_rate": 2e-05,
+      "loss": 0.03297665,
+      "step": 9473
+    },
+    {
+      "epoch": 18.948,
+      "grad_norm": 1.6262844800949097,
+      "learning_rate": 2e-05,
+      "loss": 0.05678248,
+      "step": 9474
+    },
+    {
+      "epoch": 18.95,
+      "grad_norm": 1.6136468648910522,
+      "learning_rate": 2e-05,
+      "loss": 0.03860488,
+      "step": 9475
+    },
+    {
+      "epoch": 18.951999999999998,
+      "grad_norm": 1.6454910039901733,
+      "learning_rate": 2e-05,
+      "loss": 0.0360266,
+      "step": 9476
+    },
+    {
+      "epoch": 18.954,
+      "grad_norm": 1.3030211925506592,
+      "learning_rate": 2e-05,
+      "loss": 0.03737117,
+      "step": 9477
+    },
+    {
+      "epoch": 18.956,
+      "grad_norm": 1.0022649765014648,
+      "learning_rate": 2e-05,
+      "loss": 0.03304816,
+      "step": 9478
+    },
+    {
+      "epoch": 18.958,
+      "grad_norm": 2.1013195514678955,
+      "learning_rate": 2e-05,
+      "loss": 0.04374126,
+      "step": 9479
+    },
+    {
+      "epoch": 18.96,
+      "grad_norm": 0.9510260224342346,
+      "learning_rate": 2e-05,
+      "loss": 0.03056635,
+      "step": 9480
+    },
+    {
+      "epoch": 18.962,
+      "grad_norm": 1.7025002241134644,
+      "learning_rate": 2e-05,
+      "loss": 0.04591848,
+      "step": 9481
+    },
+    {
+      "epoch": 18.964,
+      "grad_norm": 2.784985303878784,
+      "learning_rate": 2e-05,
+      "loss": 0.04702611,
+      "step": 9482
+    },
+    {
+      "epoch": 18.966,
+      "grad_norm": 1.0321415662765503,
+      "learning_rate": 2e-05,
+      "loss": 0.03965008,
+      "step": 9483
+    },
+    {
+      "epoch": 18.968,
+      "grad_norm": 1.3998851776123047,
+      "learning_rate": 2e-05,
+      "loss": 0.04656575,
+      "step": 9484
+    },
+    {
+      "epoch": 18.97,
+      "grad_norm": 1.053924322128296,
+      "learning_rate": 2e-05,
+      "loss": 0.04150043,
+      "step": 9485
+    },
+    {
+      "epoch": 18.972,
+      "grad_norm": 1.0211212635040283,
+      "learning_rate": 2e-05,
+      "loss": 0.03201551,
+      "step": 9486
+    },
+    {
+      "epoch": 18.974,
+      "grad_norm": 1.109754204750061,
+      "learning_rate": 2e-05,
+      "loss": 0.03495897,
+      "step": 9487
+    },
+    {
+      "epoch": 18.976,
+      "grad_norm": 0.9835900664329529,
+      "learning_rate": 2e-05,
+      "loss": 0.03508302,
+      "step": 9488
+    },
+    {
+      "epoch": 18.978,
+      "grad_norm": 1.2372853755950928,
+      "learning_rate": 2e-05,
+      "loss": 0.02763392,
+      "step": 9489
+    },
+    {
+      "epoch": 18.98,
+      "grad_norm": 1.1497222185134888,
+      "learning_rate": 2e-05,
+      "loss": 0.04323982,
+      "step": 9490
+    },
+    {
+      "epoch": 18.982,
+      "grad_norm": 0.9017625451087952,
+      "learning_rate": 2e-05,
+      "loss": 0.03254882,
+      "step": 9491
+    },
+    {
+      "epoch": 18.984,
+      "grad_norm": 1.0615708827972412,
+      "learning_rate": 2e-05,
+      "loss": 0.02894014,
+      "step": 9492
+    },
+    {
+      "epoch": 18.986,
+      "grad_norm": 1.5303292274475098,
+      "learning_rate": 2e-05,
+      "loss": 0.0392815,
+      "step": 9493
+    },
+    {
+      "epoch": 18.988,
+      "grad_norm": 1.4052002429962158,
+      "learning_rate": 2e-05,
+      "loss": 0.0530249,
+      "step": 9494
+    },
+    {
+      "epoch": 18.99,
+      "grad_norm": 1.3222413063049316,
+      "learning_rate": 2e-05,
+      "loss": 0.03406934,
+      "step": 9495
+    },
+    {
+      "epoch": 18.992,
+      "grad_norm": 1.365356683731079,
+      "learning_rate": 2e-05,
+      "loss": 0.03618102,
+      "step": 9496
+    },
+    {
+      "epoch": 18.994,
+      "grad_norm": 1.526374101638794,
+      "learning_rate": 2e-05,
+      "loss": 0.03938029,
+      "step": 9497
+    },
+    {
+      "epoch": 18.996,
+      "grad_norm": 1.2375332117080688,
+      "learning_rate": 2e-05,
+      "loss": 0.04605676,
+      "step": 9498
+    },
+    {
+      "epoch": 18.998,
+      "grad_norm": 1.3788480758666992,
+      "learning_rate": 2e-05,
+      "loss": 0.04089382,
+      "step": 9499
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 1.1874690055847168,
+      "learning_rate": 2e-05,
+      "loss": 0.03390002,
+      "step": 9500
+    },
+    {
+      "epoch": 19.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.992,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9720558882235529,
+        "Equal_1": 0.994,
+        "Equal_2": 0.9720558882235529,
+        "Equal_3": 0.9620758483033932,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9940119760479041,
+        "Parallel_1": 0.9899799599198397,
+        "Parallel_2": 0.9979959919839679,
+        "Parallel_3": 0.992,
+        "Perpendicular_1": 0.994,
+        "Perpendicular_2": 0.986,
+        "Perpendicular_3": 0.7935871743486974,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9973333333333333,
+        "PointLiesOnCircle_3": 0.9933333333333333,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9780439121756487
+      },
+      "eval_runtime": 319.3618,
+      "eval_samples_per_second": 32.878,
+      "eval_steps_per_second": 0.658,
+      "step": 9500
+    },
+    {
+      "epoch": 19.002,
+      "grad_norm": 0.7328594923019409,
+      "learning_rate": 2e-05,
+      "loss": 0.02319442,
+      "step": 9501
+    },
+    {
+      "epoch": 19.004,
+      "grad_norm": 1.1689836978912354,
+      "learning_rate": 2e-05,
+      "loss": 0.03362505,
+      "step": 9502
+    },
+    {
+      "epoch": 19.006,
+      "grad_norm": 1.2494908571243286,
+      "learning_rate": 2e-05,
+      "loss": 0.04948422,
+      "step": 9503
+    },
+    {
+      "epoch": 19.008,
+      "grad_norm": 1.5237233638763428,
+      "learning_rate": 2e-05,
+      "loss": 0.03897671,
+      "step": 9504
+    },
+    {
+      "epoch": 19.01,
+      "grad_norm": 1.9613434076309204,
+      "learning_rate": 2e-05,
+      "loss": 0.0261043,
+      "step": 9505
+    },
+    {
+      "epoch": 19.012,
+      "grad_norm": 0.9122869372367859,
+      "learning_rate": 2e-05,
+      "loss": 0.0235957,
+      "step": 9506
+    },
+    {
+      "epoch": 19.014,
+      "grad_norm": 2.309724807739258,
+      "learning_rate": 2e-05,
+      "loss": 0.04885814,
+      "step": 9507
+    },
+    {
+      "epoch": 19.016,
+      "grad_norm": 1.0690571069717407,
+      "learning_rate": 2e-05,
+      "loss": 0.0407089,
+      "step": 9508
+    },
+    {
+      "epoch": 19.018,
+      "grad_norm": 1.2194948196411133,
+      "learning_rate": 2e-05,
+      "loss": 0.03530276,
+      "step": 9509
+    },
+    {
+      "epoch": 19.02,
+      "grad_norm": 1.3204545974731445,
+      "learning_rate": 2e-05,
+      "loss": 0.03693368,
+      "step": 9510
+    },
+    {
+      "epoch": 19.022,
+      "grad_norm": 1.1874244213104248,
+      "learning_rate": 2e-05,
+      "loss": 0.03080766,
+      "step": 9511
+    },
+    {
+      "epoch": 19.024,
+      "grad_norm": 2.5260133743286133,
+      "learning_rate": 2e-05,
+      "loss": 0.04506269,
+      "step": 9512
+    },
+    {
+      "epoch": 19.026,
+      "grad_norm": 0.8363938331604004,
+      "learning_rate": 2e-05,
+      "loss": 0.02124131,
+      "step": 9513
+    },
+    {
+      "epoch": 19.028,
+      "grad_norm": 2.5137112140655518,
+      "learning_rate": 2e-05,
+      "loss": 0.03363831,
+      "step": 9514
+    },
+    {
+      "epoch": 19.03,
+      "grad_norm": 1.983594536781311,
+      "learning_rate": 2e-05,
+      "loss": 0.04035079,
+      "step": 9515
+    },
+    {
+      "epoch": 19.032,
+      "grad_norm": 2.5143167972564697,
+      "learning_rate": 2e-05,
+      "loss": 0.05136511,
+      "step": 9516
+    },
+    {
+      "epoch": 19.034,
+      "grad_norm": 1.130037784576416,
+      "learning_rate": 2e-05,
+      "loss": 0.03966995,
+      "step": 9517
+    },
+    {
+      "epoch": 19.036,
+      "grad_norm": 1.764394760131836,
+      "learning_rate": 2e-05,
+      "loss": 0.04287095,
+      "step": 9518
+    },
+    {
+      "epoch": 19.038,
+      "grad_norm": 1.359204888343811,
+      "learning_rate": 2e-05,
+      "loss": 0.03334723,
+      "step": 9519
+    },
+    {
+      "epoch": 19.04,
+      "grad_norm": 1.1975325345993042,
+      "learning_rate": 2e-05,
+      "loss": 0.03555122,
+      "step": 9520
+    },
+    {
+      "epoch": 19.042,
+      "grad_norm": 0.9960451126098633,
+      "learning_rate": 2e-05,
+      "loss": 0.02776124,
+      "step": 9521
+    },
+    {
+      "epoch": 19.044,
+      "grad_norm": 1.2696477174758911,
+      "learning_rate": 2e-05,
+      "loss": 0.02704079,
+      "step": 9522
+    },
+    {
+      "epoch": 19.046,
+      "grad_norm": 1.127091646194458,
+      "learning_rate": 2e-05,
+      "loss": 0.03138428,
+      "step": 9523
+    },
+    {
+      "epoch": 19.048,
+      "grad_norm": 2.801011800765991,
+      "learning_rate": 2e-05,
+      "loss": 0.04420168,
+      "step": 9524
+    },
+    {
+      "epoch": 19.05,
+      "grad_norm": 1.672225832939148,
+      "learning_rate": 2e-05,
+      "loss": 0.03045794,
+      "step": 9525
+    },
+    {
+      "epoch": 19.052,
+      "grad_norm": 1.3001759052276611,
+      "learning_rate": 2e-05,
+      "loss": 0.03507664,
+      "step": 9526
+    },
+    {
+      "epoch": 19.054,
+      "grad_norm": 1.307578444480896,
+      "learning_rate": 2e-05,
+      "loss": 0.03051586,
+      "step": 9527
+    },
+    {
+      "epoch": 19.056,
+      "grad_norm": 2.060009479522705,
+      "learning_rate": 2e-05,
+      "loss": 0.04484456,
+      "step": 9528
+    },
+    {
+      "epoch": 19.058,
+      "grad_norm": 1.9255805015563965,
+      "learning_rate": 2e-05,
+      "loss": 0.03917917,
+      "step": 9529
+    },
+    {
+      "epoch": 19.06,
+      "grad_norm": 0.9870061874389648,
+      "learning_rate": 2e-05,
+      "loss": 0.02856109,
+      "step": 9530
+    },
+    {
+      "epoch": 19.062,
+      "grad_norm": 1.398775577545166,
+      "learning_rate": 2e-05,
+      "loss": 0.04870712,
+      "step": 9531
+    },
+    {
+      "epoch": 19.064,
+      "grad_norm": 1.2352436780929565,
+      "learning_rate": 2e-05,
+      "loss": 0.04471589,
+      "step": 9532
+    },
+    {
+      "epoch": 19.066,
+      "grad_norm": 1.8325438499450684,
+      "learning_rate": 2e-05,
+      "loss": 0.04382673,
+      "step": 9533
+    },
+    {
+      "epoch": 19.068,
+      "grad_norm": 1.0861464738845825,
+      "learning_rate": 2e-05,
+      "loss": 0.02810282,
+      "step": 9534
+    },
+    {
+      "epoch": 19.07,
+      "grad_norm": 3.1042890548706055,
+      "learning_rate": 2e-05,
+      "loss": 0.04670685,
+      "step": 9535
+    },
+    {
+      "epoch": 19.072,
+      "grad_norm": 1.2983977794647217,
+      "learning_rate": 2e-05,
+      "loss": 0.03450656,
+      "step": 9536
+    },
+    {
+      "epoch": 19.074,
+      "grad_norm": 2.436805248260498,
+      "learning_rate": 2e-05,
+      "loss": 0.03273325,
+      "step": 9537
+    },
+    {
+      "epoch": 19.076,
+      "grad_norm": 1.8270798921585083,
+      "learning_rate": 2e-05,
+      "loss": 0.05142706,
+      "step": 9538
+    },
+    {
+      "epoch": 19.078,
+      "grad_norm": 1.7410194873809814,
+      "learning_rate": 2e-05,
+      "loss": 0.04839787,
+      "step": 9539
+    },
+    {
+      "epoch": 19.08,
+      "grad_norm": 0.7884871959686279,
+      "learning_rate": 2e-05,
+      "loss": 0.02119577,
+      "step": 9540
+    },
+    {
+      "epoch": 19.082,
+      "grad_norm": 1.4480359554290771,
+      "learning_rate": 2e-05,
+      "loss": 0.03697205,
+      "step": 9541
+    },
+    {
+      "epoch": 19.084,
+      "grad_norm": 0.9422207474708557,
+      "learning_rate": 2e-05,
+      "loss": 0.02769065,
+      "step": 9542
+    },
+    {
+      "epoch": 19.086,
+      "grad_norm": 0.791770339012146,
+      "learning_rate": 2e-05,
+      "loss": 0.01657294,
+      "step": 9543
+    },
+    {
+      "epoch": 19.088,
+      "grad_norm": 1.4777929782867432,
+      "learning_rate": 2e-05,
+      "loss": 0.04559475,
+      "step": 9544
+    },
+    {
+      "epoch": 19.09,
+      "grad_norm": 1.4669833183288574,
+      "learning_rate": 2e-05,
+      "loss": 0.03167301,
+      "step": 9545
+    },
+    {
+      "epoch": 19.092,
+      "grad_norm": 1.2533551454544067,
+      "learning_rate": 2e-05,
+      "loss": 0.02627946,
+      "step": 9546
+    },
+    {
+      "epoch": 19.094,
+      "grad_norm": 1.4709750413894653,
+      "learning_rate": 2e-05,
+      "loss": 0.04461734,
+      "step": 9547
+    },
+    {
+      "epoch": 19.096,
+      "grad_norm": 1.149090051651001,
+      "learning_rate": 2e-05,
+      "loss": 0.03871008,
+      "step": 9548
+    },
+    {
+      "epoch": 19.098,
+      "grad_norm": 1.1180267333984375,
+      "learning_rate": 2e-05,
+      "loss": 0.03474029,
+      "step": 9549
+    },
+    {
+      "epoch": 19.1,
+      "grad_norm": 1.3500217199325562,
+      "learning_rate": 2e-05,
+      "loss": 0.02843937,
+      "step": 9550
+    },
+    {
+      "epoch": 19.102,
+      "grad_norm": 1.7513254880905151,
+      "learning_rate": 2e-05,
+      "loss": 0.0540585,
+      "step": 9551
+    },
+    {
+      "epoch": 19.104,
+      "grad_norm": 1.781827688217163,
+      "learning_rate": 2e-05,
+      "loss": 0.04126269,
+      "step": 9552
+    },
+    {
+      "epoch": 19.106,
+      "grad_norm": 1.1269136667251587,
+      "learning_rate": 2e-05,
+      "loss": 0.03399577,
+      "step": 9553
+    },
+    {
+      "epoch": 19.108,
+      "grad_norm": 1.8535258769989014,
+      "learning_rate": 2e-05,
+      "loss": 0.05306114,
+      "step": 9554
+    },
+    {
+      "epoch": 19.11,
+      "grad_norm": 2.0714800357818604,
+      "learning_rate": 2e-05,
+      "loss": 0.0468117,
+      "step": 9555
+    },
+    {
+      "epoch": 19.112,
+      "grad_norm": 1.0497852563858032,
+      "learning_rate": 2e-05,
+      "loss": 0.02673732,
+      "step": 9556
+    },
+    {
+      "epoch": 19.114,
+      "grad_norm": 0.9615012407302856,
+      "learning_rate": 2e-05,
+      "loss": 0.02658118,
+      "step": 9557
+    },
+    {
+      "epoch": 19.116,
+      "grad_norm": 3.7830493450164795,
+      "learning_rate": 2e-05,
+      "loss": 0.04481242,
+      "step": 9558
+    },
+    {
+      "epoch": 19.118,
+      "grad_norm": 1.9422591924667358,
+      "learning_rate": 2e-05,
+      "loss": 0.05550155,
+      "step": 9559
+    },
+    {
+      "epoch": 19.12,
+      "grad_norm": 1.294343113899231,
+      "learning_rate": 2e-05,
+      "loss": 0.03953799,
+      "step": 9560
+    },
+    {
+      "epoch": 19.122,
+      "grad_norm": 1.5340063571929932,
+      "learning_rate": 2e-05,
+      "loss": 0.03284139,
+      "step": 9561
+    },
+    {
+      "epoch": 19.124,
+      "grad_norm": 1.1982046365737915,
+      "learning_rate": 2e-05,
+      "loss": 0.04478461,
+      "step": 9562
+    },
+    {
+      "epoch": 19.126,
+      "grad_norm": 1.1274042129516602,
+      "learning_rate": 2e-05,
+      "loss": 0.03823815,
+      "step": 9563
+    },
+    {
+      "epoch": 19.128,
+      "grad_norm": 1.3324651718139648,
+      "learning_rate": 2e-05,
+      "loss": 0.04417039,
+      "step": 9564
+    },
+    {
+      "epoch": 19.13,
+      "grad_norm": 3.788226366043091,
+      "learning_rate": 2e-05,
+      "loss": 0.04432316,
+      "step": 9565
+    },
+    {
+      "epoch": 19.132,
+      "grad_norm": 1.0010766983032227,
+      "learning_rate": 2e-05,
+      "loss": 0.02808233,
+      "step": 9566
+    },
+    {
+      "epoch": 19.134,
+      "grad_norm": 1.5182632207870483,
+      "learning_rate": 2e-05,
+      "loss": 0.04026213,
+      "step": 9567
+    },
+    {
+      "epoch": 19.136,
+      "grad_norm": 1.1371148824691772,
+      "learning_rate": 2e-05,
+      "loss": 0.03047445,
+      "step": 9568
+    },
+    {
+      "epoch": 19.138,
+      "grad_norm": 1.003008246421814,
+      "learning_rate": 2e-05,
+      "loss": 0.03161223,
+      "step": 9569
+    },
+    {
+      "epoch": 19.14,
+      "grad_norm": 1.9005447626113892,
+      "learning_rate": 2e-05,
+      "loss": 0.04390565,
+      "step": 9570
+    },
+    {
+      "epoch": 19.142,
+      "grad_norm": 1.1964303255081177,
+      "learning_rate": 2e-05,
+      "loss": 0.03318737,
+      "step": 9571
+    },
+    {
+      "epoch": 19.144,
+      "grad_norm": 1.471183180809021,
+      "learning_rate": 2e-05,
+      "loss": 0.04245757,
+      "step": 9572
+    },
+    {
+      "epoch": 19.146,
+      "grad_norm": 1.0873215198516846,
+      "learning_rate": 2e-05,
+      "loss": 0.03344093,
+      "step": 9573
+    },
+    {
+      "epoch": 19.148,
+      "grad_norm": 1.1384050846099854,
+      "learning_rate": 2e-05,
+      "loss": 0.03804214,
+      "step": 9574
+    },
+    {
+      "epoch": 19.15,
+      "grad_norm": 1.8956702947616577,
+      "learning_rate": 2e-05,
+      "loss": 0.04579899,
+      "step": 9575
+    },
+    {
+      "epoch": 19.152,
+      "grad_norm": 0.9021958112716675,
+      "learning_rate": 2e-05,
+      "loss": 0.02364812,
+      "step": 9576
+    },
+    {
+      "epoch": 19.154,
+      "grad_norm": 0.9730253219604492,
+      "learning_rate": 2e-05,
+      "loss": 0.0266248,
+      "step": 9577
+    },
+    {
+      "epoch": 19.156,
+      "grad_norm": 1.2630407810211182,
+      "learning_rate": 2e-05,
+      "loss": 0.03151676,
+      "step": 9578
+    },
+    {
+      "epoch": 19.158,
+      "grad_norm": 1.0007412433624268,
+      "learning_rate": 2e-05,
+      "loss": 0.03561787,
+      "step": 9579
+    },
+    {
+      "epoch": 19.16,
+      "grad_norm": 1.62027907371521,
+      "learning_rate": 2e-05,
+      "loss": 0.05210748,
+      "step": 9580
+    },
+    {
+      "epoch": 19.162,
+      "grad_norm": 0.950829029083252,
+      "learning_rate": 2e-05,
+      "loss": 0.03114957,
+      "step": 9581
+    },
+    {
+      "epoch": 19.164,
+      "grad_norm": 1.3806723356246948,
+      "learning_rate": 2e-05,
+      "loss": 0.04665224,
+      "step": 9582
+    },
+    {
+      "epoch": 19.166,
+      "grad_norm": 4.24110221862793,
+      "learning_rate": 2e-05,
+      "loss": 0.03417417,
+      "step": 9583
+    },
+    {
+      "epoch": 19.168,
+      "grad_norm": 1.3316015005111694,
+      "learning_rate": 2e-05,
+      "loss": 0.04237635,
+      "step": 9584
+    },
+    {
+      "epoch": 19.17,
+      "grad_norm": 1.7277883291244507,
+      "learning_rate": 2e-05,
+      "loss": 0.03498271,
+      "step": 9585
+    },
+    {
+      "epoch": 19.172,
+      "grad_norm": 1.2385425567626953,
+      "learning_rate": 2e-05,
+      "loss": 0.03356936,
+      "step": 9586
+    },
+    {
+      "epoch": 19.174,
+      "grad_norm": 0.8457931280136108,
+      "learning_rate": 2e-05,
+      "loss": 0.01994565,
+      "step": 9587
+    },
+    {
+      "epoch": 19.176,
+      "grad_norm": 1.2455838918685913,
+      "learning_rate": 2e-05,
+      "loss": 0.03502472,
+      "step": 9588
+    },
+    {
+      "epoch": 19.178,
+      "grad_norm": 1.9572054147720337,
+      "learning_rate": 2e-05,
+      "loss": 0.03634415,
+      "step": 9589
+    },
+    {
+      "epoch": 19.18,
+      "grad_norm": 1.2908679246902466,
+      "learning_rate": 2e-05,
+      "loss": 0.03044868,
+      "step": 9590
+    },
+    {
+      "epoch": 19.182,
+      "grad_norm": 2.5042431354522705,
+      "learning_rate": 2e-05,
+      "loss": 0.04156593,
+      "step": 9591
+    },
+    {
+      "epoch": 19.184,
+      "grad_norm": 1.0194517374038696,
+      "learning_rate": 2e-05,
+      "loss": 0.02526481,
+      "step": 9592
+    },
+    {
+      "epoch": 19.186,
+      "grad_norm": 1.0619885921478271,
+      "learning_rate": 2e-05,
+      "loss": 0.0298758,
+      "step": 9593
+    },
+    {
+      "epoch": 19.188,
+      "grad_norm": 1.263814926147461,
+      "learning_rate": 2e-05,
+      "loss": 0.03626692,
+      "step": 9594
+    },
+    {
+      "epoch": 19.19,
+      "grad_norm": 2.5439209938049316,
+      "learning_rate": 2e-05,
+      "loss": 0.05030162,
+      "step": 9595
+    },
+    {
+      "epoch": 19.192,
+      "grad_norm": 1.6095026731491089,
+      "learning_rate": 2e-05,
+      "loss": 0.04484971,
+      "step": 9596
+    },
+    {
+      "epoch": 19.194,
+      "grad_norm": 1.921481728553772,
+      "learning_rate": 2e-05,
+      "loss": 0.05408298,
+      "step": 9597
+    },
+    {
+      "epoch": 19.196,
+      "grad_norm": 2.8024349212646484,
+      "learning_rate": 2e-05,
+      "loss": 0.05455287,
+      "step": 9598
+    },
+    {
+      "epoch": 19.198,
+      "grad_norm": 0.9602892398834229,
+      "learning_rate": 2e-05,
+      "loss": 0.02859609,
+      "step": 9599
+    },
+    {
+      "epoch": 19.2,
+      "grad_norm": 1.6992692947387695,
+      "learning_rate": 2e-05,
+      "loss": 0.03378526,
+      "step": 9600
+    },
+    {
+      "epoch": 19.202,
+      "grad_norm": 2.3561947345733643,
+      "learning_rate": 2e-05,
+      "loss": 0.03298889,
+      "step": 9601
+    },
+    {
+      "epoch": 19.204,
+      "grad_norm": 1.4221720695495605,
+      "learning_rate": 2e-05,
+      "loss": 0.04421243,
+      "step": 9602
+    },
+    {
+      "epoch": 19.206,
+      "grad_norm": 1.0201853513717651,
+      "learning_rate": 2e-05,
+      "loss": 0.02438156,
+      "step": 9603
+    },
+    {
+      "epoch": 19.208,
+      "grad_norm": 0.9587017893791199,
+      "learning_rate": 2e-05,
+      "loss": 0.02957552,
+      "step": 9604
+    },
+    {
+      "epoch": 19.21,
+      "grad_norm": 1.138653039932251,
+      "learning_rate": 2e-05,
+      "loss": 0.02991896,
+      "step": 9605
+    },
+    {
+      "epoch": 19.212,
+      "grad_norm": 2.0702502727508545,
+      "learning_rate": 2e-05,
+      "loss": 0.03139401,
+      "step": 9606
+    },
+    {
+      "epoch": 19.214,
+      "grad_norm": 1.0454638004302979,
+      "learning_rate": 2e-05,
+      "loss": 0.02659562,
+      "step": 9607
+    },
+    {
+      "epoch": 19.216,
+      "grad_norm": 2.350715160369873,
+      "learning_rate": 2e-05,
+      "loss": 0.05148138,
+      "step": 9608
+    },
+    {
+      "epoch": 19.218,
+      "grad_norm": 0.7986869812011719,
+      "learning_rate": 2e-05,
+      "loss": 0.02145108,
+      "step": 9609
+    },
+    {
+      "epoch": 19.22,
+      "grad_norm": 1.9818875789642334,
+      "learning_rate": 2e-05,
+      "loss": 0.04348661,
+      "step": 9610
+    },
+    {
+      "epoch": 19.222,
+      "grad_norm": 1.72350013256073,
+      "learning_rate": 2e-05,
+      "loss": 0.03537084,
+      "step": 9611
+    },
+    {
+      "epoch": 19.224,
+      "grad_norm": 0.917288601398468,
+      "learning_rate": 2e-05,
+      "loss": 0.02850287,
+      "step": 9612
+    },
+    {
+      "epoch": 19.226,
+      "grad_norm": 2.157785177230835,
+      "learning_rate": 2e-05,
+      "loss": 0.04212007,
+      "step": 9613
+    },
+    {
+      "epoch": 19.228,
+      "grad_norm": 1.6661065816879272,
+      "learning_rate": 2e-05,
+      "loss": 0.04836546,
+      "step": 9614
+    },
+    {
+      "epoch": 19.23,
+      "grad_norm": 1.0794106721878052,
+      "learning_rate": 2e-05,
+      "loss": 0.0291114,
+      "step": 9615
+    },
+    {
+      "epoch": 19.232,
+      "grad_norm": 1.1897625923156738,
+      "learning_rate": 2e-05,
+      "loss": 0.04130687,
+      "step": 9616
+    },
+    {
+      "epoch": 19.234,
+      "grad_norm": 1.4216676950454712,
+      "learning_rate": 2e-05,
+      "loss": 0.03022526,
+      "step": 9617
+    },
+    {
+      "epoch": 19.236,
+      "grad_norm": 1.3475738763809204,
+      "learning_rate": 2e-05,
+      "loss": 0.04719062,
+      "step": 9618
+    },
+    {
+      "epoch": 19.238,
+      "grad_norm": 1.4195927381515503,
+      "learning_rate": 2e-05,
+      "loss": 0.03779185,
+      "step": 9619
+    },
+    {
+      "epoch": 19.24,
+      "grad_norm": 1.1476095914840698,
+      "learning_rate": 2e-05,
+      "loss": 0.0434536,
+      "step": 9620
+    },
+    {
+      "epoch": 19.242,
+      "grad_norm": 1.682695984840393,
+      "learning_rate": 2e-05,
+      "loss": 0.05591667,
+      "step": 9621
+    },
+    {
+      "epoch": 19.244,
+      "grad_norm": 1.1905351877212524,
+      "learning_rate": 2e-05,
+      "loss": 0.0391013,
+      "step": 9622
+    },
+    {
+      "epoch": 19.246,
+      "grad_norm": 1.3827733993530273,
+      "learning_rate": 2e-05,
+      "loss": 0.02527856,
+      "step": 9623
+    },
+    {
+      "epoch": 19.248,
+      "grad_norm": 1.235648512840271,
+      "learning_rate": 2e-05,
+      "loss": 0.03677778,
+      "step": 9624
+    },
+    {
+      "epoch": 19.25,
+      "grad_norm": 2.980347156524658,
+      "learning_rate": 2e-05,
+      "loss": 0.06639814,
+      "step": 9625
+    },
+    {
+      "epoch": 19.252,
+      "grad_norm": 1.079358458518982,
+      "learning_rate": 2e-05,
+      "loss": 0.0342607,
+      "step": 9626
+    },
+    {
+      "epoch": 19.254,
+      "grad_norm": 1.3070192337036133,
+      "learning_rate": 2e-05,
+      "loss": 0.03877965,
+      "step": 9627
+    },
+    {
+      "epoch": 19.256,
+      "grad_norm": 1.132627010345459,
+      "learning_rate": 2e-05,
+      "loss": 0.03127551,
+      "step": 9628
+    },
+    {
+      "epoch": 19.258,
+      "grad_norm": 1.1576032638549805,
+      "learning_rate": 2e-05,
+      "loss": 0.03139985,
+      "step": 9629
+    },
+    {
+      "epoch": 19.26,
+      "grad_norm": 1.2397043704986572,
+      "learning_rate": 2e-05,
+      "loss": 0.04228866,
+      "step": 9630
+    },
+    {
+      "epoch": 19.262,
+      "grad_norm": 1.0662872791290283,
+      "learning_rate": 2e-05,
+      "loss": 0.03128354,
+      "step": 9631
+    },
+    {
+      "epoch": 19.264,
+      "grad_norm": 3.369060754776001,
+      "learning_rate": 2e-05,
+      "loss": 0.03271961,
+      "step": 9632
+    },
+    {
+      "epoch": 19.266,
+      "grad_norm": 1.7384709119796753,
+      "learning_rate": 2e-05,
+      "loss": 0.04713779,
+      "step": 9633
+    },
+    {
+      "epoch": 19.268,
+      "grad_norm": 1.8644956350326538,
+      "learning_rate": 2e-05,
+      "loss": 0.03739468,
+      "step": 9634
+    },
+    {
+      "epoch": 19.27,
+      "grad_norm": 1.260280966758728,
+      "learning_rate": 2e-05,
+      "loss": 0.0264209,
+      "step": 9635
+    },
+    {
+      "epoch": 19.272,
+      "grad_norm": 1.2313249111175537,
+      "learning_rate": 2e-05,
+      "loss": 0.03959552,
+      "step": 9636
+    },
+    {
+      "epoch": 19.274,
+      "grad_norm": 1.3005576133728027,
+      "learning_rate": 2e-05,
+      "loss": 0.03051363,
+      "step": 9637
+    },
+    {
+      "epoch": 19.276,
+      "grad_norm": 1.6750763654708862,
+      "learning_rate": 2e-05,
+      "loss": 0.05048709,
+      "step": 9638
+    },
+    {
+      "epoch": 19.278,
+      "grad_norm": 1.6315701007843018,
+      "learning_rate": 2e-05,
+      "loss": 0.02604922,
+      "step": 9639
+    },
+    {
+      "epoch": 19.28,
+      "grad_norm": 1.7863506078720093,
+      "learning_rate": 2e-05,
+      "loss": 0.03610089,
+      "step": 9640
+    },
+    {
+      "epoch": 19.282,
+      "grad_norm": 1.3061033487319946,
+      "learning_rate": 2e-05,
+      "loss": 0.04164078,
+      "step": 9641
+    },
+    {
+      "epoch": 19.284,
+      "grad_norm": 1.4289368391036987,
+      "learning_rate": 2e-05,
+      "loss": 0.02338367,
+      "step": 9642
+    },
+    {
+      "epoch": 19.286,
+      "grad_norm": 1.3021084070205688,
+      "learning_rate": 2e-05,
+      "loss": 0.04391333,
+      "step": 9643
+    },
+    {
+      "epoch": 19.288,
+      "grad_norm": 1.178717851638794,
+      "learning_rate": 2e-05,
+      "loss": 0.03647342,
+      "step": 9644
+    },
+    {
+      "epoch": 19.29,
+      "grad_norm": 0.8449578881263733,
+      "learning_rate": 2e-05,
+      "loss": 0.02405243,
+      "step": 9645
+    },
+    {
+      "epoch": 19.292,
+      "grad_norm": 1.4764927625656128,
+      "learning_rate": 2e-05,
+      "loss": 0.05233892,
+      "step": 9646
+    },
+    {
+      "epoch": 19.294,
+      "grad_norm": 1.8029993772506714,
+      "learning_rate": 2e-05,
+      "loss": 0.03435803,
+      "step": 9647
+    },
+    {
+      "epoch": 19.296,
+      "grad_norm": 1.4311555624008179,
+      "learning_rate": 2e-05,
+      "loss": 0.03782001,
+      "step": 9648
+    },
+    {
+      "epoch": 19.298,
+      "grad_norm": 1.300187587738037,
+      "learning_rate": 2e-05,
+      "loss": 0.05648835,
+      "step": 9649
+    },
+    {
+      "epoch": 19.3,
+      "grad_norm": 1.4080634117126465,
+      "learning_rate": 2e-05,
+      "loss": 0.03893708,
+      "step": 9650
+    },
+    {
+      "epoch": 19.302,
+      "grad_norm": 2.201977014541626,
+      "learning_rate": 2e-05,
+      "loss": 0.04349912,
+      "step": 9651
+    },
+    {
+      "epoch": 19.304,
+      "grad_norm": 0.9720236659049988,
+      "learning_rate": 2e-05,
+      "loss": 0.02697369,
+      "step": 9652
+    },
+    {
+      "epoch": 19.306,
+      "grad_norm": 1.6441378593444824,
+      "learning_rate": 2e-05,
+      "loss": 0.05696252,
+      "step": 9653
+    },
+    {
+      "epoch": 19.308,
+      "grad_norm": 1.6382756233215332,
+      "learning_rate": 2e-05,
+      "loss": 0.03626692,
+      "step": 9654
+    },
+    {
+      "epoch": 19.31,
+      "grad_norm": 1.034172534942627,
+      "learning_rate": 2e-05,
+      "loss": 0.03107112,
+      "step": 9655
+    },
+    {
+      "epoch": 19.312,
+      "grad_norm": 1.183516263961792,
+      "learning_rate": 2e-05,
+      "loss": 0.03426999,
+      "step": 9656
+    },
+    {
+      "epoch": 19.314,
+      "grad_norm": 1.6998095512390137,
+      "learning_rate": 2e-05,
+      "loss": 0.03958786,
+      "step": 9657
+    },
+    {
+      "epoch": 19.316,
+      "grad_norm": 1.598720669746399,
+      "learning_rate": 2e-05,
+      "loss": 0.03835572,
+      "step": 9658
+    },
+    {
+      "epoch": 19.318,
+      "grad_norm": 1.1745725870132446,
+      "learning_rate": 2e-05,
+      "loss": 0.03925322,
+      "step": 9659
+    },
+    {
+      "epoch": 19.32,
+      "grad_norm": 1.0180588960647583,
+      "learning_rate": 2e-05,
+      "loss": 0.03550194,
+      "step": 9660
+    },
+    {
+      "epoch": 19.322,
+      "grad_norm": 0.9355264902114868,
+      "learning_rate": 2e-05,
+      "loss": 0.03325229,
+      "step": 9661
+    },
+    {
+      "epoch": 19.324,
+      "grad_norm": 1.2007991075515747,
+      "learning_rate": 2e-05,
+      "loss": 0.04493001,
+      "step": 9662
+    },
+    {
+      "epoch": 19.326,
+      "grad_norm": 0.86708664894104,
+      "learning_rate": 2e-05,
+      "loss": 0.02545637,
+      "step": 9663
+    },
+    {
+      "epoch": 19.328,
+      "grad_norm": 1.0623421669006348,
+      "learning_rate": 2e-05,
+      "loss": 0.03543958,
+      "step": 9664
+    },
+    {
+      "epoch": 19.33,
+      "grad_norm": 1.6305174827575684,
+      "learning_rate": 2e-05,
+      "loss": 0.04481956,
+      "step": 9665
+    },
+    {
+      "epoch": 19.332,
+      "grad_norm": 2.0141186714172363,
+      "learning_rate": 2e-05,
+      "loss": 0.04622241,
+      "step": 9666
+    },
+    {
+      "epoch": 19.334,
+      "grad_norm": 1.1043932437896729,
+      "learning_rate": 2e-05,
+      "loss": 0.03127149,
+      "step": 9667
+    },
+    {
+      "epoch": 19.336,
+      "grad_norm": 1.262362003326416,
+      "learning_rate": 2e-05,
+      "loss": 0.02693222,
+      "step": 9668
+    },
+    {
+      "epoch": 19.338,
+      "grad_norm": 2.6822168827056885,
+      "learning_rate": 2e-05,
+      "loss": 0.03670801,
+      "step": 9669
+    },
+    {
+      "epoch": 19.34,
+      "grad_norm": 1.137508749961853,
+      "learning_rate": 2e-05,
+      "loss": 0.03024655,
+      "step": 9670
+    },
+    {
+      "epoch": 19.342,
+      "grad_norm": 1.265586018562317,
+      "learning_rate": 2e-05,
+      "loss": 0.03883039,
+      "step": 9671
+    },
+    {
+      "epoch": 19.344,
+      "grad_norm": 1.6028434038162231,
+      "learning_rate": 2e-05,
+      "loss": 0.04389777,
+      "step": 9672
+    },
+    {
+      "epoch": 19.346,
+      "grad_norm": 2.5654678344726562,
+      "learning_rate": 2e-05,
+      "loss": 0.049094,
+      "step": 9673
+    },
+    {
+      "epoch": 19.348,
+      "grad_norm": 1.247218370437622,
+      "learning_rate": 2e-05,
+      "loss": 0.04227669,
+      "step": 9674
+    },
+    {
+      "epoch": 19.35,
+      "grad_norm": 0.9476075768470764,
+      "learning_rate": 2e-05,
+      "loss": 0.02782707,
+      "step": 9675
+    },
+    {
+      "epoch": 19.352,
+      "grad_norm": 1.2325526475906372,
+      "learning_rate": 2e-05,
+      "loss": 0.02918253,
+      "step": 9676
+    },
+    {
+      "epoch": 19.354,
+      "grad_norm": 1.1856415271759033,
+      "learning_rate": 2e-05,
+      "loss": 0.0287297,
+      "step": 9677
+    },
+    {
+      "epoch": 19.356,
+      "grad_norm": 0.832682728767395,
+      "learning_rate": 2e-05,
+      "loss": 0.02484965,
+      "step": 9678
+    },
+    {
+      "epoch": 19.358,
+      "grad_norm": 2.037794828414917,
+      "learning_rate": 2e-05,
+      "loss": 0.0416809,
+      "step": 9679
+    },
+    {
+      "epoch": 19.36,
+      "grad_norm": 1.2113049030303955,
+      "learning_rate": 2e-05,
+      "loss": 0.03574871,
+      "step": 9680
+    },
+    {
+      "epoch": 19.362,
+      "grad_norm": 1.0858124494552612,
+      "learning_rate": 2e-05,
+      "loss": 0.03182752,
+      "step": 9681
+    },
+    {
+      "epoch": 19.364,
+      "grad_norm": 1.3244341611862183,
+      "learning_rate": 2e-05,
+      "loss": 0.0379288,
+      "step": 9682
+    },
+    {
+      "epoch": 19.366,
+      "grad_norm": 0.96357661485672,
+      "learning_rate": 2e-05,
+      "loss": 0.03285439,
+      "step": 9683
+    },
+    {
+      "epoch": 19.368,
+      "grad_norm": 0.9765337109565735,
+      "learning_rate": 2e-05,
+      "loss": 0.02620794,
+      "step": 9684
+    },
+    {
+      "epoch": 19.37,
+      "grad_norm": 0.8831155300140381,
+      "learning_rate": 2e-05,
+      "loss": 0.02626178,
+      "step": 9685
+    },
+    {
+      "epoch": 19.372,
+      "grad_norm": 1.1981676816940308,
+      "learning_rate": 2e-05,
+      "loss": 0.02839826,
+      "step": 9686
+    },
+    {
+      "epoch": 19.374,
+      "grad_norm": 3.4404029846191406,
+      "learning_rate": 2e-05,
+      "loss": 0.04103406,
+      "step": 9687
+    },
+    {
+      "epoch": 19.376,
+      "grad_norm": 1.0752460956573486,
+      "learning_rate": 2e-05,
+      "loss": 0.02952642,
+      "step": 9688
+    },
+    {
+      "epoch": 19.378,
+      "grad_norm": 1.3563824892044067,
+      "learning_rate": 2e-05,
+      "loss": 0.0425565,
+      "step": 9689
+    },
+    {
+      "epoch": 19.38,
+      "grad_norm": 1.359243631362915,
+      "learning_rate": 2e-05,
+      "loss": 0.02972773,
+      "step": 9690
+    },
+    {
+      "epoch": 19.382,
+      "grad_norm": 1.3436927795410156,
+      "learning_rate": 2e-05,
+      "loss": 0.04154352,
+      "step": 9691
+    },
+    {
+      "epoch": 19.384,
+      "grad_norm": 1.4160605669021606,
+      "learning_rate": 2e-05,
+      "loss": 0.04017557,
+      "step": 9692
+    },
+    {
+      "epoch": 19.386,
+      "grad_norm": 2.5221076011657715,
+      "learning_rate": 2e-05,
+      "loss": 0.04263007,
+      "step": 9693
+    },
+    {
+      "epoch": 19.388,
+      "grad_norm": 1.3207120895385742,
+      "learning_rate": 2e-05,
+      "loss": 0.04798321,
+      "step": 9694
+    },
+    {
+      "epoch": 19.39,
+      "grad_norm": 1.1256715059280396,
+      "learning_rate": 2e-05,
+      "loss": 0.03117888,
+      "step": 9695
+    },
+    {
+      "epoch": 19.392,
+      "grad_norm": 1.1344276666641235,
+      "learning_rate": 2e-05,
+      "loss": 0.03253318,
+      "step": 9696
+    },
+    {
+      "epoch": 19.394,
+      "grad_norm": 1.3823455572128296,
+      "learning_rate": 2e-05,
+      "loss": 0.03994993,
+      "step": 9697
+    },
+    {
+      "epoch": 19.396,
+      "grad_norm": 2.88213849067688,
+      "learning_rate": 2e-05,
+      "loss": 0.03269661,
+      "step": 9698
+    },
+    {
+      "epoch": 19.398,
+      "grad_norm": 1.369133472442627,
+      "learning_rate": 2e-05,
+      "loss": 0.04676585,
+      "step": 9699
+    },
+    {
+      "epoch": 19.4,
+      "grad_norm": 1.3740583658218384,
+      "learning_rate": 2e-05,
+      "loss": 0.03353403,
+      "step": 9700
+    },
+    {
+      "epoch": 19.402,
+      "grad_norm": 1.107534646987915,
+      "learning_rate": 2e-05,
+      "loss": 0.03524677,
+      "step": 9701
+    },
+    {
+      "epoch": 19.404,
+      "grad_norm": 1.0297796726226807,
+      "learning_rate": 2e-05,
+      "loss": 0.03444082,
+      "step": 9702
+    },
+    {
+      "epoch": 19.406,
+      "grad_norm": 1.664433240890503,
+      "learning_rate": 2e-05,
+      "loss": 0.04659187,
+      "step": 9703
+    },
+    {
+      "epoch": 19.408,
+      "grad_norm": 1.3409240245819092,
+      "learning_rate": 2e-05,
+      "loss": 0.03735002,
+      "step": 9704
+    },
+    {
+      "epoch": 19.41,
+      "grad_norm": 2.016294479370117,
+      "learning_rate": 2e-05,
+      "loss": 0.03130295,
+      "step": 9705
+    },
+    {
+      "epoch": 19.412,
+      "grad_norm": 2.1538240909576416,
+      "learning_rate": 2e-05,
+      "loss": 0.05228285,
+      "step": 9706
+    },
+    {
+      "epoch": 19.414,
+      "grad_norm": 1.5620522499084473,
+      "learning_rate": 2e-05,
+      "loss": 0.03496569,
+      "step": 9707
+    },
+    {
+      "epoch": 19.416,
+      "grad_norm": 2.113233804702759,
+      "learning_rate": 2e-05,
+      "loss": 0.03931812,
+      "step": 9708
+    },
+    {
+      "epoch": 19.418,
+      "grad_norm": 1.4287680387496948,
+      "learning_rate": 2e-05,
+      "loss": 0.02812807,
+      "step": 9709
+    },
+    {
+      "epoch": 19.42,
+      "grad_norm": 1.4849001169204712,
+      "learning_rate": 2e-05,
+      "loss": 0.03003873,
+      "step": 9710
+    },
+    {
+      "epoch": 19.422,
+      "grad_norm": 2.0336673259735107,
+      "learning_rate": 2e-05,
+      "loss": 0.05090691,
+      "step": 9711
+    },
+    {
+      "epoch": 19.424,
+      "grad_norm": 1.0571931600570679,
+      "learning_rate": 2e-05,
+      "loss": 0.0383536,
+      "step": 9712
+    },
+    {
+      "epoch": 19.426,
+      "grad_norm": 0.924149751663208,
+      "learning_rate": 2e-05,
+      "loss": 0.02016873,
+      "step": 9713
+    },
+    {
+      "epoch": 19.428,
+      "grad_norm": 1.2180832624435425,
+      "learning_rate": 2e-05,
+      "loss": 0.02738358,
+      "step": 9714
+    },
+    {
+      "epoch": 19.43,
+      "grad_norm": 3.771353006362915,
+      "learning_rate": 2e-05,
+      "loss": 0.03492292,
+      "step": 9715
+    },
+    {
+      "epoch": 19.432,
+      "grad_norm": 0.9769635796546936,
+      "learning_rate": 2e-05,
+      "loss": 0.0286561,
+      "step": 9716
+    },
+    {
+      "epoch": 19.434,
+      "grad_norm": 1.5093841552734375,
+      "learning_rate": 2e-05,
+      "loss": 0.03438292,
+      "step": 9717
+    },
+    {
+      "epoch": 19.436,
+      "grad_norm": 1.6384273767471313,
+      "learning_rate": 2e-05,
+      "loss": 0.03842306,
+      "step": 9718
+    },
+    {
+      "epoch": 19.438,
+      "grad_norm": 1.3384544849395752,
+      "learning_rate": 2e-05,
+      "loss": 0.03926391,
+      "step": 9719
+    },
+    {
+      "epoch": 19.44,
+      "grad_norm": 1.1322165727615356,
+      "learning_rate": 2e-05,
+      "loss": 0.04284387,
+      "step": 9720
+    },
+    {
+      "epoch": 19.442,
+      "grad_norm": 1.9500755071640015,
+      "learning_rate": 2e-05,
+      "loss": 0.03185847,
+      "step": 9721
+    },
+    {
+      "epoch": 19.444,
+      "grad_norm": 2.0812883377075195,
+      "learning_rate": 2e-05,
+      "loss": 0.05557717,
+      "step": 9722
+    },
+    {
+      "epoch": 19.446,
+      "grad_norm": 0.8484706878662109,
+      "learning_rate": 2e-05,
+      "loss": 0.01811722,
+      "step": 9723
+    },
+    {
+      "epoch": 19.448,
+      "grad_norm": 1.0412821769714355,
+      "learning_rate": 2e-05,
+      "loss": 0.0315608,
+      "step": 9724
+    },
+    {
+      "epoch": 19.45,
+      "grad_norm": 1.6890451908111572,
+      "learning_rate": 2e-05,
+      "loss": 0.0433359,
+      "step": 9725
+    },
+    {
+      "epoch": 19.452,
+      "grad_norm": 1.1269593238830566,
+      "learning_rate": 2e-05,
+      "loss": 0.0354099,
+      "step": 9726
+    },
+    {
+      "epoch": 19.454,
+      "grad_norm": 1.0882933139801025,
+      "learning_rate": 2e-05,
+      "loss": 0.02488111,
+      "step": 9727
+    },
+    {
+      "epoch": 19.456,
+      "grad_norm": 2.212448835372925,
+      "learning_rate": 2e-05,
+      "loss": 0.0372538,
+      "step": 9728
+    },
+    {
+      "epoch": 19.458,
+      "grad_norm": 1.6880683898925781,
+      "learning_rate": 2e-05,
+      "loss": 0.03757115,
+      "step": 9729
+    },
+    {
+      "epoch": 19.46,
+      "grad_norm": 1.3081451654434204,
+      "learning_rate": 2e-05,
+      "loss": 0.04177973,
+      "step": 9730
+    },
+    {
+      "epoch": 19.462,
+      "grad_norm": 1.5396312475204468,
+      "learning_rate": 2e-05,
+      "loss": 0.0427272,
+      "step": 9731
+    },
+    {
+      "epoch": 19.464,
+      "grad_norm": 1.6483396291732788,
+      "learning_rate": 2e-05,
+      "loss": 0.04573702,
+      "step": 9732
+    },
+    {
+      "epoch": 19.466,
+      "grad_norm": 2.9675261974334717,
+      "learning_rate": 2e-05,
+      "loss": 0.05595337,
+      "step": 9733
+    },
+    {
+      "epoch": 19.468,
+      "grad_norm": 1.0395759344100952,
+      "learning_rate": 2e-05,
+      "loss": 0.03035801,
+      "step": 9734
+    },
+    {
+      "epoch": 19.47,
+      "grad_norm": 1.7170274257659912,
+      "learning_rate": 2e-05,
+      "loss": 0.04833669,
+      "step": 9735
+    },
+    {
+      "epoch": 19.472,
+      "grad_norm": 0.9310649633407593,
+      "learning_rate": 2e-05,
+      "loss": 0.03122351,
+      "step": 9736
+    },
+    {
+      "epoch": 19.474,
+      "grad_norm": 1.8986892700195312,
+      "learning_rate": 2e-05,
+      "loss": 0.0391847,
+      "step": 9737
+    },
+    {
+      "epoch": 19.476,
+      "grad_norm": 1.0801862478256226,
+      "learning_rate": 2e-05,
+      "loss": 0.04195695,
+      "step": 9738
+    },
+    {
+      "epoch": 19.478,
+      "grad_norm": 2.033712148666382,
+      "learning_rate": 2e-05,
+      "loss": 0.03731752,
+      "step": 9739
+    },
+    {
+      "epoch": 19.48,
+      "grad_norm": 1.0139687061309814,
+      "learning_rate": 2e-05,
+      "loss": 0.03789478,
+      "step": 9740
+    },
+    {
+      "epoch": 19.482,
+      "grad_norm": 1.770564317703247,
+      "learning_rate": 2e-05,
+      "loss": 0.04420698,
+      "step": 9741
+    },
+    {
+      "epoch": 19.484,
+      "grad_norm": 1.256616234779358,
+      "learning_rate": 2e-05,
+      "loss": 0.03891748,
+      "step": 9742
+    },
+    {
+      "epoch": 19.486,
+      "grad_norm": 1.4175989627838135,
+      "learning_rate": 2e-05,
+      "loss": 0.03451204,
+      "step": 9743
+    },
+    {
+      "epoch": 19.488,
+      "grad_norm": 1.5900920629501343,
+      "learning_rate": 2e-05,
+      "loss": 0.05277488,
+      "step": 9744
+    },
+    {
+      "epoch": 19.49,
+      "grad_norm": 1.8523023128509521,
+      "learning_rate": 2e-05,
+      "loss": 0.04177346,
+      "step": 9745
+    },
+    {
+      "epoch": 19.492,
+      "grad_norm": 1.2272021770477295,
+      "learning_rate": 2e-05,
+      "loss": 0.02621971,
+      "step": 9746
+    },
+    {
+      "epoch": 19.494,
+      "grad_norm": 2.3542540073394775,
+      "learning_rate": 2e-05,
+      "loss": 0.04458754,
+      "step": 9747
+    },
+    {
+      "epoch": 19.496,
+      "grad_norm": 1.0178302526474,
+      "learning_rate": 2e-05,
+      "loss": 0.02959592,
+      "step": 9748
+    },
+    {
+      "epoch": 19.498,
+      "grad_norm": 1.0469385385513306,
+      "learning_rate": 2e-05,
+      "loss": 0.03299112,
+      "step": 9749
+    },
+    {
+      "epoch": 19.5,
+      "grad_norm": 1.7686867713928223,
+      "learning_rate": 2e-05,
+      "loss": 0.03709473,
+      "step": 9750
+    },
+    {
+      "epoch": 19.502,
+      "grad_norm": 1.1636873483657837,
+      "learning_rate": 2e-05,
+      "loss": 0.0360143,
+      "step": 9751
+    },
+    {
+      "epoch": 19.504,
+      "grad_norm": 1.5657356977462769,
+      "learning_rate": 2e-05,
+      "loss": 0.02885246,
+      "step": 9752
+    },
+    {
+      "epoch": 19.506,
+      "grad_norm": 1.2220760583877563,
+      "learning_rate": 2e-05,
+      "loss": 0.03826221,
+      "step": 9753
+    },
+    {
+      "epoch": 19.508,
+      "grad_norm": 1.124636173248291,
+      "learning_rate": 2e-05,
+      "loss": 0.03203499,
+      "step": 9754
+    },
+    {
+      "epoch": 19.51,
+      "grad_norm": 1.681024193763733,
+      "learning_rate": 2e-05,
+      "loss": 0.03180341,
+      "step": 9755
+    },
+    {
+      "epoch": 19.512,
+      "grad_norm": 2.359029531478882,
+      "learning_rate": 2e-05,
+      "loss": 0.04386077,
+      "step": 9756
+    },
+    {
+      "epoch": 19.514,
+      "grad_norm": 1.4641876220703125,
+      "learning_rate": 2e-05,
+      "loss": 0.03449754,
+      "step": 9757
+    },
+    {
+      "epoch": 19.516,
+      "grad_norm": 1.5766197443008423,
+      "learning_rate": 2e-05,
+      "loss": 0.02979608,
+      "step": 9758
+    },
+    {
+      "epoch": 19.518,
+      "grad_norm": 2.0075721740722656,
+      "learning_rate": 2e-05,
+      "loss": 0.0430277,
+      "step": 9759
+    },
+    {
+      "epoch": 19.52,
+      "grad_norm": 1.359182596206665,
+      "learning_rate": 2e-05,
+      "loss": 0.03539016,
+      "step": 9760
+    },
+    {
+      "epoch": 19.522,
+      "grad_norm": 1.0913418531417847,
+      "learning_rate": 2e-05,
+      "loss": 0.03243357,
+      "step": 9761
+    },
+    {
+      "epoch": 19.524,
+      "grad_norm": 1.3150811195373535,
+      "learning_rate": 2e-05,
+      "loss": 0.03395069,
+      "step": 9762
+    },
+    {
+      "epoch": 19.526,
+      "grad_norm": 1.6575753688812256,
+      "learning_rate": 2e-05,
+      "loss": 0.03828706,
+      "step": 9763
+    },
+    {
+      "epoch": 19.528,
+      "grad_norm": 1.700451374053955,
+      "learning_rate": 2e-05,
+      "loss": 0.05022026,
+      "step": 9764
+    },
+    {
+      "epoch": 19.53,
+      "grad_norm": 1.0677083730697632,
+      "learning_rate": 2e-05,
+      "loss": 0.03512496,
+      "step": 9765
+    },
+    {
+      "epoch": 19.532,
+      "grad_norm": 1.0207821130752563,
+      "learning_rate": 2e-05,
+      "loss": 0.03350664,
+      "step": 9766
+    },
+    {
+      "epoch": 19.534,
+      "grad_norm": 1.220189094543457,
+      "learning_rate": 2e-05,
+      "loss": 0.03028592,
+      "step": 9767
+    },
+    {
+      "epoch": 19.536,
+      "grad_norm": 1.0478119850158691,
+      "learning_rate": 2e-05,
+      "loss": 0.02295764,
+      "step": 9768
+    },
+    {
+      "epoch": 19.538,
+      "grad_norm": 1.4206011295318604,
+      "learning_rate": 2e-05,
+      "loss": 0.04312293,
+      "step": 9769
+    },
+    {
+      "epoch": 19.54,
+      "grad_norm": 2.566202163696289,
+      "learning_rate": 2e-05,
+      "loss": 0.04163282,
+      "step": 9770
+    },
+    {
+      "epoch": 19.542,
+      "grad_norm": 1.8683357238769531,
+      "learning_rate": 2e-05,
+      "loss": 0.02420815,
+      "step": 9771
+    },
+    {
+      "epoch": 19.544,
+      "grad_norm": 1.4970320463180542,
+      "learning_rate": 2e-05,
+      "loss": 0.04130397,
+      "step": 9772
+    },
+    {
+      "epoch": 19.546,
+      "grad_norm": 1.3153479099273682,
+      "learning_rate": 2e-05,
+      "loss": 0.04295473,
+      "step": 9773
+    },
+    {
+      "epoch": 19.548000000000002,
+      "grad_norm": 1.5751969814300537,
+      "learning_rate": 2e-05,
+      "loss": 0.05208491,
+      "step": 9774
+    },
+    {
+      "epoch": 19.55,
+      "grad_norm": 1.0567619800567627,
+      "learning_rate": 2e-05,
+      "loss": 0.02717445,
+      "step": 9775
+    },
+    {
+      "epoch": 19.552,
+      "grad_norm": 1.5886588096618652,
+      "learning_rate": 2e-05,
+      "loss": 0.05369675,
+      "step": 9776
+    },
+    {
+      "epoch": 19.554,
+      "grad_norm": 1.6979105472564697,
+      "learning_rate": 2e-05,
+      "loss": 0.04253804,
+      "step": 9777
+    },
+    {
+      "epoch": 19.556,
+      "grad_norm": 1.3951151371002197,
+      "learning_rate": 2e-05,
+      "loss": 0.04301263,
+      "step": 9778
+    },
+    {
+      "epoch": 19.558,
+      "grad_norm": 3.395491361618042,
+      "learning_rate": 2e-05,
+      "loss": 0.03041153,
+      "step": 9779
+    },
+    {
+      "epoch": 19.56,
+      "grad_norm": 1.3777649402618408,
+      "learning_rate": 2e-05,
+      "loss": 0.04786427,
+      "step": 9780
+    },
+    {
+      "epoch": 19.562,
+      "grad_norm": 1.122922420501709,
+      "learning_rate": 2e-05,
+      "loss": 0.03683935,
+      "step": 9781
+    },
+    {
+      "epoch": 19.564,
+      "grad_norm": 2.4130606651306152,
+      "learning_rate": 2e-05,
+      "loss": 0.04994403,
+      "step": 9782
+    },
+    {
+      "epoch": 19.566,
+      "grad_norm": 2.533428907394409,
+      "learning_rate": 2e-05,
+      "loss": 0.03640559,
+      "step": 9783
+    },
+    {
+      "epoch": 19.568,
+      "grad_norm": 1.6431173086166382,
+      "learning_rate": 2e-05,
+      "loss": 0.03542903,
+      "step": 9784
+    },
+    {
+      "epoch": 19.57,
+      "grad_norm": 0.8823850750923157,
+      "learning_rate": 2e-05,
+      "loss": 0.02428854,
+      "step": 9785
+    },
+    {
+      "epoch": 19.572,
+      "grad_norm": 1.7447413206100464,
+      "learning_rate": 2e-05,
+      "loss": 0.05290275,
+      "step": 9786
+    },
+    {
+      "epoch": 19.574,
+      "grad_norm": 1.2419053316116333,
+      "learning_rate": 2e-05,
+      "loss": 0.04086098,
+      "step": 9787
+    },
+    {
+      "epoch": 19.576,
+      "grad_norm": 2.450349807739258,
+      "learning_rate": 2e-05,
+      "loss": 0.0346169,
+      "step": 9788
+    },
+    {
+      "epoch": 19.578,
+      "grad_norm": 1.867328405380249,
+      "learning_rate": 2e-05,
+      "loss": 0.04954457,
+      "step": 9789
+    },
+    {
+      "epoch": 19.58,
+      "grad_norm": 1.358357310295105,
+      "learning_rate": 2e-05,
+      "loss": 0.04382439,
+      "step": 9790
+    },
+    {
+      "epoch": 19.582,
+      "grad_norm": 1.9822919368743896,
+      "learning_rate": 2e-05,
+      "loss": 0.03268323,
+      "step": 9791
+    },
+    {
+      "epoch": 19.584,
+      "grad_norm": 1.2838702201843262,
+      "learning_rate": 2e-05,
+      "loss": 0.02432247,
+      "step": 9792
+    },
+    {
+      "epoch": 19.586,
+      "grad_norm": 1.2836079597473145,
+      "learning_rate": 2e-05,
+      "loss": 0.03453143,
+      "step": 9793
+    },
+    {
+      "epoch": 19.588,
+      "grad_norm": 1.7768176794052124,
+      "learning_rate": 2e-05,
+      "loss": 0.03774252,
+      "step": 9794
+    },
+    {
+      "epoch": 19.59,
+      "grad_norm": 1.0157089233398438,
+      "learning_rate": 2e-05,
+      "loss": 0.03381804,
+      "step": 9795
+    },
+    {
+      "epoch": 19.592,
+      "grad_norm": 2.1426620483398438,
+      "learning_rate": 2e-05,
+      "loss": 0.04200416,
+      "step": 9796
+    },
+    {
+      "epoch": 19.594,
+      "grad_norm": 1.004624605178833,
+      "learning_rate": 2e-05,
+      "loss": 0.02966635,
+      "step": 9797
+    },
+    {
+      "epoch": 19.596,
+      "grad_norm": 1.0430629253387451,
+      "learning_rate": 2e-05,
+      "loss": 0.03210926,
+      "step": 9798
+    },
+    {
+      "epoch": 19.598,
+      "grad_norm": 0.9733067750930786,
+      "learning_rate": 2e-05,
+      "loss": 0.02974671,
+      "step": 9799
+    },
+    {
+      "epoch": 19.6,
+      "grad_norm": 2.1266605854034424,
+      "learning_rate": 2e-05,
+      "loss": 0.04483049,
+      "step": 9800
+    },
+    {
+      "epoch": 19.602,
+      "grad_norm": 1.2150872945785522,
+      "learning_rate": 2e-05,
+      "loss": 0.04423954,
+      "step": 9801
+    },
+    {
+      "epoch": 19.604,
+      "grad_norm": 1.1437227725982666,
+      "learning_rate": 2e-05,
+      "loss": 0.03814924,
+      "step": 9802
+    },
+    {
+      "epoch": 19.606,
+      "grad_norm": 1.234365701675415,
+      "learning_rate": 2e-05,
+      "loss": 0.04211161,
+      "step": 9803
+    },
+    {
+      "epoch": 19.608,
+      "grad_norm": 1.2904542684555054,
+      "learning_rate": 2e-05,
+      "loss": 0.039315,
+      "step": 9804
+    },
+    {
+      "epoch": 19.61,
+      "grad_norm": 1.0852253437042236,
+      "learning_rate": 2e-05,
+      "loss": 0.03015836,
+      "step": 9805
+    },
+    {
+      "epoch": 19.612,
+      "grad_norm": 2.3996999263763428,
+      "learning_rate": 2e-05,
+      "loss": 0.04053651,
+      "step": 9806
+    },
+    {
+      "epoch": 19.614,
+      "grad_norm": 1.0746995210647583,
+      "learning_rate": 2e-05,
+      "loss": 0.03122765,
+      "step": 9807
+    },
+    {
+      "epoch": 19.616,
+      "grad_norm": 1.0592286586761475,
+      "learning_rate": 2e-05,
+      "loss": 0.03569851,
+      "step": 9808
+    },
+    {
+      "epoch": 19.618,
+      "grad_norm": 1.2815927267074585,
+      "learning_rate": 2e-05,
+      "loss": 0.03232668,
+      "step": 9809
+    },
+    {
+      "epoch": 19.62,
+      "grad_norm": 1.2451739311218262,
+      "learning_rate": 2e-05,
+      "loss": 0.02683158,
+      "step": 9810
+    },
+    {
+      "epoch": 19.622,
+      "grad_norm": 0.9535860419273376,
+      "learning_rate": 2e-05,
+      "loss": 0.02343744,
+      "step": 9811
+    },
+    {
+      "epoch": 19.624,
+      "grad_norm": 2.1150941848754883,
+      "learning_rate": 2e-05,
+      "loss": 0.03646648,
+      "step": 9812
+    },
+    {
+      "epoch": 19.626,
+      "grad_norm": 1.4279240369796753,
+      "learning_rate": 2e-05,
+      "loss": 0.03931186,
+      "step": 9813
+    },
+    {
+      "epoch": 19.628,
+      "grad_norm": 1.9593764543533325,
+      "learning_rate": 2e-05,
+      "loss": 0.04083652,
+      "step": 9814
+    },
+    {
+      "epoch": 19.63,
+      "grad_norm": 1.289267659187317,
+      "learning_rate": 2e-05,
+      "loss": 0.03806636,
+      "step": 9815
+    },
+    {
+      "epoch": 19.632,
+      "grad_norm": 2.0705997943878174,
+      "learning_rate": 2e-05,
+      "loss": 0.0471487,
+      "step": 9816
+    },
+    {
+      "epoch": 19.634,
+      "grad_norm": 1.577562928199768,
+      "learning_rate": 2e-05,
+      "loss": 0.02332294,
+      "step": 9817
+    },
+    {
+      "epoch": 19.636,
+      "grad_norm": 1.0667752027511597,
+      "learning_rate": 2e-05,
+      "loss": 0.03316651,
+      "step": 9818
+    },
+    {
+      "epoch": 19.638,
+      "grad_norm": 1.272627592086792,
+      "learning_rate": 2e-05,
+      "loss": 0.0328108,
+      "step": 9819
+    },
+    {
+      "epoch": 19.64,
+      "grad_norm": 1.1442145109176636,
+      "learning_rate": 2e-05,
+      "loss": 0.03424655,
+      "step": 9820
+    },
+    {
+      "epoch": 19.642,
+      "grad_norm": 0.977202832698822,
+      "learning_rate": 2e-05,
+      "loss": 0.02730237,
+      "step": 9821
+    },
+    {
+      "epoch": 19.644,
+      "grad_norm": 1.2419651746749878,
+      "learning_rate": 2e-05,
+      "loss": 0.03938533,
+      "step": 9822
+    },
+    {
+      "epoch": 19.646,
+      "grad_norm": 1.5671879053115845,
+      "learning_rate": 2e-05,
+      "loss": 0.03334674,
+      "step": 9823
+    },
+    {
+      "epoch": 19.648,
+      "grad_norm": 1.2008931636810303,
+      "learning_rate": 2e-05,
+      "loss": 0.03198599,
+      "step": 9824
+    },
+    {
+      "epoch": 19.65,
+      "grad_norm": 0.8605839610099792,
+      "learning_rate": 2e-05,
+      "loss": 0.02071075,
+      "step": 9825
+    },
+    {
+      "epoch": 19.652,
+      "grad_norm": 1.1936160326004028,
+      "learning_rate": 2e-05,
+      "loss": 0.041041,
+      "step": 9826
+    },
+    {
+      "epoch": 19.654,
+      "grad_norm": 0.7941514253616333,
+      "learning_rate": 2e-05,
+      "loss": 0.01732956,
+      "step": 9827
+    },
+    {
+      "epoch": 19.656,
+      "grad_norm": 0.9034144282341003,
+      "learning_rate": 2e-05,
+      "loss": 0.02284335,
+      "step": 9828
+    },
+    {
+      "epoch": 19.658,
+      "grad_norm": 2.9886972904205322,
+      "learning_rate": 2e-05,
+      "loss": 0.05936865,
+      "step": 9829
+    },
+    {
+      "epoch": 19.66,
+      "grad_norm": 1.7433563470840454,
+      "learning_rate": 2e-05,
+      "loss": 0.04660515,
+      "step": 9830
+    },
+    {
+      "epoch": 19.662,
+      "grad_norm": 1.4639521837234497,
+      "learning_rate": 2e-05,
+      "loss": 0.03722664,
+      "step": 9831
+    },
+    {
+      "epoch": 19.664,
+      "grad_norm": 1.6266924142837524,
+      "learning_rate": 2e-05,
+      "loss": 0.04071777,
+      "step": 9832
+    },
+    {
+      "epoch": 19.666,
+      "grad_norm": 1.52934730052948,
+      "learning_rate": 2e-05,
+      "loss": 0.03618839,
+      "step": 9833
+    },
+    {
+      "epoch": 19.668,
+      "grad_norm": 1.321384310722351,
+      "learning_rate": 2e-05,
+      "loss": 0.04987688,
+      "step": 9834
+    },
+    {
+      "epoch": 19.67,
+      "grad_norm": 1.1635421514511108,
+      "learning_rate": 2e-05,
+      "loss": 0.03412395,
+      "step": 9835
+    },
+    {
+      "epoch": 19.672,
+      "grad_norm": 2.087972402572632,
+      "learning_rate": 2e-05,
+      "loss": 0.02932746,
+      "step": 9836
+    },
+    {
+      "epoch": 19.674,
+      "grad_norm": 1.8116962909698486,
+      "learning_rate": 2e-05,
+      "loss": 0.05201422,
+      "step": 9837
+    },
+    {
+      "epoch": 19.676,
+      "grad_norm": 1.7317609786987305,
+      "learning_rate": 2e-05,
+      "loss": 0.03608788,
+      "step": 9838
+    },
+    {
+      "epoch": 19.678,
+      "grad_norm": 1.6292537450790405,
+      "learning_rate": 2e-05,
+      "loss": 0.03296152,
+      "step": 9839
+    },
+    {
+      "epoch": 19.68,
+      "grad_norm": 1.4448356628417969,
+      "learning_rate": 2e-05,
+      "loss": 0.04338595,
+      "step": 9840
+    },
+    {
+      "epoch": 19.682,
+      "grad_norm": 1.1970990896224976,
+      "learning_rate": 2e-05,
+      "loss": 0.02565513,
+      "step": 9841
+    },
+    {
+      "epoch": 19.684,
+      "grad_norm": 1.4799765348434448,
+      "learning_rate": 2e-05,
+      "loss": 0.04556567,
+      "step": 9842
+    },
+    {
+      "epoch": 19.686,
+      "grad_norm": 1.1363059282302856,
+      "learning_rate": 2e-05,
+      "loss": 0.03603154,
+      "step": 9843
+    },
+    {
+      "epoch": 19.688,
+      "grad_norm": 2.854661703109741,
+      "learning_rate": 2e-05,
+      "loss": 0.05047087,
+      "step": 9844
+    },
+    {
+      "epoch": 19.69,
+      "grad_norm": 0.8889420032501221,
+      "learning_rate": 2e-05,
+      "loss": 0.02277106,
+      "step": 9845
+    },
+    {
+      "epoch": 19.692,
+      "grad_norm": 1.6678924560546875,
+      "learning_rate": 2e-05,
+      "loss": 0.03258391,
+      "step": 9846
+    },
+    {
+      "epoch": 19.694,
+      "grad_norm": 3.123685598373413,
+      "learning_rate": 2e-05,
+      "loss": 0.03615994,
+      "step": 9847
+    },
+    {
+      "epoch": 19.696,
+      "grad_norm": 1.1917728185653687,
+      "learning_rate": 2e-05,
+      "loss": 0.0365363,
+      "step": 9848
+    },
+    {
+      "epoch": 19.698,
+      "grad_norm": 1.2014756202697754,
+      "learning_rate": 2e-05,
+      "loss": 0.03089717,
+      "step": 9849
+    },
+    {
+      "epoch": 19.7,
+      "grad_norm": 2.1376969814300537,
+      "learning_rate": 2e-05,
+      "loss": 0.04704689,
+      "step": 9850
+    },
+    {
+      "epoch": 19.701999999999998,
+      "grad_norm": 1.4255774021148682,
+      "learning_rate": 2e-05,
+      "loss": 0.02486222,
+      "step": 9851
+    },
+    {
+      "epoch": 19.704,
+      "grad_norm": 1.1981037855148315,
+      "learning_rate": 2e-05,
+      "loss": 0.04301386,
+      "step": 9852
+    },
+    {
+      "epoch": 19.706,
+      "grad_norm": 1.5258287191390991,
+      "learning_rate": 2e-05,
+      "loss": 0.05485792,
+      "step": 9853
+    },
+    {
+      "epoch": 19.708,
+      "grad_norm": 1.6070016622543335,
+      "learning_rate": 2e-05,
+      "loss": 0.03380585,
+      "step": 9854
+    },
+    {
+      "epoch": 19.71,
+      "grad_norm": 1.0304168462753296,
+      "learning_rate": 2e-05,
+      "loss": 0.02755873,
+      "step": 9855
+    },
+    {
+      "epoch": 19.712,
+      "grad_norm": 2.085618257522583,
+      "learning_rate": 2e-05,
+      "loss": 0.04198353,
+      "step": 9856
+    },
+    {
+      "epoch": 19.714,
+      "grad_norm": 1.2231016159057617,
+      "learning_rate": 2e-05,
+      "loss": 0.03908198,
+      "step": 9857
+    },
+    {
+      "epoch": 19.716,
+      "grad_norm": 2.0941994190216064,
+      "learning_rate": 2e-05,
+      "loss": 0.04153078,
+      "step": 9858
+    },
+    {
+      "epoch": 19.718,
+      "grad_norm": 1.3140674829483032,
+      "learning_rate": 2e-05,
+      "loss": 0.03461843,
+      "step": 9859
+    },
+    {
+      "epoch": 19.72,
+      "grad_norm": 2.5073509216308594,
+      "learning_rate": 2e-05,
+      "loss": 0.05611764,
+      "step": 9860
+    },
+    {
+      "epoch": 19.722,
+      "grad_norm": 1.2403875589370728,
+      "learning_rate": 2e-05,
+      "loss": 0.04040193,
+      "step": 9861
+    },
+    {
+      "epoch": 19.724,
+      "grad_norm": 1.2236120700836182,
+      "learning_rate": 2e-05,
+      "loss": 0.04124648,
+      "step": 9862
+    },
+    {
+      "epoch": 19.726,
+      "grad_norm": 1.410003662109375,
+      "learning_rate": 2e-05,
+      "loss": 0.0384649,
+      "step": 9863
+    },
+    {
+      "epoch": 19.728,
+      "grad_norm": 1.1842060089111328,
+      "learning_rate": 2e-05,
+      "loss": 0.03703492,
+      "step": 9864
+    },
+    {
+      "epoch": 19.73,
+      "grad_norm": 1.596753478050232,
+      "learning_rate": 2e-05,
+      "loss": 0.03340092,
+      "step": 9865
+    },
+    {
+      "epoch": 19.732,
+      "grad_norm": 1.6975880861282349,
+      "learning_rate": 2e-05,
+      "loss": 0.04212178,
+      "step": 9866
+    },
+    {
+      "epoch": 19.734,
+      "grad_norm": 2.210099458694458,
+      "learning_rate": 2e-05,
+      "loss": 0.03261279,
+      "step": 9867
+    },
+    {
+      "epoch": 19.736,
+      "grad_norm": 1.1055867671966553,
+      "learning_rate": 2e-05,
+      "loss": 0.03092555,
+      "step": 9868
+    },
+    {
+      "epoch": 19.738,
+      "grad_norm": 1.1590442657470703,
+      "learning_rate": 2e-05,
+      "loss": 0.02638486,
+      "step": 9869
+    },
+    {
+      "epoch": 19.74,
+      "grad_norm": 1.1977529525756836,
+      "learning_rate": 2e-05,
+      "loss": 0.03209082,
+      "step": 9870
+    },
+    {
+      "epoch": 19.742,
+      "grad_norm": 1.3903778791427612,
+      "learning_rate": 2e-05,
+      "loss": 0.03176203,
+      "step": 9871
+    },
+    {
+      "epoch": 19.744,
+      "grad_norm": 1.7422199249267578,
+      "learning_rate": 2e-05,
+      "loss": 0.03671364,
+      "step": 9872
+    },
+    {
+      "epoch": 19.746,
+      "grad_norm": 1.0997118949890137,
+      "learning_rate": 2e-05,
+      "loss": 0.03464361,
+      "step": 9873
+    },
+    {
+      "epoch": 19.748,
+      "grad_norm": 1.4066072702407837,
+      "learning_rate": 2e-05,
+      "loss": 0.03441656,
+      "step": 9874
+    },
+    {
+      "epoch": 19.75,
+      "grad_norm": 1.0262517929077148,
+      "learning_rate": 2e-05,
+      "loss": 0.03322373,
+      "step": 9875
+    },
+    {
+      "epoch": 19.752,
+      "grad_norm": 1.3730692863464355,
+      "learning_rate": 2e-05,
+      "loss": 0.04439168,
+      "step": 9876
+    },
+    {
+      "epoch": 19.754,
+      "grad_norm": 1.2978415489196777,
+      "learning_rate": 2e-05,
+      "loss": 0.04093964,
+      "step": 9877
+    },
+    {
+      "epoch": 19.756,
+      "grad_norm": 1.0929715633392334,
+      "learning_rate": 2e-05,
+      "loss": 0.0319971,
+      "step": 9878
+    },
+    {
+      "epoch": 19.758,
+      "grad_norm": 1.8838789463043213,
+      "learning_rate": 2e-05,
+      "loss": 0.0498081,
+      "step": 9879
+    },
+    {
+      "epoch": 19.76,
+      "grad_norm": 1.343527913093567,
+      "learning_rate": 2e-05,
+      "loss": 0.03659369,
+      "step": 9880
+    },
+    {
+      "epoch": 19.762,
+      "grad_norm": 2.9458236694335938,
+      "learning_rate": 2e-05,
+      "loss": 0.05308362,
+      "step": 9881
+    },
+    {
+      "epoch": 19.764,
+      "grad_norm": 1.1302599906921387,
+      "learning_rate": 2e-05,
+      "loss": 0.03225252,
+      "step": 9882
+    },
+    {
+      "epoch": 19.766,
+      "grad_norm": 1.005309820175171,
+      "learning_rate": 2e-05,
+      "loss": 0.02554948,
+      "step": 9883
+    },
+    {
+      "epoch": 19.768,
+      "grad_norm": 1.148569107055664,
+      "learning_rate": 2e-05,
+      "loss": 0.0364852,
+      "step": 9884
+    },
+    {
+      "epoch": 19.77,
+      "grad_norm": 1.1067266464233398,
+      "learning_rate": 2e-05,
+      "loss": 0.03569014,
+      "step": 9885
+    },
+    {
+      "epoch": 19.772,
+      "grad_norm": 1.2811511754989624,
+      "learning_rate": 2e-05,
+      "loss": 0.02505126,
+      "step": 9886
+    },
+    {
+      "epoch": 19.774,
+      "grad_norm": 1.8366814851760864,
+      "learning_rate": 2e-05,
+      "loss": 0.04801732,
+      "step": 9887
+    },
+    {
+      "epoch": 19.776,
+      "grad_norm": 1.0430152416229248,
+      "learning_rate": 2e-05,
+      "loss": 0.02715734,
+      "step": 9888
+    },
+    {
+      "epoch": 19.778,
+      "grad_norm": 1.6979142427444458,
+      "learning_rate": 2e-05,
+      "loss": 0.02943612,
+      "step": 9889
+    },
+    {
+      "epoch": 19.78,
+      "grad_norm": 2.1606314182281494,
+      "learning_rate": 2e-05,
+      "loss": 0.05470781,
+      "step": 9890
+    },
+    {
+      "epoch": 19.782,
+      "grad_norm": 1.1862667798995972,
+      "learning_rate": 2e-05,
+      "loss": 0.0372127,
+      "step": 9891
+    },
+    {
+      "epoch": 19.784,
+      "grad_norm": 1.5462552309036255,
+      "learning_rate": 2e-05,
+      "loss": 0.04436889,
+      "step": 9892
+    },
+    {
+      "epoch": 19.786,
+      "grad_norm": 1.1272681951522827,
+      "learning_rate": 2e-05,
+      "loss": 0.02969773,
+      "step": 9893
+    },
+    {
+      "epoch": 19.788,
+      "grad_norm": 1.8199870586395264,
+      "learning_rate": 2e-05,
+      "loss": 0.04257311,
+      "step": 9894
+    },
+    {
+      "epoch": 19.79,
+      "grad_norm": 1.1364978551864624,
+      "learning_rate": 2e-05,
+      "loss": 0.03960226,
+      "step": 9895
+    },
+    {
+      "epoch": 19.792,
+      "grad_norm": 5.6248908042907715,
+      "learning_rate": 2e-05,
+      "loss": 0.05272412,
+      "step": 9896
+    },
+    {
+      "epoch": 19.794,
+      "grad_norm": 1.9296268224716187,
+      "learning_rate": 2e-05,
+      "loss": 0.0427717,
+      "step": 9897
+    },
+    {
+      "epoch": 19.796,
+      "grad_norm": 1.5250190496444702,
+      "learning_rate": 2e-05,
+      "loss": 0.04301464,
+      "step": 9898
+    },
+    {
+      "epoch": 19.798000000000002,
+      "grad_norm": 3.7405781745910645,
+      "learning_rate": 2e-05,
+      "loss": 0.03769503,
+      "step": 9899
+    },
+    {
+      "epoch": 19.8,
+      "grad_norm": 1.0620787143707275,
+      "learning_rate": 2e-05,
+      "loss": 0.02631702,
+      "step": 9900
+    },
+    {
+      "epoch": 19.802,
+      "grad_norm": 1.9227813482284546,
+      "learning_rate": 2e-05,
+      "loss": 0.03936617,
+      "step": 9901
+    },
+    {
+      "epoch": 19.804,
+      "grad_norm": 1.4326848983764648,
+      "learning_rate": 2e-05,
+      "loss": 0.03811381,
+      "step": 9902
+    },
+    {
+      "epoch": 19.806,
+      "grad_norm": 0.9656258821487427,
+      "learning_rate": 2e-05,
+      "loss": 0.02461326,
+      "step": 9903
+    },
+    {
+      "epoch": 19.808,
+      "grad_norm": 1.468076229095459,
+      "learning_rate": 2e-05,
+      "loss": 0.04942961,
+      "step": 9904
+    },
+    {
+      "epoch": 19.81,
+      "grad_norm": 1.0729457139968872,
+      "learning_rate": 2e-05,
+      "loss": 0.03775813,
+      "step": 9905
+    },
+    {
+      "epoch": 19.812,
+      "grad_norm": 2.371782064437866,
+      "learning_rate": 2e-05,
+      "loss": 0.04209847,
+      "step": 9906
+    },
+    {
+      "epoch": 19.814,
+      "grad_norm": 1.1211527585983276,
+      "learning_rate": 2e-05,
+      "loss": 0.03618193,
+      "step": 9907
+    },
+    {
+      "epoch": 19.816,
+      "grad_norm": 1.2822222709655762,
+      "learning_rate": 2e-05,
+      "loss": 0.02676849,
+      "step": 9908
+    },
+    {
+      "epoch": 19.818,
+      "grad_norm": 1.2275272607803345,
+      "learning_rate": 2e-05,
+      "loss": 0.03364244,
+      "step": 9909
+    },
+    {
+      "epoch": 19.82,
+      "grad_norm": 1.4807791709899902,
+      "learning_rate": 2e-05,
+      "loss": 0.03341034,
+      "step": 9910
+    },
+    {
+      "epoch": 19.822,
+      "grad_norm": 0.8294656872749329,
+      "learning_rate": 2e-05,
+      "loss": 0.02359098,
+      "step": 9911
+    },
+    {
+      "epoch": 19.824,
+      "grad_norm": 1.1592644453048706,
+      "learning_rate": 2e-05,
+      "loss": 0.0353507,
+      "step": 9912
+    },
+    {
+      "epoch": 19.826,
+      "grad_norm": 1.078616976737976,
+      "learning_rate": 2e-05,
+      "loss": 0.02933605,
+      "step": 9913
+    },
+    {
+      "epoch": 19.828,
+      "grad_norm": 1.74703848361969,
+      "learning_rate": 2e-05,
+      "loss": 0.03395237,
+      "step": 9914
+    },
+    {
+      "epoch": 19.83,
+      "grad_norm": 0.9906914830207825,
+      "learning_rate": 2e-05,
+      "loss": 0.03834459,
+      "step": 9915
+    },
+    {
+      "epoch": 19.832,
+      "grad_norm": 1.5938894748687744,
+      "learning_rate": 2e-05,
+      "loss": 0.04390221,
+      "step": 9916
+    },
+    {
+      "epoch": 19.834,
+      "grad_norm": 0.8304893374443054,
+      "learning_rate": 2e-05,
+      "loss": 0.02339101,
+      "step": 9917
+    },
+    {
+      "epoch": 19.836,
+      "grad_norm": 2.1542248725891113,
+      "learning_rate": 2e-05,
+      "loss": 0.04835535,
+      "step": 9918
+    },
+    {
+      "epoch": 19.838,
+      "grad_norm": 1.7245725393295288,
+      "learning_rate": 2e-05,
+      "loss": 0.04382911,
+      "step": 9919
+    },
+    {
+      "epoch": 19.84,
+      "grad_norm": 1.0681260824203491,
+      "learning_rate": 2e-05,
+      "loss": 0.0333902,
+      "step": 9920
+    },
+    {
+      "epoch": 19.842,
+      "grad_norm": 1.5578216314315796,
+      "learning_rate": 2e-05,
+      "loss": 0.03641013,
+      "step": 9921
+    },
+    {
+      "epoch": 19.844,
+      "grad_norm": 0.9164597392082214,
+      "learning_rate": 2e-05,
+      "loss": 0.02537297,
+      "step": 9922
+    },
+    {
+      "epoch": 19.846,
+      "grad_norm": 2.094348192214966,
+      "learning_rate": 2e-05,
+      "loss": 0.04775488,
+      "step": 9923
+    },
+    {
+      "epoch": 19.848,
+      "grad_norm": 2.1640682220458984,
+      "learning_rate": 2e-05,
+      "loss": 0.03262997,
+      "step": 9924
+    },
+    {
+      "epoch": 19.85,
+      "grad_norm": 1.3162791728973389,
+      "learning_rate": 2e-05,
+      "loss": 0.03451164,
+      "step": 9925
+    },
+    {
+      "epoch": 19.852,
+      "grad_norm": 1.047416090965271,
+      "learning_rate": 2e-05,
+      "loss": 0.02686568,
+      "step": 9926
+    },
+    {
+      "epoch": 19.854,
+      "grad_norm": 1.4626446962356567,
+      "learning_rate": 2e-05,
+      "loss": 0.05131056,
+      "step": 9927
+    },
+    {
+      "epoch": 19.856,
+      "grad_norm": 1.1286470890045166,
+      "learning_rate": 2e-05,
+      "loss": 0.03642206,
+      "step": 9928
+    },
+    {
+      "epoch": 19.858,
+      "grad_norm": 1.7644599676132202,
+      "learning_rate": 2e-05,
+      "loss": 0.04343008,
+      "step": 9929
+    },
+    {
+      "epoch": 19.86,
+      "grad_norm": 0.8793272972106934,
+      "learning_rate": 2e-05,
+      "loss": 0.02633665,
+      "step": 9930
+    },
+    {
+      "epoch": 19.862,
+      "grad_norm": 1.3826559782028198,
+      "learning_rate": 2e-05,
+      "loss": 0.03667331,
+      "step": 9931
+    },
+    {
+      "epoch": 19.864,
+      "grad_norm": 1.5260701179504395,
+      "learning_rate": 2e-05,
+      "loss": 0.03625732,
+      "step": 9932
+    },
+    {
+      "epoch": 19.866,
+      "grad_norm": 1.5434454679489136,
+      "learning_rate": 2e-05,
+      "loss": 0.04190221,
+      "step": 9933
+    },
+    {
+      "epoch": 19.868,
+      "grad_norm": 1.4295403957366943,
+      "learning_rate": 2e-05,
+      "loss": 0.03398123,
+      "step": 9934
+    },
+    {
+      "epoch": 19.87,
+      "grad_norm": 1.359604835510254,
+      "learning_rate": 2e-05,
+      "loss": 0.04142676,
+      "step": 9935
+    },
+    {
+      "epoch": 19.872,
+      "grad_norm": 2.0198137760162354,
+      "learning_rate": 2e-05,
+      "loss": 0.05195599,
+      "step": 9936
+    },
+    {
+      "epoch": 19.874,
+      "grad_norm": 0.9484997987747192,
+      "learning_rate": 2e-05,
+      "loss": 0.0277728,
+      "step": 9937
+    },
+    {
+      "epoch": 19.876,
+      "grad_norm": 1.3963159322738647,
+      "learning_rate": 2e-05,
+      "loss": 0.05602345,
+      "step": 9938
+    },
+    {
+      "epoch": 19.878,
+      "grad_norm": 1.419621229171753,
+      "learning_rate": 2e-05,
+      "loss": 0.02756134,
+      "step": 9939
+    },
+    {
+      "epoch": 19.88,
+      "grad_norm": 2.4479565620422363,
+      "learning_rate": 2e-05,
+      "loss": 0.0412005,
+      "step": 9940
+    },
+    {
+      "epoch": 19.882,
+      "grad_norm": 1.392535924911499,
+      "learning_rate": 2e-05,
+      "loss": 0.03324911,
+      "step": 9941
+    },
+    {
+      "epoch": 19.884,
+      "grad_norm": 0.9907611608505249,
+      "learning_rate": 2e-05,
+      "loss": 0.03080793,
+      "step": 9942
+    },
+    {
+      "epoch": 19.886,
+      "grad_norm": 1.6985739469528198,
+      "learning_rate": 2e-05,
+      "loss": 0.04235946,
+      "step": 9943
+    },
+    {
+      "epoch": 19.888,
+      "grad_norm": 1.201528787612915,
+      "learning_rate": 2e-05,
+      "loss": 0.02874477,
+      "step": 9944
+    },
+    {
+      "epoch": 19.89,
+      "grad_norm": 1.272756814956665,
+      "learning_rate": 2e-05,
+      "loss": 0.03114101,
+      "step": 9945
+    },
+    {
+      "epoch": 19.892,
+      "grad_norm": 2.536895513534546,
+      "learning_rate": 2e-05,
+      "loss": 0.04412911,
+      "step": 9946
+    },
+    {
+      "epoch": 19.894,
+      "grad_norm": 1.345763087272644,
+      "learning_rate": 2e-05,
+      "loss": 0.03955846,
+      "step": 9947
+    },
+    {
+      "epoch": 19.896,
+      "grad_norm": 1.703199863433838,
+      "learning_rate": 2e-05,
+      "loss": 0.04491033,
+      "step": 9948
+    },
+    {
+      "epoch": 19.898,
+      "grad_norm": 2.1777453422546387,
+      "learning_rate": 2e-05,
+      "loss": 0.0519994,
+      "step": 9949
+    },
+    {
+      "epoch": 19.9,
+      "grad_norm": 1.9769400358200073,
+      "learning_rate": 2e-05,
+      "loss": 0.04786195,
+      "step": 9950
+    },
+    {
+      "epoch": 19.902,
+      "grad_norm": 0.9934007525444031,
+      "learning_rate": 2e-05,
+      "loss": 0.02457796,
+      "step": 9951
+    },
+    {
+      "epoch": 19.904,
+      "grad_norm": 1.0393855571746826,
+      "learning_rate": 2e-05,
+      "loss": 0.03693616,
+      "step": 9952
+    },
+    {
+      "epoch": 19.906,
+      "grad_norm": 1.4611011743545532,
+      "learning_rate": 2e-05,
+      "loss": 0.03181311,
+      "step": 9953
+    },
+    {
+      "epoch": 19.908,
+      "grad_norm": 1.9323896169662476,
+      "learning_rate": 2e-05,
+      "loss": 0.04420885,
+      "step": 9954
+    },
+    {
+      "epoch": 19.91,
+      "grad_norm": 1.2806192636489868,
+      "learning_rate": 2e-05,
+      "loss": 0.0216569,
+      "step": 9955
+    },
+    {
+      "epoch": 19.912,
+      "grad_norm": 2.6190237998962402,
+      "learning_rate": 2e-05,
+      "loss": 0.04326738,
+      "step": 9956
+    },
+    {
+      "epoch": 19.914,
+      "grad_norm": 1.5687636137008667,
+      "learning_rate": 2e-05,
+      "loss": 0.03183849,
+      "step": 9957
+    },
+    {
+      "epoch": 19.916,
+      "grad_norm": 1.1878702640533447,
+      "learning_rate": 2e-05,
+      "loss": 0.03523634,
+      "step": 9958
+    },
+    {
+      "epoch": 19.918,
+      "grad_norm": 1.3091981410980225,
+      "learning_rate": 2e-05,
+      "loss": 0.03142114,
+      "step": 9959
+    },
+    {
+      "epoch": 19.92,
+      "grad_norm": 2.1375269889831543,
+      "learning_rate": 2e-05,
+      "loss": 0.04402493,
+      "step": 9960
+    },
+    {
+      "epoch": 19.922,
+      "grad_norm": 1.4389981031417847,
+      "learning_rate": 2e-05,
+      "loss": 0.03660287,
+      "step": 9961
+    },
+    {
+      "epoch": 19.924,
+      "grad_norm": 1.4258970022201538,
+      "learning_rate": 2e-05,
+      "loss": 0.03248418,
+      "step": 9962
+    },
+    {
+      "epoch": 19.926,
+      "grad_norm": 1.94060480594635,
+      "learning_rate": 2e-05,
+      "loss": 0.05550197,
+      "step": 9963
+    },
+    {
+      "epoch": 19.928,
+      "grad_norm": 2.0639617443084717,
+      "learning_rate": 2e-05,
+      "loss": 0.04555191,
+      "step": 9964
+    },
+    {
+      "epoch": 19.93,
+      "grad_norm": 1.4543213844299316,
+      "learning_rate": 2e-05,
+      "loss": 0.02980587,
+      "step": 9965
+    },
+    {
+      "epoch": 19.932,
+      "grad_norm": 1.0913479328155518,
+      "learning_rate": 2e-05,
+      "loss": 0.03366387,
+      "step": 9966
+    },
+    {
+      "epoch": 19.934,
+      "grad_norm": 1.459293246269226,
+      "learning_rate": 2e-05,
+      "loss": 0.03716648,
+      "step": 9967
+    },
+    {
+      "epoch": 19.936,
+      "grad_norm": 1.5800535678863525,
+      "learning_rate": 2e-05,
+      "loss": 0.04051695,
+      "step": 9968
+    },
+    {
+      "epoch": 19.938,
+      "grad_norm": 1.6184855699539185,
+      "learning_rate": 2e-05,
+      "loss": 0.0413382,
+      "step": 9969
+    },
+    {
+      "epoch": 19.94,
+      "grad_norm": 1.287768840789795,
+      "learning_rate": 2e-05,
+      "loss": 0.02909556,
+      "step": 9970
+    },
+    {
+      "epoch": 19.942,
+      "grad_norm": 2.3062894344329834,
+      "learning_rate": 2e-05,
+      "loss": 0.04965471,
+      "step": 9971
+    },
+    {
+      "epoch": 19.944,
+      "grad_norm": 1.639174222946167,
+      "learning_rate": 2e-05,
+      "loss": 0.03276775,
+      "step": 9972
+    },
+    {
+      "epoch": 19.946,
+      "grad_norm": 1.1336901187896729,
+      "learning_rate": 2e-05,
+      "loss": 0.03398212,
+      "step": 9973
+    },
+    {
+      "epoch": 19.948,
+      "grad_norm": 1.7506204843521118,
+      "learning_rate": 2e-05,
+      "loss": 0.05709363,
+      "step": 9974
+    },
+    {
+      "epoch": 19.95,
+      "grad_norm": 1.0985453128814697,
+      "learning_rate": 2e-05,
+      "loss": 0.03749266,
+      "step": 9975
+    },
+    {
+      "epoch": 19.951999999999998,
+      "grad_norm": 1.533744215965271,
+      "learning_rate": 2e-05,
+      "loss": 0.04838315,
+      "step": 9976
+    },
+    {
+      "epoch": 19.954,
+      "grad_norm": 1.545249104499817,
+      "learning_rate": 2e-05,
+      "loss": 0.03769622,
+      "step": 9977
+    },
+    {
+      "epoch": 19.956,
+      "grad_norm": 1.0320427417755127,
+      "learning_rate": 2e-05,
+      "loss": 0.02421537,
+      "step": 9978
+    },
+    {
+      "epoch": 19.958,
+      "grad_norm": 1.2649873495101929,
+      "learning_rate": 2e-05,
+      "loss": 0.03971934,
+      "step": 9979
+    },
+    {
+      "epoch": 19.96,
+      "grad_norm": 1.2545596361160278,
+      "learning_rate": 2e-05,
+      "loss": 0.04188591,
+      "step": 9980
+    },
+    {
+      "epoch": 19.962,
+      "grad_norm": 1.0163325071334839,
+      "learning_rate": 2e-05,
+      "loss": 0.02098066,
+      "step": 9981
+    },
+    {
+      "epoch": 19.964,
+      "grad_norm": 0.8341543674468994,
+      "learning_rate": 2e-05,
+      "loss": 0.01667216,
+      "step": 9982
+    },
+    {
+      "epoch": 19.966,
+      "grad_norm": 1.1888387203216553,
+      "learning_rate": 2e-05,
+      "loss": 0.0333435,
+      "step": 9983
+    },
+    {
+      "epoch": 19.968,
+      "grad_norm": 1.6849191188812256,
+      "learning_rate": 2e-05,
+      "loss": 0.04021031,
+      "step": 9984
+    },
+    {
+      "epoch": 19.97,
+      "grad_norm": 1.1306986808776855,
+      "learning_rate": 2e-05,
+      "loss": 0.03524192,
+      "step": 9985
+    },
+    {
+      "epoch": 19.972,
+      "grad_norm": 1.2705979347229004,
+      "learning_rate": 2e-05,
+      "loss": 0.045315,
+      "step": 9986
+    },
+    {
+      "epoch": 19.974,
+      "grad_norm": 3.2684991359710693,
+      "learning_rate": 2e-05,
+      "loss": 0.05361638,
+      "step": 9987
+    },
+    {
+      "epoch": 19.976,
+      "grad_norm": 2.3001787662506104,
+      "learning_rate": 2e-05,
+      "loss": 0.0321941,
+      "step": 9988
+    },
+    {
+      "epoch": 19.978,
+      "grad_norm": 1.2926708459854126,
+      "learning_rate": 2e-05,
+      "loss": 0.03842488,
+      "step": 9989
+    },
+    {
+      "epoch": 19.98,
+      "grad_norm": 1.6358253955841064,
+      "learning_rate": 2e-05,
+      "loss": 0.04522712,
+      "step": 9990
+    },
+    {
+      "epoch": 19.982,
+      "grad_norm": 1.0329595804214478,
+      "learning_rate": 2e-05,
+      "loss": 0.02501506,
+      "step": 9991
+    },
+    {
+      "epoch": 19.984,
+      "grad_norm": 1.3090020418167114,
+      "learning_rate": 2e-05,
+      "loss": 0.03367433,
+      "step": 9992
+    },
+    {
+      "epoch": 19.986,
+      "grad_norm": 1.0016107559204102,
+      "learning_rate": 2e-05,
+      "loss": 0.02978981,
+      "step": 9993
+    },
+    {
+      "epoch": 19.988,
+      "grad_norm": 1.311272382736206,
+      "learning_rate": 2e-05,
+      "loss": 0.04261601,
+      "step": 9994
+    },
+    {
+      "epoch": 19.99,
+      "grad_norm": 0.9918799996376038,
+      "learning_rate": 2e-05,
+      "loss": 0.02678192,
+      "step": 9995
+    },
+    {
+      "epoch": 19.992,
+      "grad_norm": 1.2438549995422363,
+      "learning_rate": 2e-05,
+      "loss": 0.04806288,
+      "step": 9996
+    },
+    {
+      "epoch": 19.994,
+      "grad_norm": 1.1533282995224,
+      "learning_rate": 2e-05,
+      "loss": 0.03929444,
+      "step": 9997
+    },
+    {
+      "epoch": 19.996,
+      "grad_norm": 1.9555519819259644,
+      "learning_rate": 2e-05,
+      "loss": 0.04684892,
+      "step": 9998
+    },
+    {
+      "epoch": 19.998,
+      "grad_norm": 1.2001152038574219,
+      "learning_rate": 2e-05,
+      "loss": 0.02799315,
+      "step": 9999
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 1.3295609951019287,
+      "learning_rate": 2e-05,
+      "loss": 0.04259039,
+      "step": 10000
+    },
+    {
+      "epoch": 20.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.99,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9740518962075848,
+        "Equal_1": 0.996,
+        "Equal_2": 0.9820359281437125,
+        "Equal_3": 0.9520958083832335,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 0.998003992015968,
+        "LineComparison_3": 0.9940119760479041,
+        "Parallel_1": 0.9879759519038076,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.99,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.986,
+        "Perpendicular_3": 0.7975951903807615,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 0.998,
+        "PointLiesOnCircle_3": 0.994,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9919839679358717,
+        "PointLiesOnLine_3": 0.9800399201596807
+      },
+      "eval_runtime": 319.7877,
+      "eval_samples_per_second": 32.834,
+      "eval_steps_per_second": 0.657,
+      "step": 10000
+    },
+    {
+      "epoch": 20.002,
+      "grad_norm": 1.157810926437378,
+      "learning_rate": 2e-05,
+      "loss": 0.02725174,
+      "step": 10001
+    },
+    {
+      "epoch": 20.004,
+      "grad_norm": 2.2255258560180664,
+      "learning_rate": 2e-05,
+      "loss": 0.04750681,
+      "step": 10002
+    },
+    {
+      "epoch": 20.006,
+      "grad_norm": 0.977530300617218,
+      "learning_rate": 2e-05,
+      "loss": 0.03366182,
+      "step": 10003
+    },
+    {
+      "epoch": 20.008,
+      "grad_norm": 2.368126392364502,
+      "learning_rate": 2e-05,
+      "loss": 0.03093569,
+      "step": 10004
+    },
+    {
+      "epoch": 20.01,
+      "grad_norm": 1.1914955377578735,
+      "learning_rate": 2e-05,
+      "loss": 0.04483454,
+      "step": 10005
+    },
+    {
+      "epoch": 20.012,
+      "grad_norm": 1.2124160528182983,
+      "learning_rate": 2e-05,
+      "loss": 0.03203357,
+      "step": 10006
+    },
+    {
+      "epoch": 20.014,
+      "grad_norm": 2.622993230819702,
+      "learning_rate": 2e-05,
+      "loss": 0.0496589,
+      "step": 10007
+    },
+    {
+      "epoch": 20.016,
+      "grad_norm": 1.2373409271240234,
+      "learning_rate": 2e-05,
+      "loss": 0.04058324,
+      "step": 10008
+    },
+    {
+      "epoch": 20.018,
+      "grad_norm": 1.6478288173675537,
+      "learning_rate": 2e-05,
+      "loss": 0.04132287,
+      "step": 10009
+    },
+    {
+      "epoch": 20.02,
+      "grad_norm": 1.0980284214019775,
+      "learning_rate": 2e-05,
+      "loss": 0.03450535,
+      "step": 10010
+    },
+    {
+      "epoch": 20.022,
+      "grad_norm": 0.9690496921539307,
+      "learning_rate": 2e-05,
+      "loss": 0.02480958,
+      "step": 10011
+    },
+    {
+      "epoch": 20.024,
+      "grad_norm": 3.7256979942321777,
+      "learning_rate": 2e-05,
+      "loss": 0.03763371,
+      "step": 10012
+    },
+    {
+      "epoch": 20.026,
+      "grad_norm": 0.9773795008659363,
+      "learning_rate": 2e-05,
+      "loss": 0.02543466,
+      "step": 10013
+    },
+    {
+      "epoch": 20.028,
+      "grad_norm": 1.7975311279296875,
+      "learning_rate": 2e-05,
+      "loss": 0.0339796,
+      "step": 10014
+    },
+    {
+      "epoch": 20.03,
+      "grad_norm": 1.9980261325836182,
+      "learning_rate": 2e-05,
+      "loss": 0.03168187,
+      "step": 10015
+    },
+    {
+      "epoch": 20.032,
+      "grad_norm": 1.4407600164413452,
+      "learning_rate": 2e-05,
+      "loss": 0.05185389,
+      "step": 10016
+    },
+    {
+      "epoch": 20.034,
+      "grad_norm": 1.0228748321533203,
+      "learning_rate": 2e-05,
+      "loss": 0.02575734,
+      "step": 10017
+    },
+    {
+      "epoch": 20.036,
+      "grad_norm": 1.1408131122589111,
+      "learning_rate": 2e-05,
+      "loss": 0.03414143,
+      "step": 10018
+    },
+    {
+      "epoch": 20.038,
+      "grad_norm": 1.2371625900268555,
+      "learning_rate": 2e-05,
+      "loss": 0.03162018,
+      "step": 10019
+    },
+    {
+      "epoch": 20.04,
+      "grad_norm": 0.982409656047821,
+      "learning_rate": 2e-05,
+      "loss": 0.0248183,
+      "step": 10020
+    },
+    {
+      "epoch": 20.042,
+      "grad_norm": 1.5073639154434204,
+      "learning_rate": 2e-05,
+      "loss": 0.04884845,
+      "step": 10021
+    },
+    {
+      "epoch": 20.044,
+      "grad_norm": 1.28328537940979,
+      "learning_rate": 2e-05,
+      "loss": 0.04167043,
+      "step": 10022
+    },
+    {
+      "epoch": 20.046,
+      "grad_norm": 1.0359996557235718,
+      "learning_rate": 2e-05,
+      "loss": 0.0376031,
+      "step": 10023
+    },
+    {
+      "epoch": 20.048,
+      "grad_norm": 1.150727391242981,
+      "learning_rate": 2e-05,
+      "loss": 0.02745938,
+      "step": 10024
+    },
+    {
+      "epoch": 20.05,
+      "grad_norm": 1.7619673013687134,
+      "learning_rate": 2e-05,
+      "loss": 0.04664058,
+      "step": 10025
+    },
+    {
+      "epoch": 20.052,
+      "grad_norm": 1.849218726158142,
+      "learning_rate": 2e-05,
+      "loss": 0.0393936,
+      "step": 10026
+    },
+    {
+      "epoch": 20.054,
+      "grad_norm": 1.425082802772522,
+      "learning_rate": 2e-05,
+      "loss": 0.03768519,
+      "step": 10027
+    },
+    {
+      "epoch": 20.056,
+      "grad_norm": 2.417065143585205,
+      "learning_rate": 2e-05,
+      "loss": 0.04438198,
+      "step": 10028
+    },
+    {
+      "epoch": 20.058,
+      "grad_norm": 1.103333592414856,
+      "learning_rate": 2e-05,
+      "loss": 0.0304841,
+      "step": 10029
+    },
+    {
+      "epoch": 20.06,
+      "grad_norm": 1.0776103734970093,
+      "learning_rate": 2e-05,
+      "loss": 0.03893217,
+      "step": 10030
+    },
+    {
+      "epoch": 20.062,
+      "grad_norm": 1.4815536737442017,
+      "learning_rate": 2e-05,
+      "loss": 0.05196786,
+      "step": 10031
+    },
+    {
+      "epoch": 20.064,
+      "grad_norm": 1.3739426136016846,
+      "learning_rate": 2e-05,
+      "loss": 0.03645986,
+      "step": 10032
+    },
+    {
+      "epoch": 20.066,
+      "grad_norm": 1.1692328453063965,
+      "learning_rate": 2e-05,
+      "loss": 0.03630241,
+      "step": 10033
+    },
+    {
+      "epoch": 20.068,
+      "grad_norm": 1.0185339450836182,
+      "learning_rate": 2e-05,
+      "loss": 0.0344325,
+      "step": 10034
+    },
+    {
+      "epoch": 20.07,
+      "grad_norm": 0.9241943359375,
+      "learning_rate": 2e-05,
+      "loss": 0.02990804,
+      "step": 10035
+    },
+    {
+      "epoch": 20.072,
+      "grad_norm": 2.4085822105407715,
+      "learning_rate": 2e-05,
+      "loss": 0.04084285,
+      "step": 10036
+    },
+    {
+      "epoch": 20.074,
+      "grad_norm": 1.6510875225067139,
+      "learning_rate": 2e-05,
+      "loss": 0.03414775,
+      "step": 10037
+    },
+    {
+      "epoch": 20.076,
+      "grad_norm": 1.0024648904800415,
+      "learning_rate": 2e-05,
+      "loss": 0.02915655,
+      "step": 10038
+    },
+    {
+      "epoch": 20.078,
+      "grad_norm": 1.1941813230514526,
+      "learning_rate": 2e-05,
+      "loss": 0.03444272,
+      "step": 10039
+    },
+    {
+      "epoch": 20.08,
+      "grad_norm": 1.5588897466659546,
+      "learning_rate": 2e-05,
+      "loss": 0.04292718,
+      "step": 10040
+    },
+    {
+      "epoch": 20.082,
+      "grad_norm": 1.1992404460906982,
+      "learning_rate": 2e-05,
+      "loss": 0.03661583,
+      "step": 10041
+    },
+    {
+      "epoch": 20.084,
+      "grad_norm": 1.6118459701538086,
+      "learning_rate": 2e-05,
+      "loss": 0.03280291,
+      "step": 10042
+    },
+    {
+      "epoch": 20.086,
+      "grad_norm": 1.184962272644043,
+      "learning_rate": 2e-05,
+      "loss": 0.03350543,
+      "step": 10043
+    },
+    {
+      "epoch": 20.088,
+      "grad_norm": 1.2698707580566406,
+      "learning_rate": 2e-05,
+      "loss": 0.03131055,
+      "step": 10044
+    },
+    {
+      "epoch": 20.09,
+      "grad_norm": 1.2608489990234375,
+      "learning_rate": 2e-05,
+      "loss": 0.04219927,
+      "step": 10045
+    },
+    {
+      "epoch": 20.092,
+      "grad_norm": 1.3150699138641357,
+      "learning_rate": 2e-05,
+      "loss": 0.04263498,
+      "step": 10046
+    },
+    {
+      "epoch": 20.094,
+      "grad_norm": 0.7644698619842529,
+      "learning_rate": 2e-05,
+      "loss": 0.01624379,
+      "step": 10047
+    },
+    {
+      "epoch": 20.096,
+      "grad_norm": 1.202438235282898,
+      "learning_rate": 2e-05,
+      "loss": 0.04117297,
+      "step": 10048
+    },
+    {
+      "epoch": 20.098,
+      "grad_norm": 1.1358537673950195,
+      "learning_rate": 2e-05,
+      "loss": 0.0320413,
+      "step": 10049
+    },
+    {
+      "epoch": 20.1,
+      "grad_norm": 1.1852686405181885,
+      "learning_rate": 2e-05,
+      "loss": 0.0407727,
+      "step": 10050
+    },
+    {
+      "epoch": 20.102,
+      "grad_norm": 1.051586389541626,
+      "learning_rate": 2e-05,
+      "loss": 0.03409174,
+      "step": 10051
+    },
+    {
+      "epoch": 20.104,
+      "grad_norm": 0.8840859532356262,
+      "learning_rate": 2e-05,
+      "loss": 0.02449731,
+      "step": 10052
+    },
+    {
+      "epoch": 20.106,
+      "grad_norm": 0.8056784868240356,
+      "learning_rate": 2e-05,
+      "loss": 0.02186309,
+      "step": 10053
+    },
+    {
+      "epoch": 20.108,
+      "grad_norm": 1.1816960573196411,
+      "learning_rate": 2e-05,
+      "loss": 0.03838727,
+      "step": 10054
+    },
+    {
+      "epoch": 20.11,
+      "grad_norm": 1.3361520767211914,
+      "learning_rate": 2e-05,
+      "loss": 0.046068,
+      "step": 10055
+    },
+    {
+      "epoch": 20.112,
+      "grad_norm": 1.7416718006134033,
+      "learning_rate": 2e-05,
+      "loss": 0.02897919,
+      "step": 10056
+    },
+    {
+      "epoch": 20.114,
+      "grad_norm": 0.8946969509124756,
+      "learning_rate": 2e-05,
+      "loss": 0.02574443,
+      "step": 10057
+    },
+    {
+      "epoch": 20.116,
+      "grad_norm": 0.9283491969108582,
+      "learning_rate": 2e-05,
+      "loss": 0.02700006,
+      "step": 10058
+    },
+    {
+      "epoch": 20.118,
+      "grad_norm": 1.316556692123413,
+      "learning_rate": 2e-05,
+      "loss": 0.03163679,
+      "step": 10059
+    },
+    {
+      "epoch": 20.12,
+      "grad_norm": 1.7087239027023315,
+      "learning_rate": 2e-05,
+      "loss": 0.05993263,
+      "step": 10060
+    },
+    {
+      "epoch": 20.122,
+      "grad_norm": 1.1801674365997314,
+      "learning_rate": 2e-05,
+      "loss": 0.03400156,
+      "step": 10061
+    },
+    {
+      "epoch": 20.124,
+      "grad_norm": 1.5751726627349854,
+      "learning_rate": 2e-05,
+      "loss": 0.03422957,
+      "step": 10062
+    },
+    {
+      "epoch": 20.126,
+      "grad_norm": 0.9739691019058228,
+      "learning_rate": 2e-05,
+      "loss": 0.02737168,
+      "step": 10063
+    },
+    {
+      "epoch": 20.128,
+      "grad_norm": 1.8991624116897583,
+      "learning_rate": 2e-05,
+      "loss": 0.05057021,
+      "step": 10064
+    },
+    {
+      "epoch": 20.13,
+      "grad_norm": 1.7493529319763184,
+      "learning_rate": 2e-05,
+      "loss": 0.03247079,
+      "step": 10065
+    },
+    {
+      "epoch": 20.132,
+      "grad_norm": 0.9996808767318726,
+      "learning_rate": 2e-05,
+      "loss": 0.03386039,
+      "step": 10066
+    },
+    {
+      "epoch": 20.134,
+      "grad_norm": 1.3784937858581543,
+      "learning_rate": 2e-05,
+      "loss": 0.04235096,
+      "step": 10067
+    },
+    {
+      "epoch": 20.136,
+      "grad_norm": 0.9882324934005737,
+      "learning_rate": 2e-05,
+      "loss": 0.03374637,
+      "step": 10068
+    },
+    {
+      "epoch": 20.138,
+      "grad_norm": 1.1563467979431152,
+      "learning_rate": 2e-05,
+      "loss": 0.03350766,
+      "step": 10069
+    },
+    {
+      "epoch": 20.14,
+      "grad_norm": 1.2875149250030518,
+      "learning_rate": 2e-05,
+      "loss": 0.03958068,
+      "step": 10070
+    },
+    {
+      "epoch": 20.142,
+      "grad_norm": 1.2082184553146362,
+      "learning_rate": 2e-05,
+      "loss": 0.03617564,
+      "step": 10071
+    },
+    {
+      "epoch": 20.144,
+      "grad_norm": 1.5155640840530396,
+      "learning_rate": 2e-05,
+      "loss": 0.03431886,
+      "step": 10072
+    },
+    {
+      "epoch": 20.146,
+      "grad_norm": 1.241598129272461,
+      "learning_rate": 2e-05,
+      "loss": 0.05081156,
+      "step": 10073
+    },
+    {
+      "epoch": 20.148,
+      "grad_norm": 1.3844859600067139,
+      "learning_rate": 2e-05,
+      "loss": 0.04880771,
+      "step": 10074
+    },
+    {
+      "epoch": 20.15,
+      "grad_norm": 1.1351920366287231,
+      "learning_rate": 2e-05,
+      "loss": 0.04456853,
+      "step": 10075
+    },
+    {
+      "epoch": 20.152,
+      "grad_norm": 1.7876904010772705,
+      "learning_rate": 2e-05,
+      "loss": 0.04733547,
+      "step": 10076
+    },
+    {
+      "epoch": 20.154,
+      "grad_norm": 1.676299810409546,
+      "learning_rate": 2e-05,
+      "loss": 0.04053468,
+      "step": 10077
+    },
+    {
+      "epoch": 20.156,
+      "grad_norm": 0.9991481304168701,
+      "learning_rate": 2e-05,
+      "loss": 0.03337016,
+      "step": 10078
+    },
+    {
+      "epoch": 20.158,
+      "grad_norm": 1.0327903032302856,
+      "learning_rate": 2e-05,
+      "loss": 0.03195174,
+      "step": 10079
+    },
+    {
+      "epoch": 20.16,
+      "grad_norm": 1.0681126117706299,
+      "learning_rate": 2e-05,
+      "loss": 0.03106238,
+      "step": 10080
+    },
+    {
+      "epoch": 20.162,
+      "grad_norm": 2.461606979370117,
+      "learning_rate": 2e-05,
+      "loss": 0.04076942,
+      "step": 10081
+    },
+    {
+      "epoch": 20.164,
+      "grad_norm": 1.4891592264175415,
+      "learning_rate": 2e-05,
+      "loss": 0.0378016,
+      "step": 10082
+    },
+    {
+      "epoch": 20.166,
+      "grad_norm": 2.7966361045837402,
+      "learning_rate": 2e-05,
+      "loss": 0.04116115,
+      "step": 10083
+    },
+    {
+      "epoch": 20.168,
+      "grad_norm": 1.0875968933105469,
+      "learning_rate": 2e-05,
+      "loss": 0.03374017,
+      "step": 10084
+    },
+    {
+      "epoch": 20.17,
+      "grad_norm": 1.6224273443222046,
+      "learning_rate": 2e-05,
+      "loss": 0.04455825,
+      "step": 10085
+    },
+    {
+      "epoch": 20.172,
+      "grad_norm": 1.2282185554504395,
+      "learning_rate": 2e-05,
+      "loss": 0.03824051,
+      "step": 10086
+    },
+    {
+      "epoch": 20.174,
+      "grad_norm": 1.5995073318481445,
+      "learning_rate": 2e-05,
+      "loss": 0.03549742,
+      "step": 10087
+    },
+    {
+      "epoch": 20.176,
+      "grad_norm": 1.2183595895767212,
+      "learning_rate": 2e-05,
+      "loss": 0.03944673,
+      "step": 10088
+    },
+    {
+      "epoch": 20.178,
+      "grad_norm": 1.4532480239868164,
+      "learning_rate": 2e-05,
+      "loss": 0.03099249,
+      "step": 10089
+    },
+    {
+      "epoch": 20.18,
+      "grad_norm": 1.6714351177215576,
+      "learning_rate": 2e-05,
+      "loss": 0.05370768,
+      "step": 10090
+    },
+    {
+      "epoch": 20.182,
+      "grad_norm": 1.2066564559936523,
+      "learning_rate": 2e-05,
+      "loss": 0.03310648,
+      "step": 10091
+    },
+    {
+      "epoch": 20.184,
+      "grad_norm": 1.9016071557998657,
+      "learning_rate": 2e-05,
+      "loss": 0.06958458,
+      "step": 10092
+    },
+    {
+      "epoch": 20.186,
+      "grad_norm": 1.0576119422912598,
+      "learning_rate": 2e-05,
+      "loss": 0.03138105,
+      "step": 10093
+    },
+    {
+      "epoch": 20.188,
+      "grad_norm": 1.632576823234558,
+      "learning_rate": 2e-05,
+      "loss": 0.05661856,
+      "step": 10094
+    },
+    {
+      "epoch": 20.19,
+      "grad_norm": 1.3058475255966187,
+      "learning_rate": 2e-05,
+      "loss": 0.02934664,
+      "step": 10095
+    },
+    {
+      "epoch": 20.192,
+      "grad_norm": 1.5141475200653076,
+      "learning_rate": 2e-05,
+      "loss": 0.04773264,
+      "step": 10096
+    },
+    {
+      "epoch": 20.194,
+      "grad_norm": 1.2892210483551025,
+      "learning_rate": 2e-05,
+      "loss": 0.0409392,
+      "step": 10097
+    },
+    {
+      "epoch": 20.196,
+      "grad_norm": 1.0145940780639648,
+      "learning_rate": 2e-05,
+      "loss": 0.03348306,
+      "step": 10098
+    },
+    {
+      "epoch": 20.198,
+      "grad_norm": 1.3150386810302734,
+      "learning_rate": 2e-05,
+      "loss": 0.04228361,
+      "step": 10099
+    },
+    {
+      "epoch": 20.2,
+      "grad_norm": 1.4747895002365112,
+      "learning_rate": 2e-05,
+      "loss": 0.04473066,
+      "step": 10100
+    },
+    {
+      "epoch": 20.202,
+      "grad_norm": 1.3429152965545654,
+      "learning_rate": 2e-05,
+      "loss": 0.03698963,
+      "step": 10101
+    },
+    {
+      "epoch": 20.204,
+      "grad_norm": 1.5666682720184326,
+      "learning_rate": 2e-05,
+      "loss": 0.03814546,
+      "step": 10102
+    },
+    {
+      "epoch": 20.206,
+      "grad_norm": 1.1224846839904785,
+      "learning_rate": 2e-05,
+      "loss": 0.03632937,
+      "step": 10103
+    },
+    {
+      "epoch": 20.208,
+      "grad_norm": 1.2551809549331665,
+      "learning_rate": 2e-05,
+      "loss": 0.05948218,
+      "step": 10104
+    },
+    {
+      "epoch": 20.21,
+      "grad_norm": 1.0920709371566772,
+      "learning_rate": 2e-05,
+      "loss": 0.03180977,
+      "step": 10105
+    },
+    {
+      "epoch": 20.212,
+      "grad_norm": 1.1358767747879028,
+      "learning_rate": 2e-05,
+      "loss": 0.05127998,
+      "step": 10106
+    },
+    {
+      "epoch": 20.214,
+      "grad_norm": 1.1630680561065674,
+      "learning_rate": 2e-05,
+      "loss": 0.03373503,
+      "step": 10107
+    },
+    {
+      "epoch": 20.216,
+      "grad_norm": 1.0587674379348755,
+      "learning_rate": 2e-05,
+      "loss": 0.03114433,
+      "step": 10108
+    },
+    {
+      "epoch": 20.218,
+      "grad_norm": 2.387716054916382,
+      "learning_rate": 2e-05,
+      "loss": 0.05491668,
+      "step": 10109
+    },
+    {
+      "epoch": 20.22,
+      "grad_norm": 1.344147801399231,
+      "learning_rate": 2e-05,
+      "loss": 0.03537369,
+      "step": 10110
+    },
+    {
+      "epoch": 20.222,
+      "grad_norm": 1.4900367259979248,
+      "learning_rate": 2e-05,
+      "loss": 0.04168044,
+      "step": 10111
+    },
+    {
+      "epoch": 20.224,
+      "grad_norm": 1.3198118209838867,
+      "learning_rate": 2e-05,
+      "loss": 0.03236744,
+      "step": 10112
+    },
+    {
+      "epoch": 20.226,
+      "grad_norm": 1.05833101272583,
+      "learning_rate": 2e-05,
+      "loss": 0.03506115,
+      "step": 10113
+    },
+    {
+      "epoch": 20.228,
+      "grad_norm": 3.1193032264709473,
+      "learning_rate": 2e-05,
+      "loss": 0.06916459,
+      "step": 10114
+    },
+    {
+      "epoch": 20.23,
+      "grad_norm": 0.9400885105133057,
+      "learning_rate": 2e-05,
+      "loss": 0.02644516,
+      "step": 10115
+    },
+    {
+      "epoch": 20.232,
+      "grad_norm": 1.744911551475525,
+      "learning_rate": 2e-05,
+      "loss": 0.03334912,
+      "step": 10116
+    },
+    {
+      "epoch": 20.234,
+      "grad_norm": 1.1233267784118652,
+      "learning_rate": 2e-05,
+      "loss": 0.02539495,
+      "step": 10117
+    },
+    {
+      "epoch": 20.236,
+      "grad_norm": 0.8864102959632874,
+      "learning_rate": 2e-05,
+      "loss": 0.02736823,
+      "step": 10118
+    },
+    {
+      "epoch": 20.238,
+      "grad_norm": 0.9077542424201965,
+      "learning_rate": 2e-05,
+      "loss": 0.02917507,
+      "step": 10119
+    },
+    {
+      "epoch": 20.24,
+      "grad_norm": 1.7056787014007568,
+      "learning_rate": 2e-05,
+      "loss": 0.04501066,
+      "step": 10120
+    },
+    {
+      "epoch": 20.242,
+      "grad_norm": 1.0653247833251953,
+      "learning_rate": 2e-05,
+      "loss": 0.03258765,
+      "step": 10121
+    },
+    {
+      "epoch": 20.244,
+      "grad_norm": 1.5916606187820435,
+      "learning_rate": 2e-05,
+      "loss": 0.03004246,
+      "step": 10122
+    },
+    {
+      "epoch": 20.246,
+      "grad_norm": 0.992498517036438,
+      "learning_rate": 2e-05,
+      "loss": 0.03313632,
+      "step": 10123
+    },
+    {
+      "epoch": 20.248,
+      "grad_norm": 1.408166527748108,
+      "learning_rate": 2e-05,
+      "loss": 0.03982632,
+      "step": 10124
+    },
+    {
+      "epoch": 20.25,
+      "grad_norm": 2.0703201293945312,
+      "learning_rate": 2e-05,
+      "loss": 0.03419951,
+      "step": 10125
+    },
+    {
+      "epoch": 20.252,
+      "grad_norm": 2.027787446975708,
+      "learning_rate": 2e-05,
+      "loss": 0.04078956,
+      "step": 10126
+    },
+    {
+      "epoch": 20.254,
+      "grad_norm": 1.7033048868179321,
+      "learning_rate": 2e-05,
+      "loss": 0.04830636,
+      "step": 10127
+    },
+    {
+      "epoch": 20.256,
+      "grad_norm": 1.3836185932159424,
+      "learning_rate": 2e-05,
+      "loss": 0.05766873,
+      "step": 10128
+    },
+    {
+      "epoch": 20.258,
+      "grad_norm": 0.874599039554596,
+      "learning_rate": 2e-05,
+      "loss": 0.02754955,
+      "step": 10129
+    },
+    {
+      "epoch": 20.26,
+      "grad_norm": 1.280648946762085,
+      "learning_rate": 2e-05,
+      "loss": 0.03249458,
+      "step": 10130
+    },
+    {
+      "epoch": 20.262,
+      "grad_norm": 3.8084380626678467,
+      "learning_rate": 2e-05,
+      "loss": 0.03238885,
+      "step": 10131
+    },
+    {
+      "epoch": 20.264,
+      "grad_norm": 1.1068285703659058,
+      "learning_rate": 2e-05,
+      "loss": 0.03581792,
+      "step": 10132
+    },
+    {
+      "epoch": 20.266,
+      "grad_norm": 1.4236106872558594,
+      "learning_rate": 2e-05,
+      "loss": 0.03960142,
+      "step": 10133
+    },
+    {
+      "epoch": 20.268,
+      "grad_norm": 1.674912452697754,
+      "learning_rate": 2e-05,
+      "loss": 0.04210643,
+      "step": 10134
+    },
+    {
+      "epoch": 20.27,
+      "grad_norm": 1.6190119981765747,
+      "learning_rate": 2e-05,
+      "loss": 0.03533464,
+      "step": 10135
+    },
+    {
+      "epoch": 20.272,
+      "grad_norm": 1.0870835781097412,
+      "learning_rate": 2e-05,
+      "loss": 0.03944698,
+      "step": 10136
+    },
+    {
+      "epoch": 20.274,
+      "grad_norm": 1.3776017427444458,
+      "learning_rate": 2e-05,
+      "loss": 0.03751071,
+      "step": 10137
+    },
+    {
+      "epoch": 20.276,
+      "grad_norm": 1.1693474054336548,
+      "learning_rate": 2e-05,
+      "loss": 0.03089845,
+      "step": 10138
+    },
+    {
+      "epoch": 20.278,
+      "grad_norm": 1.5422534942626953,
+      "learning_rate": 2e-05,
+      "loss": 0.0364728,
+      "step": 10139
+    },
+    {
+      "epoch": 20.28,
+      "grad_norm": 1.130757212638855,
+      "learning_rate": 2e-05,
+      "loss": 0.03968713,
+      "step": 10140
+    },
+    {
+      "epoch": 20.282,
+      "grad_norm": 1.601479172706604,
+      "learning_rate": 2e-05,
+      "loss": 0.0454113,
+      "step": 10141
+    },
+    {
+      "epoch": 20.284,
+      "grad_norm": 1.0961538553237915,
+      "learning_rate": 2e-05,
+      "loss": 0.03847424,
+      "step": 10142
+    },
+    {
+      "epoch": 20.286,
+      "grad_norm": 1.3003602027893066,
+      "learning_rate": 2e-05,
+      "loss": 0.03098981,
+      "step": 10143
+    },
+    {
+      "epoch": 20.288,
+      "grad_norm": 1.9535208940505981,
+      "learning_rate": 2e-05,
+      "loss": 0.03620074,
+      "step": 10144
+    },
+    {
+      "epoch": 20.29,
+      "grad_norm": 1.0742915868759155,
+      "learning_rate": 2e-05,
+      "loss": 0.03768191,
+      "step": 10145
+    },
+    {
+      "epoch": 20.292,
+      "grad_norm": 0.8740436434745789,
+      "learning_rate": 2e-05,
+      "loss": 0.0226828,
+      "step": 10146
+    },
+    {
+      "epoch": 20.294,
+      "grad_norm": 0.9478847980499268,
+      "learning_rate": 2e-05,
+      "loss": 0.03354997,
+      "step": 10147
+    },
+    {
+      "epoch": 20.296,
+      "grad_norm": 1.195736289024353,
+      "learning_rate": 2e-05,
+      "loss": 0.03266739,
+      "step": 10148
+    },
+    {
+      "epoch": 20.298,
+      "grad_norm": 1.1402510404586792,
+      "learning_rate": 2e-05,
+      "loss": 0.03164797,
+      "step": 10149
+    },
+    {
+      "epoch": 20.3,
+      "grad_norm": 0.9935315847396851,
+      "learning_rate": 2e-05,
+      "loss": 0.0263091,
+      "step": 10150
+    },
+    {
+      "epoch": 20.302,
+      "grad_norm": 1.2046669721603394,
+      "learning_rate": 2e-05,
+      "loss": 0.03377987,
+      "step": 10151
+    },
+    {
+      "epoch": 20.304,
+      "grad_norm": 1.2782913446426392,
+      "learning_rate": 2e-05,
+      "loss": 0.03555286,
+      "step": 10152
+    },
+    {
+      "epoch": 20.306,
+      "grad_norm": 1.155943512916565,
+      "learning_rate": 2e-05,
+      "loss": 0.03053265,
+      "step": 10153
+    },
+    {
+      "epoch": 20.308,
+      "grad_norm": 1.3335613012313843,
+      "learning_rate": 2e-05,
+      "loss": 0.02797567,
+      "step": 10154
+    },
+    {
+      "epoch": 20.31,
+      "grad_norm": 1.3252575397491455,
+      "learning_rate": 2e-05,
+      "loss": 0.04167391,
+      "step": 10155
+    },
+    {
+      "epoch": 20.312,
+      "grad_norm": 1.257401704788208,
+      "learning_rate": 2e-05,
+      "loss": 0.04301036,
+      "step": 10156
+    },
+    {
+      "epoch": 20.314,
+      "grad_norm": 1.2897112369537354,
+      "learning_rate": 2e-05,
+      "loss": 0.02559963,
+      "step": 10157
+    },
+    {
+      "epoch": 20.316,
+      "grad_norm": 1.0978411436080933,
+      "learning_rate": 2e-05,
+      "loss": 0.03113418,
+      "step": 10158
+    },
+    {
+      "epoch": 20.318,
+      "grad_norm": 0.9940638542175293,
+      "learning_rate": 2e-05,
+      "loss": 0.02741957,
+      "step": 10159
+    },
+    {
+      "epoch": 20.32,
+      "grad_norm": 1.1258987188339233,
+      "learning_rate": 2e-05,
+      "loss": 0.03458284,
+      "step": 10160
+    },
+    {
+      "epoch": 20.322,
+      "grad_norm": 1.3880150318145752,
+      "learning_rate": 2e-05,
+      "loss": 0.02557427,
+      "step": 10161
+    },
+    {
+      "epoch": 20.324,
+      "grad_norm": 0.89255690574646,
+      "learning_rate": 2e-05,
+      "loss": 0.02698753,
+      "step": 10162
+    },
+    {
+      "epoch": 20.326,
+      "grad_norm": 0.9891429543495178,
+      "learning_rate": 2e-05,
+      "loss": 0.03858247,
+      "step": 10163
+    },
+    {
+      "epoch": 20.328,
+      "grad_norm": 3.275198459625244,
+      "learning_rate": 2e-05,
+      "loss": 0.04408592,
+      "step": 10164
+    },
+    {
+      "epoch": 20.33,
+      "grad_norm": 2.566772699356079,
+      "learning_rate": 2e-05,
+      "loss": 0.04836412,
+      "step": 10165
+    },
+    {
+      "epoch": 20.332,
+      "grad_norm": 1.2465749979019165,
+      "learning_rate": 2e-05,
+      "loss": 0.03609654,
+      "step": 10166
+    },
+    {
+      "epoch": 20.334,
+      "grad_norm": 0.8518936634063721,
+      "learning_rate": 2e-05,
+      "loss": 0.02846691,
+      "step": 10167
+    },
+    {
+      "epoch": 20.336,
+      "grad_norm": 0.8058186769485474,
+      "learning_rate": 2e-05,
+      "loss": 0.02809364,
+      "step": 10168
+    },
+    {
+      "epoch": 20.338,
+      "grad_norm": 1.2480473518371582,
+      "learning_rate": 2e-05,
+      "loss": 0.03842864,
+      "step": 10169
+    },
+    {
+      "epoch": 20.34,
+      "grad_norm": 1.6874932050704956,
+      "learning_rate": 2e-05,
+      "loss": 0.02990905,
+      "step": 10170
+    },
+    {
+      "epoch": 20.342,
+      "grad_norm": 1.2645959854125977,
+      "learning_rate": 2e-05,
+      "loss": 0.02560621,
+      "step": 10171
+    },
+    {
+      "epoch": 20.344,
+      "grad_norm": 1.209860920906067,
+      "learning_rate": 2e-05,
+      "loss": 0.04289609,
+      "step": 10172
+    },
+    {
+      "epoch": 20.346,
+      "grad_norm": 0.9277303814888,
+      "learning_rate": 2e-05,
+      "loss": 0.02414119,
+      "step": 10173
+    },
+    {
+      "epoch": 20.348,
+      "grad_norm": 1.4776111841201782,
+      "learning_rate": 2e-05,
+      "loss": 0.03888512,
+      "step": 10174
+    },
+    {
+      "epoch": 20.35,
+      "grad_norm": 10.23653507232666,
+      "learning_rate": 2e-05,
+      "loss": 0.04151426,
+      "step": 10175
+    },
+    {
+      "epoch": 20.352,
+      "grad_norm": 1.4618253707885742,
+      "learning_rate": 2e-05,
+      "loss": 0.04169954,
+      "step": 10176
+    },
+    {
+      "epoch": 20.354,
+      "grad_norm": 1.517190933227539,
+      "learning_rate": 2e-05,
+      "loss": 0.03129986,
+      "step": 10177
+    },
+    {
+      "epoch": 20.356,
+      "grad_norm": 1.265041470527649,
+      "learning_rate": 2e-05,
+      "loss": 0.04625013,
+      "step": 10178
+    },
+    {
+      "epoch": 20.358,
+      "grad_norm": 0.9929906725883484,
+      "learning_rate": 2e-05,
+      "loss": 0.03318879,
+      "step": 10179
+    },
+    {
+      "epoch": 20.36,
+      "grad_norm": 1.1843352317810059,
+      "learning_rate": 2e-05,
+      "loss": 0.02937111,
+      "step": 10180
+    },
+    {
+      "epoch": 20.362,
+      "grad_norm": 3.79577374458313,
+      "learning_rate": 2e-05,
+      "loss": 0.04123313,
+      "step": 10181
+    },
+    {
+      "epoch": 20.364,
+      "grad_norm": 1.2206071615219116,
+      "learning_rate": 2e-05,
+      "loss": 0.03874102,
+      "step": 10182
+    },
+    {
+      "epoch": 20.366,
+      "grad_norm": 1.6392779350280762,
+      "learning_rate": 2e-05,
+      "loss": 0.03447268,
+      "step": 10183
+    },
+    {
+      "epoch": 20.368,
+      "grad_norm": 1.166060447692871,
+      "learning_rate": 2e-05,
+      "loss": 0.04190687,
+      "step": 10184
+    },
+    {
+      "epoch": 20.37,
+      "grad_norm": 1.3748960494995117,
+      "learning_rate": 2e-05,
+      "loss": 0.05042437,
+      "step": 10185
+    },
+    {
+      "epoch": 20.372,
+      "grad_norm": 1.016833782196045,
+      "learning_rate": 2e-05,
+      "loss": 0.0323528,
+      "step": 10186
+    },
+    {
+      "epoch": 20.374,
+      "grad_norm": 1.248270869255066,
+      "learning_rate": 2e-05,
+      "loss": 0.03355975,
+      "step": 10187
+    },
+    {
+      "epoch": 20.376,
+      "grad_norm": 2.965949058532715,
+      "learning_rate": 2e-05,
+      "loss": 0.03507435,
+      "step": 10188
+    },
+    {
+      "epoch": 20.378,
+      "grad_norm": 0.9657095074653625,
+      "learning_rate": 2e-05,
+      "loss": 0.03347619,
+      "step": 10189
+    },
+    {
+      "epoch": 20.38,
+      "grad_norm": 1.3561325073242188,
+      "learning_rate": 2e-05,
+      "loss": 0.03161833,
+      "step": 10190
+    },
+    {
+      "epoch": 20.382,
+      "grad_norm": 1.271353006362915,
+      "learning_rate": 2e-05,
+      "loss": 0.04460255,
+      "step": 10191
+    },
+    {
+      "epoch": 20.384,
+      "grad_norm": 1.1262338161468506,
+      "learning_rate": 2e-05,
+      "loss": 0.03606068,
+      "step": 10192
+    },
+    {
+      "epoch": 20.386,
+      "grad_norm": 1.0850212574005127,
+      "learning_rate": 2e-05,
+      "loss": 0.03550104,
+      "step": 10193
+    },
+    {
+      "epoch": 20.388,
+      "grad_norm": 1.1879736185073853,
+      "learning_rate": 2e-05,
+      "loss": 0.03706096,
+      "step": 10194
+    },
+    {
+      "epoch": 20.39,
+      "grad_norm": 0.8170276284217834,
+      "learning_rate": 2e-05,
+      "loss": 0.02326233,
+      "step": 10195
+    },
+    {
+      "epoch": 20.392,
+      "grad_norm": 1.2504932880401611,
+      "learning_rate": 2e-05,
+      "loss": 0.0375908,
+      "step": 10196
+    },
+    {
+      "epoch": 20.394,
+      "grad_norm": 2.0169646739959717,
+      "learning_rate": 2e-05,
+      "loss": 0.03000656,
+      "step": 10197
+    },
+    {
+      "epoch": 20.396,
+      "grad_norm": 1.223647952079773,
+      "learning_rate": 2e-05,
+      "loss": 0.03923958,
+      "step": 10198
+    },
+    {
+      "epoch": 20.398,
+      "grad_norm": 1.207067847251892,
+      "learning_rate": 2e-05,
+      "loss": 0.03308685,
+      "step": 10199
+    },
+    {
+      "epoch": 20.4,
+      "grad_norm": 1.9212299585342407,
+      "learning_rate": 2e-05,
+      "loss": 0.03529272,
+      "step": 10200
+    },
+    {
+      "epoch": 20.402,
+      "grad_norm": 2.4769113063812256,
+      "learning_rate": 2e-05,
+      "loss": 0.03606252,
+      "step": 10201
+    },
+    {
+      "epoch": 20.404,
+      "grad_norm": 1.2097878456115723,
+      "learning_rate": 2e-05,
+      "loss": 0.03837938,
+      "step": 10202
+    },
+    {
+      "epoch": 20.406,
+      "grad_norm": 1.7725907564163208,
+      "learning_rate": 2e-05,
+      "loss": 0.03187569,
+      "step": 10203
+    },
+    {
+      "epoch": 20.408,
+      "grad_norm": 1.7797393798828125,
+      "learning_rate": 2e-05,
+      "loss": 0.04215318,
+      "step": 10204
+    },
+    {
+      "epoch": 20.41,
+      "grad_norm": 1.4638700485229492,
+      "learning_rate": 2e-05,
+      "loss": 0.04161576,
+      "step": 10205
+    },
+    {
+      "epoch": 20.412,
+      "grad_norm": 1.297315001487732,
+      "learning_rate": 2e-05,
+      "loss": 0.03515932,
+      "step": 10206
+    },
+    {
+      "epoch": 20.414,
+      "grad_norm": 1.8025538921356201,
+      "learning_rate": 2e-05,
+      "loss": 0.03362041,
+      "step": 10207
+    },
+    {
+      "epoch": 20.416,
+      "grad_norm": 1.6176317930221558,
+      "learning_rate": 2e-05,
+      "loss": 0.040805,
+      "step": 10208
+    },
+    {
+      "epoch": 20.418,
+      "grad_norm": 0.9706657528877258,
+      "learning_rate": 2e-05,
+      "loss": 0.02987682,
+      "step": 10209
+    },
+    {
+      "epoch": 20.42,
+      "grad_norm": 1.238422155380249,
+      "learning_rate": 2e-05,
+      "loss": 0.03280898,
+      "step": 10210
+    },
+    {
+      "epoch": 20.422,
+      "grad_norm": 1.4637218713760376,
+      "learning_rate": 2e-05,
+      "loss": 0.03068679,
+      "step": 10211
+    },
+    {
+      "epoch": 20.424,
+      "grad_norm": 2.2752456665039062,
+      "learning_rate": 2e-05,
+      "loss": 0.04319573,
+      "step": 10212
+    },
+    {
+      "epoch": 20.426,
+      "grad_norm": 0.9194338917732239,
+      "learning_rate": 2e-05,
+      "loss": 0.03249656,
+      "step": 10213
+    },
+    {
+      "epoch": 20.428,
+      "grad_norm": 1.2353630065917969,
+      "learning_rate": 2e-05,
+      "loss": 0.02999087,
+      "step": 10214
+    },
+    {
+      "epoch": 20.43,
+      "grad_norm": 1.2627156972885132,
+      "learning_rate": 2e-05,
+      "loss": 0.04242412,
+      "step": 10215
+    },
+    {
+      "epoch": 20.432,
+      "grad_norm": 1.1771798133850098,
+      "learning_rate": 2e-05,
+      "loss": 0.04588514,
+      "step": 10216
+    },
+    {
+      "epoch": 20.434,
+      "grad_norm": 1.0947905778884888,
+      "learning_rate": 2e-05,
+      "loss": 0.03065883,
+      "step": 10217
+    },
+    {
+      "epoch": 20.436,
+      "grad_norm": 1.1154054403305054,
+      "learning_rate": 2e-05,
+      "loss": 0.02652307,
+      "step": 10218
+    },
+    {
+      "epoch": 20.438,
+      "grad_norm": 1.1405092477798462,
+      "learning_rate": 2e-05,
+      "loss": 0.01361668,
+      "step": 10219
+    },
+    {
+      "epoch": 20.44,
+      "grad_norm": 1.1158807277679443,
+      "learning_rate": 2e-05,
+      "loss": 0.02150419,
+      "step": 10220
+    },
+    {
+      "epoch": 20.442,
+      "grad_norm": 1.2909730672836304,
+      "learning_rate": 2e-05,
+      "loss": 0.03332622,
+      "step": 10221
+    },
+    {
+      "epoch": 20.444,
+      "grad_norm": 1.8945097923278809,
+      "learning_rate": 2e-05,
+      "loss": 0.05206165,
+      "step": 10222
+    },
+    {
+      "epoch": 20.446,
+      "grad_norm": 1.1864949464797974,
+      "learning_rate": 2e-05,
+      "loss": 0.02879405,
+      "step": 10223
+    },
+    {
+      "epoch": 20.448,
+      "grad_norm": 1.5208016633987427,
+      "learning_rate": 2e-05,
+      "loss": 0.04240531,
+      "step": 10224
+    },
+    {
+      "epoch": 20.45,
+      "grad_norm": 1.8025298118591309,
+      "learning_rate": 2e-05,
+      "loss": 0.04271449,
+      "step": 10225
+    },
+    {
+      "epoch": 20.452,
+      "grad_norm": 1.7837144136428833,
+      "learning_rate": 2e-05,
+      "loss": 0.0441253,
+      "step": 10226
+    },
+    {
+      "epoch": 20.454,
+      "grad_norm": 1.146601915359497,
+      "learning_rate": 2e-05,
+      "loss": 0.03485624,
+      "step": 10227
+    },
+    {
+      "epoch": 20.456,
+      "grad_norm": 0.8954287171363831,
+      "learning_rate": 2e-05,
+      "loss": 0.02285667,
+      "step": 10228
+    },
+    {
+      "epoch": 20.458,
+      "grad_norm": 2.9269511699676514,
+      "learning_rate": 2e-05,
+      "loss": 0.03113767,
+      "step": 10229
+    },
+    {
+      "epoch": 20.46,
+      "grad_norm": 0.9090030193328857,
+      "learning_rate": 2e-05,
+      "loss": 0.02447719,
+      "step": 10230
+    },
+    {
+      "epoch": 20.462,
+      "grad_norm": 1.004408836364746,
+      "learning_rate": 2e-05,
+      "loss": 0.03344005,
+      "step": 10231
+    },
+    {
+      "epoch": 20.464,
+      "grad_norm": 1.294810175895691,
+      "learning_rate": 2e-05,
+      "loss": 0.02832571,
+      "step": 10232
+    },
+    {
+      "epoch": 20.466,
+      "grad_norm": 1.3520246744155884,
+      "learning_rate": 2e-05,
+      "loss": 0.05982265,
+      "step": 10233
+    },
+    {
+      "epoch": 20.468,
+      "grad_norm": 1.4492565393447876,
+      "learning_rate": 2e-05,
+      "loss": 0.04218536,
+      "step": 10234
+    },
+    {
+      "epoch": 20.47,
+      "grad_norm": 2.6897168159484863,
+      "learning_rate": 2e-05,
+      "loss": 0.04501551,
+      "step": 10235
+    },
+    {
+      "epoch": 20.472,
+      "grad_norm": 1.8297394514083862,
+      "learning_rate": 2e-05,
+      "loss": 0.03225635,
+      "step": 10236
+    },
+    {
+      "epoch": 20.474,
+      "grad_norm": 1.2002131938934326,
+      "learning_rate": 2e-05,
+      "loss": 0.03257868,
+      "step": 10237
+    },
+    {
+      "epoch": 20.476,
+      "grad_norm": 1.413238525390625,
+      "learning_rate": 2e-05,
+      "loss": 0.04716073,
+      "step": 10238
+    },
+    {
+      "epoch": 20.478,
+      "grad_norm": 1.317274808883667,
+      "learning_rate": 2e-05,
+      "loss": 0.04736836,
+      "step": 10239
+    },
+    {
+      "epoch": 20.48,
+      "grad_norm": 1.25471031665802,
+      "learning_rate": 2e-05,
+      "loss": 0.03034903,
+      "step": 10240
+    },
+    {
+      "epoch": 20.482,
+      "grad_norm": 0.9913167357444763,
+      "learning_rate": 2e-05,
+      "loss": 0.02995886,
+      "step": 10241
+    },
+    {
+      "epoch": 20.484,
+      "grad_norm": 1.0410956144332886,
+      "learning_rate": 2e-05,
+      "loss": 0.03229685,
+      "step": 10242
+    },
+    {
+      "epoch": 20.486,
+      "grad_norm": 1.0938940048217773,
+      "learning_rate": 2e-05,
+      "loss": 0.02566743,
+      "step": 10243
+    },
+    {
+      "epoch": 20.488,
+      "grad_norm": 1.5702996253967285,
+      "learning_rate": 2e-05,
+      "loss": 0.03801944,
+      "step": 10244
+    },
+    {
+      "epoch": 20.49,
+      "grad_norm": 1.0842231512069702,
+      "learning_rate": 2e-05,
+      "loss": 0.03229133,
+      "step": 10245
+    },
+    {
+      "epoch": 20.492,
+      "grad_norm": 4.291932106018066,
+      "learning_rate": 2e-05,
+      "loss": 0.03782872,
+      "step": 10246
+    },
+    {
+      "epoch": 20.494,
+      "grad_norm": 1.3879879713058472,
+      "learning_rate": 2e-05,
+      "loss": 0.05402232,
+      "step": 10247
+    },
+    {
+      "epoch": 20.496,
+      "grad_norm": 1.004504919052124,
+      "learning_rate": 2e-05,
+      "loss": 0.03432404,
+      "step": 10248
+    },
+    {
+      "epoch": 20.498,
+      "grad_norm": 0.948047399520874,
+      "learning_rate": 2e-05,
+      "loss": 0.02306466,
+      "step": 10249
+    },
+    {
+      "epoch": 20.5,
+      "grad_norm": 1.270635724067688,
+      "learning_rate": 2e-05,
+      "loss": 0.03679778,
+      "step": 10250
+    },
+    {
+      "epoch": 20.502,
+      "grad_norm": 1.9827309846878052,
+      "learning_rate": 2e-05,
+      "loss": 0.04181923,
+      "step": 10251
+    },
+    {
+      "epoch": 20.504,
+      "grad_norm": 1.788630485534668,
+      "learning_rate": 2e-05,
+      "loss": 0.03067524,
+      "step": 10252
+    },
+    {
+      "epoch": 20.506,
+      "grad_norm": 1.5751676559448242,
+      "learning_rate": 2e-05,
+      "loss": 0.04409302,
+      "step": 10253
+    },
+    {
+      "epoch": 20.508,
+      "grad_norm": 1.0470032691955566,
+      "learning_rate": 2e-05,
+      "loss": 0.02919145,
+      "step": 10254
+    },
+    {
+      "epoch": 20.51,
+      "grad_norm": 1.5532153844833374,
+      "learning_rate": 2e-05,
+      "loss": 0.0342193,
+      "step": 10255
+    },
+    {
+      "epoch": 20.512,
+      "grad_norm": 1.2492958307266235,
+      "learning_rate": 2e-05,
+      "loss": 0.03348667,
+      "step": 10256
+    },
+    {
+      "epoch": 20.514,
+      "grad_norm": 1.015598177909851,
+      "learning_rate": 2e-05,
+      "loss": 0.02103408,
+      "step": 10257
+    },
+    {
+      "epoch": 20.516,
+      "grad_norm": 1.0364876985549927,
+      "learning_rate": 2e-05,
+      "loss": 0.03669195,
+      "step": 10258
+    },
+    {
+      "epoch": 20.518,
+      "grad_norm": 1.7574011087417603,
+      "learning_rate": 2e-05,
+      "loss": 0.02889457,
+      "step": 10259
+    },
+    {
+      "epoch": 20.52,
+      "grad_norm": 2.7406435012817383,
+      "learning_rate": 2e-05,
+      "loss": 0.03528633,
+      "step": 10260
+    },
+    {
+      "epoch": 20.522,
+      "grad_norm": 1.9497075080871582,
+      "learning_rate": 2e-05,
+      "loss": 0.03676779,
+      "step": 10261
+    },
+    {
+      "epoch": 20.524,
+      "grad_norm": 1.2621787786483765,
+      "learning_rate": 2e-05,
+      "loss": 0.03754564,
+      "step": 10262
+    },
+    {
+      "epoch": 20.526,
+      "grad_norm": 1.5117218494415283,
+      "learning_rate": 2e-05,
+      "loss": 0.03292262,
+      "step": 10263
+    },
+    {
+      "epoch": 20.528,
+      "grad_norm": 0.9753475785255432,
+      "learning_rate": 2e-05,
+      "loss": 0.02898685,
+      "step": 10264
+    },
+    {
+      "epoch": 20.53,
+      "grad_norm": 1.2703803777694702,
+      "learning_rate": 2e-05,
+      "loss": 0.03447834,
+      "step": 10265
+    },
+    {
+      "epoch": 20.532,
+      "grad_norm": 1.037750244140625,
+      "learning_rate": 2e-05,
+      "loss": 0.02343212,
+      "step": 10266
+    },
+    {
+      "epoch": 20.534,
+      "grad_norm": 1.8119739294052124,
+      "learning_rate": 2e-05,
+      "loss": 0.0345272,
+      "step": 10267
+    },
+    {
+      "epoch": 20.536,
+      "grad_norm": 0.9879088401794434,
+      "learning_rate": 2e-05,
+      "loss": 0.03324077,
+      "step": 10268
+    },
+    {
+      "epoch": 20.538,
+      "grad_norm": 1.1989351511001587,
+      "learning_rate": 2e-05,
+      "loss": 0.04151801,
+      "step": 10269
+    },
+    {
+      "epoch": 20.54,
+      "grad_norm": 1.0397365093231201,
+      "learning_rate": 2e-05,
+      "loss": 0.03603576,
+      "step": 10270
+    },
+    {
+      "epoch": 20.542,
+      "grad_norm": 3.7377114295959473,
+      "learning_rate": 2e-05,
+      "loss": 0.03584925,
+      "step": 10271
+    },
+    {
+      "epoch": 20.544,
+      "grad_norm": 0.9478249549865723,
+      "learning_rate": 2e-05,
+      "loss": 0.02031682,
+      "step": 10272
+    },
+    {
+      "epoch": 20.546,
+      "grad_norm": 1.2334407567977905,
+      "learning_rate": 2e-05,
+      "loss": 0.04748309,
+      "step": 10273
+    },
+    {
+      "epoch": 20.548000000000002,
+      "grad_norm": 2.86969256401062,
+      "learning_rate": 2e-05,
+      "loss": 0.04576512,
+      "step": 10274
+    },
+    {
+      "epoch": 20.55,
+      "grad_norm": 1.3003041744232178,
+      "learning_rate": 2e-05,
+      "loss": 0.03236845,
+      "step": 10275
+    },
+    {
+      "epoch": 20.552,
+      "grad_norm": 2.296776533126831,
+      "learning_rate": 2e-05,
+      "loss": 0.04013178,
+      "step": 10276
+    },
+    {
+      "epoch": 20.554,
+      "grad_norm": 1.155997395515442,
+      "learning_rate": 2e-05,
+      "loss": 0.03783529,
+      "step": 10277
+    },
+    {
+      "epoch": 20.556,
+      "grad_norm": 1.140368938446045,
+      "learning_rate": 2e-05,
+      "loss": 0.03774282,
+      "step": 10278
+    },
+    {
+      "epoch": 20.558,
+      "grad_norm": 1.4872280359268188,
+      "learning_rate": 2e-05,
+      "loss": 0.03347996,
+      "step": 10279
+    },
+    {
+      "epoch": 20.56,
+      "grad_norm": 2.4217143058776855,
+      "learning_rate": 2e-05,
+      "loss": 0.0531043,
+      "step": 10280
+    },
+    {
+      "epoch": 20.562,
+      "grad_norm": 1.4787334203720093,
+      "learning_rate": 2e-05,
+      "loss": 0.03938647,
+      "step": 10281
+    },
+    {
+      "epoch": 20.564,
+      "grad_norm": 1.2541735172271729,
+      "learning_rate": 2e-05,
+      "loss": 0.03302208,
+      "step": 10282
+    },
+    {
+      "epoch": 20.566,
+      "grad_norm": 1.1605701446533203,
+      "learning_rate": 2e-05,
+      "loss": 0.03414555,
+      "step": 10283
+    },
+    {
+      "epoch": 20.568,
+      "grad_norm": 2.4995267391204834,
+      "learning_rate": 2e-05,
+      "loss": 0.03342891,
+      "step": 10284
+    },
+    {
+      "epoch": 20.57,
+      "grad_norm": 1.6872836351394653,
+      "learning_rate": 2e-05,
+      "loss": 0.0238953,
+      "step": 10285
+    },
+    {
+      "epoch": 20.572,
+      "grad_norm": 1.2707072496414185,
+      "learning_rate": 2e-05,
+      "loss": 0.03610934,
+      "step": 10286
+    },
+    {
+      "epoch": 20.574,
+      "grad_norm": 1.908408761024475,
+      "learning_rate": 2e-05,
+      "loss": 0.02757995,
+      "step": 10287
+    },
+    {
+      "epoch": 20.576,
+      "grad_norm": 0.9678316712379456,
+      "learning_rate": 2e-05,
+      "loss": 0.02639127,
+      "step": 10288
+    },
+    {
+      "epoch": 20.578,
+      "grad_norm": 1.0835262537002563,
+      "learning_rate": 2e-05,
+      "loss": 0.03382458,
+      "step": 10289
+    },
+    {
+      "epoch": 20.58,
+      "grad_norm": 1.21681547164917,
+      "learning_rate": 2e-05,
+      "loss": 0.02967013,
+      "step": 10290
+    },
+    {
+      "epoch": 20.582,
+      "grad_norm": 1.723157286643982,
+      "learning_rate": 2e-05,
+      "loss": 0.03698352,
+      "step": 10291
+    },
+    {
+      "epoch": 20.584,
+      "grad_norm": 1.0680166482925415,
+      "learning_rate": 2e-05,
+      "loss": 0.02927051,
+      "step": 10292
+    },
+    {
+      "epoch": 20.586,
+      "grad_norm": 0.9486793279647827,
+      "learning_rate": 2e-05,
+      "loss": 0.0270779,
+      "step": 10293
+    },
+    {
+      "epoch": 20.588,
+      "grad_norm": 2.921025037765503,
+      "learning_rate": 2e-05,
+      "loss": 0.04973118,
+      "step": 10294
+    },
+    {
+      "epoch": 20.59,
+      "grad_norm": 1.2367773056030273,
+      "learning_rate": 2e-05,
+      "loss": 0.04252058,
+      "step": 10295
+    },
+    {
+      "epoch": 20.592,
+      "grad_norm": 1.1579737663269043,
+      "learning_rate": 2e-05,
+      "loss": 0.03316213,
+      "step": 10296
+    },
+    {
+      "epoch": 20.594,
+      "grad_norm": 2.2136647701263428,
+      "learning_rate": 2e-05,
+      "loss": 0.04563174,
+      "step": 10297
+    },
+    {
+      "epoch": 20.596,
+      "grad_norm": 1.8893747329711914,
+      "learning_rate": 2e-05,
+      "loss": 0.03102884,
+      "step": 10298
+    },
+    {
+      "epoch": 20.598,
+      "grad_norm": 2.0692083835601807,
+      "learning_rate": 2e-05,
+      "loss": 0.05330367,
+      "step": 10299
+    },
+    {
+      "epoch": 20.6,
+      "grad_norm": 0.9594127535820007,
+      "learning_rate": 2e-05,
+      "loss": 0.02779751,
+      "step": 10300
+    },
+    {
+      "epoch": 20.602,
+      "grad_norm": 2.0204217433929443,
+      "learning_rate": 2e-05,
+      "loss": 0.05240842,
+      "step": 10301
+    },
+    {
+      "epoch": 20.604,
+      "grad_norm": 1.9438049793243408,
+      "learning_rate": 2e-05,
+      "loss": 0.0536239,
+      "step": 10302
+    },
+    {
+      "epoch": 20.606,
+      "grad_norm": 1.0696160793304443,
+      "learning_rate": 2e-05,
+      "loss": 0.02945369,
+      "step": 10303
+    },
+    {
+      "epoch": 20.608,
+      "grad_norm": 1.0507266521453857,
+      "learning_rate": 2e-05,
+      "loss": 0.0354307,
+      "step": 10304
+    },
+    {
+      "epoch": 20.61,
+      "grad_norm": 1.37172269821167,
+      "learning_rate": 2e-05,
+      "loss": 0.04699483,
+      "step": 10305
+    },
+    {
+      "epoch": 20.612,
+      "grad_norm": 1.324902057647705,
+      "learning_rate": 2e-05,
+      "loss": 0.02831972,
+      "step": 10306
+    },
+    {
+      "epoch": 20.614,
+      "grad_norm": 1.9394724369049072,
+      "learning_rate": 2e-05,
+      "loss": 0.03686791,
+      "step": 10307
+    },
+    {
+      "epoch": 20.616,
+      "grad_norm": 1.221810221672058,
+      "learning_rate": 2e-05,
+      "loss": 0.04077676,
+      "step": 10308
+    },
+    {
+      "epoch": 20.618,
+      "grad_norm": 1.411195993423462,
+      "learning_rate": 2e-05,
+      "loss": 0.02411193,
+      "step": 10309
+    },
+    {
+      "epoch": 20.62,
+      "grad_norm": 1.5662122964859009,
+      "learning_rate": 2e-05,
+      "loss": 0.04722676,
+      "step": 10310
+    },
+    {
+      "epoch": 20.622,
+      "grad_norm": 0.8571231365203857,
+      "learning_rate": 2e-05,
+      "loss": 0.03123869,
+      "step": 10311
+    },
+    {
+      "epoch": 20.624,
+      "grad_norm": 0.9552671313285828,
+      "learning_rate": 2e-05,
+      "loss": 0.02956947,
+      "step": 10312
+    },
+    {
+      "epoch": 20.626,
+      "grad_norm": 1.1970704793930054,
+      "learning_rate": 2e-05,
+      "loss": 0.03908085,
+      "step": 10313
+    },
+    {
+      "epoch": 20.628,
+      "grad_norm": 1.5940030813217163,
+      "learning_rate": 2e-05,
+      "loss": 0.03680157,
+      "step": 10314
+    },
+    {
+      "epoch": 20.63,
+      "grad_norm": 1.1858786344528198,
+      "learning_rate": 2e-05,
+      "loss": 0.0258833,
+      "step": 10315
+    },
+    {
+      "epoch": 20.632,
+      "grad_norm": 1.1859554052352905,
+      "learning_rate": 2e-05,
+      "loss": 0.0370981,
+      "step": 10316
+    },
+    {
+      "epoch": 20.634,
+      "grad_norm": 0.9557812809944153,
+      "learning_rate": 2e-05,
+      "loss": 0.03161231,
+      "step": 10317
+    },
+    {
+      "epoch": 20.636,
+      "grad_norm": 1.269998550415039,
+      "learning_rate": 2e-05,
+      "loss": 0.03466748,
+      "step": 10318
+    },
+    {
+      "epoch": 20.638,
+      "grad_norm": 1.5743147134780884,
+      "learning_rate": 2e-05,
+      "loss": 0.03005154,
+      "step": 10319
+    },
+    {
+      "epoch": 20.64,
+      "grad_norm": 1.3234727382659912,
+      "learning_rate": 2e-05,
+      "loss": 0.03330294,
+      "step": 10320
+    },
+    {
+      "epoch": 20.642,
+      "grad_norm": 1.6464325189590454,
+      "learning_rate": 2e-05,
+      "loss": 0.04935025,
+      "step": 10321
+    },
+    {
+      "epoch": 20.644,
+      "grad_norm": 1.5875946283340454,
+      "learning_rate": 2e-05,
+      "loss": 0.03549227,
+      "step": 10322
+    },
+    {
+      "epoch": 20.646,
+      "grad_norm": 1.4373325109481812,
+      "learning_rate": 2e-05,
+      "loss": 0.03930869,
+      "step": 10323
+    },
+    {
+      "epoch": 20.648,
+      "grad_norm": 0.8817594647407532,
+      "learning_rate": 2e-05,
+      "loss": 0.0271128,
+      "step": 10324
+    },
+    {
+      "epoch": 20.65,
+      "grad_norm": 1.6336750984191895,
+      "learning_rate": 2e-05,
+      "loss": 0.03263681,
+      "step": 10325
+    },
+    {
+      "epoch": 20.652,
+      "grad_norm": 1.517414927482605,
+      "learning_rate": 2e-05,
+      "loss": 0.04617567,
+      "step": 10326
+    },
+    {
+      "epoch": 20.654,
+      "grad_norm": 4.604348182678223,
+      "learning_rate": 2e-05,
+      "loss": 0.02881363,
+      "step": 10327
+    },
+    {
+      "epoch": 20.656,
+      "grad_norm": 1.4577995538711548,
+      "learning_rate": 2e-05,
+      "loss": 0.03543629,
+      "step": 10328
+    },
+    {
+      "epoch": 20.658,
+      "grad_norm": 0.9887872338294983,
+      "learning_rate": 2e-05,
+      "loss": 0.02208835,
+      "step": 10329
+    },
+    {
+      "epoch": 20.66,
+      "grad_norm": 1.5389630794525146,
+      "learning_rate": 2e-05,
+      "loss": 0.02897654,
+      "step": 10330
+    },
+    {
+      "epoch": 20.662,
+      "grad_norm": 1.308394432067871,
+      "learning_rate": 2e-05,
+      "loss": 0.03866008,
+      "step": 10331
+    },
+    {
+      "epoch": 20.664,
+      "grad_norm": 0.8264133334159851,
+      "learning_rate": 2e-05,
+      "loss": 0.02847798,
+      "step": 10332
+    },
+    {
+      "epoch": 20.666,
+      "grad_norm": 1.7936815023422241,
+      "learning_rate": 2e-05,
+      "loss": 0.03791363,
+      "step": 10333
+    },
+    {
+      "epoch": 20.668,
+      "grad_norm": 1.3060663938522339,
+      "learning_rate": 2e-05,
+      "loss": 0.03301517,
+      "step": 10334
+    },
+    {
+      "epoch": 20.67,
+      "grad_norm": 1.709163784980774,
+      "learning_rate": 2e-05,
+      "loss": 0.03455461,
+      "step": 10335
+    },
+    {
+      "epoch": 20.672,
+      "grad_norm": 1.8301975727081299,
+      "learning_rate": 2e-05,
+      "loss": 0.02920924,
+      "step": 10336
+    },
+    {
+      "epoch": 20.674,
+      "grad_norm": 1.4292941093444824,
+      "learning_rate": 2e-05,
+      "loss": 0.03628965,
+      "step": 10337
+    },
+    {
+      "epoch": 20.676,
+      "grad_norm": 1.2578301429748535,
+      "learning_rate": 2e-05,
+      "loss": 0.02768565,
+      "step": 10338
+    },
+    {
+      "epoch": 20.678,
+      "grad_norm": 1.4743505716323853,
+      "learning_rate": 2e-05,
+      "loss": 0.03460295,
+      "step": 10339
+    },
+    {
+      "epoch": 20.68,
+      "grad_norm": 0.9008990526199341,
+      "learning_rate": 2e-05,
+      "loss": 0.02599203,
+      "step": 10340
+    },
+    {
+      "epoch": 20.682,
+      "grad_norm": 2.693477153778076,
+      "learning_rate": 2e-05,
+      "loss": 0.02892848,
+      "step": 10341
+    },
+    {
+      "epoch": 20.684,
+      "grad_norm": 1.2961777448654175,
+      "learning_rate": 2e-05,
+      "loss": 0.03386326,
+      "step": 10342
+    },
+    {
+      "epoch": 20.686,
+      "grad_norm": 1.1865119934082031,
+      "learning_rate": 2e-05,
+      "loss": 0.02271389,
+      "step": 10343
+    },
+    {
+      "epoch": 20.688,
+      "grad_norm": 1.4448591470718384,
+      "learning_rate": 2e-05,
+      "loss": 0.03761415,
+      "step": 10344
+    },
+    {
+      "epoch": 20.69,
+      "grad_norm": 1.2941032648086548,
+      "learning_rate": 2e-05,
+      "loss": 0.04060592,
+      "step": 10345
+    },
+    {
+      "epoch": 20.692,
+      "grad_norm": 1.4364266395568848,
+      "learning_rate": 2e-05,
+      "loss": 0.0526364,
+      "step": 10346
+    },
+    {
+      "epoch": 20.694,
+      "grad_norm": 1.228283405303955,
+      "learning_rate": 2e-05,
+      "loss": 0.03359667,
+      "step": 10347
+    },
+    {
+      "epoch": 20.696,
+      "grad_norm": 1.265254259109497,
+      "learning_rate": 2e-05,
+      "loss": 0.04058044,
+      "step": 10348
+    },
+    {
+      "epoch": 20.698,
+      "grad_norm": 1.298516035079956,
+      "learning_rate": 2e-05,
+      "loss": 0.04035057,
+      "step": 10349
+    },
+    {
+      "epoch": 20.7,
+      "grad_norm": 1.272953748703003,
+      "learning_rate": 2e-05,
+      "loss": 0.03609696,
+      "step": 10350
+    },
+    {
+      "epoch": 20.701999999999998,
+      "grad_norm": 1.0742324590682983,
+      "learning_rate": 2e-05,
+      "loss": 0.03620451,
+      "step": 10351
+    },
+    {
+      "epoch": 20.704,
+      "grad_norm": 1.4301596879959106,
+      "learning_rate": 2e-05,
+      "loss": 0.04321437,
+      "step": 10352
+    },
+    {
+      "epoch": 20.706,
+      "grad_norm": 1.9156485795974731,
+      "learning_rate": 2e-05,
+      "loss": 0.03270895,
+      "step": 10353
+    },
+    {
+      "epoch": 20.708,
+      "grad_norm": 1.2552930116653442,
+      "learning_rate": 2e-05,
+      "loss": 0.02602671,
+      "step": 10354
+    },
+    {
+      "epoch": 20.71,
+      "grad_norm": 1.193481683731079,
+      "learning_rate": 2e-05,
+      "loss": 0.03754304,
+      "step": 10355
+    },
+    {
+      "epoch": 20.712,
+      "grad_norm": 1.4162143468856812,
+      "learning_rate": 2e-05,
+      "loss": 0.04919789,
+      "step": 10356
+    },
+    {
+      "epoch": 20.714,
+      "grad_norm": 1.1660418510437012,
+      "learning_rate": 2e-05,
+      "loss": 0.03052988,
+      "step": 10357
+    },
+    {
+      "epoch": 20.716,
+      "grad_norm": 0.9906470775604248,
+      "learning_rate": 2e-05,
+      "loss": 0.0268365,
+      "step": 10358
+    },
+    {
+      "epoch": 20.718,
+      "grad_norm": 1.7477552890777588,
+      "learning_rate": 2e-05,
+      "loss": 0.05105488,
+      "step": 10359
+    },
+    {
+      "epoch": 20.72,
+      "grad_norm": 1.1876134872436523,
+      "learning_rate": 2e-05,
+      "loss": 0.02859118,
+      "step": 10360
+    },
+    {
+      "epoch": 20.722,
+      "grad_norm": 1.3102295398712158,
+      "learning_rate": 2e-05,
+      "loss": 0.03058487,
+      "step": 10361
+    },
+    {
+      "epoch": 20.724,
+      "grad_norm": 1.7961902618408203,
+      "learning_rate": 2e-05,
+      "loss": 0.04488046,
+      "step": 10362
+    },
+    {
+      "epoch": 20.726,
+      "grad_norm": 1.058826208114624,
+      "learning_rate": 2e-05,
+      "loss": 0.03246313,
+      "step": 10363
+    },
+    {
+      "epoch": 20.728,
+      "grad_norm": 1.0731538534164429,
+      "learning_rate": 2e-05,
+      "loss": 0.02816553,
+      "step": 10364
+    },
+    {
+      "epoch": 20.73,
+      "grad_norm": 2.4187989234924316,
+      "learning_rate": 2e-05,
+      "loss": 0.04112451,
+      "step": 10365
+    },
+    {
+      "epoch": 20.732,
+      "grad_norm": 1.7762290239334106,
+      "learning_rate": 2e-05,
+      "loss": 0.04738872,
+      "step": 10366
+    },
+    {
+      "epoch": 20.734,
+      "grad_norm": 1.1634255647659302,
+      "learning_rate": 2e-05,
+      "loss": 0.03303593,
+      "step": 10367
+    },
+    {
+      "epoch": 20.736,
+      "grad_norm": 1.1194519996643066,
+      "learning_rate": 2e-05,
+      "loss": 0.03090698,
+      "step": 10368
+    },
+    {
+      "epoch": 20.738,
+      "grad_norm": 1.9328291416168213,
+      "learning_rate": 2e-05,
+      "loss": 0.03827959,
+      "step": 10369
+    },
+    {
+      "epoch": 20.74,
+      "grad_norm": 0.9472059607505798,
+      "learning_rate": 2e-05,
+      "loss": 0.02610377,
+      "step": 10370
+    },
+    {
+      "epoch": 20.742,
+      "grad_norm": 1.0899560451507568,
+      "learning_rate": 2e-05,
+      "loss": 0.04154932,
+      "step": 10371
+    },
+    {
+      "epoch": 20.744,
+      "grad_norm": 1.071056604385376,
+      "learning_rate": 2e-05,
+      "loss": 0.02690003,
+      "step": 10372
+    },
+    {
+      "epoch": 20.746,
+      "grad_norm": 1.2740556001663208,
+      "learning_rate": 2e-05,
+      "loss": 0.03140601,
+      "step": 10373
+    },
+    {
+      "epoch": 20.748,
+      "grad_norm": 1.2829887866973877,
+      "learning_rate": 2e-05,
+      "loss": 0.0347194,
+      "step": 10374
+    },
+    {
+      "epoch": 20.75,
+      "grad_norm": 0.9616041779518127,
+      "learning_rate": 2e-05,
+      "loss": 0.03521022,
+      "step": 10375
+    },
+    {
+      "epoch": 20.752,
+      "grad_norm": 1.2607543468475342,
+      "learning_rate": 2e-05,
+      "loss": 0.03321379,
+      "step": 10376
+    },
+    {
+      "epoch": 20.754,
+      "grad_norm": 1.4216960668563843,
+      "learning_rate": 2e-05,
+      "loss": 0.03590911,
+      "step": 10377
+    },
+    {
+      "epoch": 20.756,
+      "grad_norm": 1.5020769834518433,
+      "learning_rate": 2e-05,
+      "loss": 0.04069631,
+      "step": 10378
+    },
+    {
+      "epoch": 20.758,
+      "grad_norm": 1.044640302658081,
+      "learning_rate": 2e-05,
+      "loss": 0.02541364,
+      "step": 10379
+    },
+    {
+      "epoch": 20.76,
+      "grad_norm": 1.2528002262115479,
+      "learning_rate": 2e-05,
+      "loss": 0.03949961,
+      "step": 10380
+    },
+    {
+      "epoch": 20.762,
+      "grad_norm": 1.23416006565094,
+      "learning_rate": 2e-05,
+      "loss": 0.02047839,
+      "step": 10381
+    },
+    {
+      "epoch": 20.764,
+      "grad_norm": 0.9270185232162476,
+      "learning_rate": 2e-05,
+      "loss": 0.03086235,
+      "step": 10382
+    },
+    {
+      "epoch": 20.766,
+      "grad_norm": 1.3459302186965942,
+      "learning_rate": 2e-05,
+      "loss": 0.04397811,
+      "step": 10383
+    },
+    {
+      "epoch": 20.768,
+      "grad_norm": 1.8959941864013672,
+      "learning_rate": 2e-05,
+      "loss": 0.02847361,
+      "step": 10384
+    },
+    {
+      "epoch": 20.77,
+      "grad_norm": 1.0249385833740234,
+      "learning_rate": 2e-05,
+      "loss": 0.02996654,
+      "step": 10385
+    },
+    {
+      "epoch": 20.772,
+      "grad_norm": 1.1461448669433594,
+      "learning_rate": 2e-05,
+      "loss": 0.04238908,
+      "step": 10386
+    },
+    {
+      "epoch": 20.774,
+      "grad_norm": 0.896342933177948,
+      "learning_rate": 2e-05,
+      "loss": 0.02898027,
+      "step": 10387
+    },
+    {
+      "epoch": 20.776,
+      "grad_norm": 1.2837777137756348,
+      "learning_rate": 2e-05,
+      "loss": 0.03371213,
+      "step": 10388
+    },
+    {
+      "epoch": 20.778,
+      "grad_norm": 1.4851652383804321,
+      "learning_rate": 2e-05,
+      "loss": 0.04785133,
+      "step": 10389
+    },
+    {
+      "epoch": 20.78,
+      "grad_norm": 1.1204262971878052,
+      "learning_rate": 2e-05,
+      "loss": 0.03940021,
+      "step": 10390
+    },
+    {
+      "epoch": 20.782,
+      "grad_norm": 1.6720068454742432,
+      "learning_rate": 2e-05,
+      "loss": 0.03579511,
+      "step": 10391
+    },
+    {
+      "epoch": 20.784,
+      "grad_norm": 0.9738611578941345,
+      "learning_rate": 2e-05,
+      "loss": 0.0249313,
+      "step": 10392
+    },
+    {
+      "epoch": 20.786,
+      "grad_norm": 1.8201998472213745,
+      "learning_rate": 2e-05,
+      "loss": 0.048961,
+      "step": 10393
+    },
+    {
+      "epoch": 20.788,
+      "grad_norm": 1.2513413429260254,
+      "learning_rate": 2e-05,
+      "loss": 0.02755917,
+      "step": 10394
+    },
+    {
+      "epoch": 20.79,
+      "grad_norm": 1.3315002918243408,
+      "learning_rate": 2e-05,
+      "loss": 0.04382675,
+      "step": 10395
+    },
+    {
+      "epoch": 20.792,
+      "grad_norm": 1.3850572109222412,
+      "learning_rate": 2e-05,
+      "loss": 0.04453446,
+      "step": 10396
+    },
+    {
+      "epoch": 20.794,
+      "grad_norm": 1.5482556819915771,
+      "learning_rate": 2e-05,
+      "loss": 0.04031963,
+      "step": 10397
+    },
+    {
+      "epoch": 20.796,
+      "grad_norm": 1.8607449531555176,
+      "learning_rate": 2e-05,
+      "loss": 0.05087737,
+      "step": 10398
+    },
+    {
+      "epoch": 20.798000000000002,
+      "grad_norm": 1.6714609861373901,
+      "learning_rate": 2e-05,
+      "loss": 0.04965856,
+      "step": 10399
+    },
+    {
+      "epoch": 20.8,
+      "grad_norm": 1.9360687732696533,
+      "learning_rate": 2e-05,
+      "loss": 0.04098749,
+      "step": 10400
+    },
+    {
+      "epoch": 20.802,
+      "grad_norm": 0.939067542552948,
+      "learning_rate": 2e-05,
+      "loss": 0.02232962,
+      "step": 10401
+    },
+    {
+      "epoch": 20.804,
+      "grad_norm": 1.1615712642669678,
+      "learning_rate": 2e-05,
+      "loss": 0.03574735,
+      "step": 10402
+    },
+    {
+      "epoch": 20.806,
+      "grad_norm": 1.2097338438034058,
+      "learning_rate": 2e-05,
+      "loss": 0.03018945,
+      "step": 10403
+    },
+    {
+      "epoch": 20.808,
+      "grad_norm": 1.0195422172546387,
+      "learning_rate": 2e-05,
+      "loss": 0.03126357,
+      "step": 10404
+    },
+    {
+      "epoch": 20.81,
+      "grad_norm": 1.9747889041900635,
+      "learning_rate": 2e-05,
+      "loss": 0.04224984,
+      "step": 10405
+    },
+    {
+      "epoch": 20.812,
+      "grad_norm": 1.3474302291870117,
+      "learning_rate": 2e-05,
+      "loss": 0.04074406,
+      "step": 10406
+    },
+    {
+      "epoch": 20.814,
+      "grad_norm": 1.5049179792404175,
+      "learning_rate": 2e-05,
+      "loss": 0.04507799,
+      "step": 10407
+    },
+    {
+      "epoch": 20.816,
+      "grad_norm": 1.5321961641311646,
+      "learning_rate": 2e-05,
+      "loss": 0.04021778,
+      "step": 10408
+    },
+    {
+      "epoch": 20.818,
+      "grad_norm": 0.8743321895599365,
+      "learning_rate": 2e-05,
+      "loss": 0.0247119,
+      "step": 10409
+    },
+    {
+      "epoch": 20.82,
+      "grad_norm": 1.6426585912704468,
+      "learning_rate": 2e-05,
+      "loss": 0.04337841,
+      "step": 10410
+    },
+    {
+      "epoch": 20.822,
+      "grad_norm": 2.048464059829712,
+      "learning_rate": 2e-05,
+      "loss": 0.04811566,
+      "step": 10411
+    },
+    {
+      "epoch": 20.824,
+      "grad_norm": 1.0955586433410645,
+      "learning_rate": 2e-05,
+      "loss": 0.02928637,
+      "step": 10412
+    },
+    {
+      "epoch": 20.826,
+      "grad_norm": 3.2851905822753906,
+      "learning_rate": 2e-05,
+      "loss": 0.03792008,
+      "step": 10413
+    },
+    {
+      "epoch": 20.828,
+      "grad_norm": 1.1193740367889404,
+      "learning_rate": 2e-05,
+      "loss": 0.03241657,
+      "step": 10414
+    },
+    {
+      "epoch": 20.83,
+      "grad_norm": 2.217038631439209,
+      "learning_rate": 2e-05,
+      "loss": 0.04027008,
+      "step": 10415
+    },
+    {
+      "epoch": 20.832,
+      "grad_norm": 1.1971979141235352,
+      "learning_rate": 2e-05,
+      "loss": 0.0313881,
+      "step": 10416
+    },
+    {
+      "epoch": 20.834,
+      "grad_norm": 1.2874393463134766,
+      "learning_rate": 2e-05,
+      "loss": 0.03874963,
+      "step": 10417
+    },
+    {
+      "epoch": 20.836,
+      "grad_norm": 1.7825161218643188,
+      "learning_rate": 2e-05,
+      "loss": 0.0345277,
+      "step": 10418
+    },
+    {
+      "epoch": 20.838,
+      "grad_norm": 2.392350196838379,
+      "learning_rate": 2e-05,
+      "loss": 0.05566229,
+      "step": 10419
+    },
+    {
+      "epoch": 20.84,
+      "grad_norm": 1.6109018325805664,
+      "learning_rate": 2e-05,
+      "loss": 0.04091034,
+      "step": 10420
+    },
+    {
+      "epoch": 20.842,
+      "grad_norm": 1.0898905992507935,
+      "learning_rate": 2e-05,
+      "loss": 0.03517729,
+      "step": 10421
+    },
+    {
+      "epoch": 20.844,
+      "grad_norm": 1.2754164934158325,
+      "learning_rate": 2e-05,
+      "loss": 0.03659736,
+      "step": 10422
+    },
+    {
+      "epoch": 20.846,
+      "grad_norm": 1.520416259765625,
+      "learning_rate": 2e-05,
+      "loss": 0.03716481,
+      "step": 10423
+    },
+    {
+      "epoch": 20.848,
+      "grad_norm": 2.5298376083374023,
+      "learning_rate": 2e-05,
+      "loss": 0.04783663,
+      "step": 10424
+    },
+    {
+      "epoch": 20.85,
+      "grad_norm": 1.6789484024047852,
+      "learning_rate": 2e-05,
+      "loss": 0.03685528,
+      "step": 10425
+    },
+    {
+      "epoch": 20.852,
+      "grad_norm": 1.5136722326278687,
+      "learning_rate": 2e-05,
+      "loss": 0.03236245,
+      "step": 10426
+    },
+    {
+      "epoch": 20.854,
+      "grad_norm": 0.9097597002983093,
+      "learning_rate": 2e-05,
+      "loss": 0.02598057,
+      "step": 10427
+    },
+    {
+      "epoch": 20.856,
+      "grad_norm": 1.2258105278015137,
+      "learning_rate": 2e-05,
+      "loss": 0.02003474,
+      "step": 10428
+    },
+    {
+      "epoch": 20.858,
+      "grad_norm": 1.2743656635284424,
+      "learning_rate": 2e-05,
+      "loss": 0.04647747,
+      "step": 10429
+    },
+    {
+      "epoch": 20.86,
+      "grad_norm": 1.199223279953003,
+      "learning_rate": 2e-05,
+      "loss": 0.03247061,
+      "step": 10430
+    },
+    {
+      "epoch": 20.862,
+      "grad_norm": 1.2215161323547363,
+      "learning_rate": 2e-05,
+      "loss": 0.03422792,
+      "step": 10431
+    },
+    {
+      "epoch": 20.864,
+      "grad_norm": 1.02229642868042,
+      "learning_rate": 2e-05,
+      "loss": 0.02587418,
+      "step": 10432
+    },
+    {
+      "epoch": 20.866,
+      "grad_norm": 1.7659459114074707,
+      "learning_rate": 2e-05,
+      "loss": 0.03618574,
+      "step": 10433
+    },
+    {
+      "epoch": 20.868,
+      "grad_norm": 1.2492812871932983,
+      "learning_rate": 2e-05,
+      "loss": 0.03435018,
+      "step": 10434
+    },
+    {
+      "epoch": 20.87,
+      "grad_norm": 1.9097106456756592,
+      "learning_rate": 2e-05,
+      "loss": 0.05099519,
+      "step": 10435
+    },
+    {
+      "epoch": 20.872,
+      "grad_norm": 2.3891992568969727,
+      "learning_rate": 2e-05,
+      "loss": 0.03493572,
+      "step": 10436
+    },
+    {
+      "epoch": 20.874,
+      "grad_norm": 6.847945690155029,
+      "learning_rate": 2e-05,
+      "loss": 0.04543907,
+      "step": 10437
+    },
+    {
+      "epoch": 20.876,
+      "grad_norm": 0.8078843951225281,
+      "learning_rate": 2e-05,
+      "loss": 0.01693185,
+      "step": 10438
+    },
+    {
+      "epoch": 20.878,
+      "grad_norm": 1.1145824193954468,
+      "learning_rate": 2e-05,
+      "loss": 0.03014626,
+      "step": 10439
+    },
+    {
+      "epoch": 20.88,
+      "grad_norm": 1.4058823585510254,
+      "learning_rate": 2e-05,
+      "loss": 0.03669397,
+      "step": 10440
+    },
+    {
+      "epoch": 20.882,
+      "grad_norm": 1.749841570854187,
+      "learning_rate": 2e-05,
+      "loss": 0.03852821,
+      "step": 10441
+    },
+    {
+      "epoch": 20.884,
+      "grad_norm": 1.228166937828064,
+      "learning_rate": 2e-05,
+      "loss": 0.04459646,
+      "step": 10442
+    },
+    {
+      "epoch": 20.886,
+      "grad_norm": 1.1540074348449707,
+      "learning_rate": 2e-05,
+      "loss": 0.02661585,
+      "step": 10443
+    },
+    {
+      "epoch": 20.888,
+      "grad_norm": 2.69791841506958,
+      "learning_rate": 2e-05,
+      "loss": 0.03573243,
+      "step": 10444
+    },
+    {
+      "epoch": 20.89,
+      "grad_norm": 1.062644124031067,
+      "learning_rate": 2e-05,
+      "loss": 0.02639816,
+      "step": 10445
+    },
+    {
+      "epoch": 20.892,
+      "grad_norm": 1.1434491872787476,
+      "learning_rate": 2e-05,
+      "loss": 0.03068436,
+      "step": 10446
+    },
+    {
+      "epoch": 20.894,
+      "grad_norm": 1.0184530019760132,
+      "learning_rate": 2e-05,
+      "loss": 0.02732141,
+      "step": 10447
+    },
+    {
+      "epoch": 20.896,
+      "grad_norm": 1.670699954032898,
+      "learning_rate": 2e-05,
+      "loss": 0.02273174,
+      "step": 10448
+    },
+    {
+      "epoch": 20.898,
+      "grad_norm": 1.2134953737258911,
+      "learning_rate": 2e-05,
+      "loss": 0.03250109,
+      "step": 10449
+    },
+    {
+      "epoch": 20.9,
+      "grad_norm": 1.2033954858779907,
+      "learning_rate": 2e-05,
+      "loss": 0.04112671,
+      "step": 10450
+    },
+    {
+      "epoch": 20.902,
+      "grad_norm": 1.2063027620315552,
+      "learning_rate": 2e-05,
+      "loss": 0.03256904,
+      "step": 10451
+    },
+    {
+      "epoch": 20.904,
+      "grad_norm": 1.2613540887832642,
+      "learning_rate": 2e-05,
+      "loss": 0.02923998,
+      "step": 10452
+    },
+    {
+      "epoch": 20.906,
+      "grad_norm": 1.8170700073242188,
+      "learning_rate": 2e-05,
+      "loss": 0.04404435,
+      "step": 10453
+    },
+    {
+      "epoch": 20.908,
+      "grad_norm": 2.074605703353882,
+      "learning_rate": 2e-05,
+      "loss": 0.06511399,
+      "step": 10454
+    },
+    {
+      "epoch": 20.91,
+      "grad_norm": 1.1869606971740723,
+      "learning_rate": 2e-05,
+      "loss": 0.03975774,
+      "step": 10455
+    },
+    {
+      "epoch": 20.912,
+      "grad_norm": 1.0519496202468872,
+      "learning_rate": 2e-05,
+      "loss": 0.03061169,
+      "step": 10456
+    },
+    {
+      "epoch": 20.914,
+      "grad_norm": 1.1407067775726318,
+      "learning_rate": 2e-05,
+      "loss": 0.02600056,
+      "step": 10457
+    },
+    {
+      "epoch": 20.916,
+      "grad_norm": 1.2726540565490723,
+      "learning_rate": 2e-05,
+      "loss": 0.03579668,
+      "step": 10458
+    },
+    {
+      "epoch": 20.918,
+      "grad_norm": 1.0577654838562012,
+      "learning_rate": 2e-05,
+      "loss": 0.0273657,
+      "step": 10459
+    },
+    {
+      "epoch": 20.92,
+      "grad_norm": 3.370119333267212,
+      "learning_rate": 2e-05,
+      "loss": 0.04482879,
+      "step": 10460
+    },
+    {
+      "epoch": 20.922,
+      "grad_norm": 0.9270197153091431,
+      "learning_rate": 2e-05,
+      "loss": 0.02642347,
+      "step": 10461
+    },
+    {
+      "epoch": 20.924,
+      "grad_norm": 2.1031179428100586,
+      "learning_rate": 2e-05,
+      "loss": 0.03457503,
+      "step": 10462
+    },
+    {
+      "epoch": 20.926,
+      "grad_norm": 1.47755765914917,
+      "learning_rate": 2e-05,
+      "loss": 0.04129061,
+      "step": 10463
+    },
+    {
+      "epoch": 20.928,
+      "grad_norm": 1.221456527709961,
+      "learning_rate": 2e-05,
+      "loss": 0.0394924,
+      "step": 10464
+    },
+    {
+      "epoch": 20.93,
+      "grad_norm": 1.5422698259353638,
+      "learning_rate": 2e-05,
+      "loss": 0.04629602,
+      "step": 10465
+    },
+    {
+      "epoch": 20.932,
+      "grad_norm": 1.2598649263381958,
+      "learning_rate": 2e-05,
+      "loss": 0.03605943,
+      "step": 10466
+    },
+    {
+      "epoch": 20.934,
+      "grad_norm": 1.0877323150634766,
+      "learning_rate": 2e-05,
+      "loss": 0.03485626,
+      "step": 10467
+    },
+    {
+      "epoch": 20.936,
+      "grad_norm": 0.8516746163368225,
+      "learning_rate": 2e-05,
+      "loss": 0.01967622,
+      "step": 10468
+    },
+    {
+      "epoch": 20.938,
+      "grad_norm": 1.01011061668396,
+      "learning_rate": 2e-05,
+      "loss": 0.0255489,
+      "step": 10469
+    },
+    {
+      "epoch": 20.94,
+      "grad_norm": 1.1266335248947144,
+      "learning_rate": 2e-05,
+      "loss": 0.02532486,
+      "step": 10470
+    },
+    {
+      "epoch": 20.942,
+      "grad_norm": 1.5323753356933594,
+      "learning_rate": 2e-05,
+      "loss": 0.03835236,
+      "step": 10471
+    },
+    {
+      "epoch": 20.944,
+      "grad_norm": 1.58487069606781,
+      "learning_rate": 2e-05,
+      "loss": 0.03736504,
+      "step": 10472
+    },
+    {
+      "epoch": 20.946,
+      "grad_norm": 1.2376452684402466,
+      "learning_rate": 2e-05,
+      "loss": 0.03332981,
+      "step": 10473
+    },
+    {
+      "epoch": 20.948,
+      "grad_norm": 1.3151975870132446,
+      "learning_rate": 2e-05,
+      "loss": 0.03374566,
+      "step": 10474
+    },
+    {
+      "epoch": 20.95,
+      "grad_norm": 1.410962700843811,
+      "learning_rate": 2e-05,
+      "loss": 0.04390511,
+      "step": 10475
+    },
+    {
+      "epoch": 20.951999999999998,
+      "grad_norm": 3.6582796573638916,
+      "learning_rate": 2e-05,
+      "loss": 0.04068881,
+      "step": 10476
+    },
+    {
+      "epoch": 20.954,
+      "grad_norm": 1.183980941772461,
+      "learning_rate": 2e-05,
+      "loss": 0.03768382,
+      "step": 10477
+    },
+    {
+      "epoch": 20.956,
+      "grad_norm": 1.9583686590194702,
+      "learning_rate": 2e-05,
+      "loss": 0.04036102,
+      "step": 10478
+    },
+    {
+      "epoch": 20.958,
+      "grad_norm": 1.1289726495742798,
+      "learning_rate": 2e-05,
+      "loss": 0.02689121,
+      "step": 10479
+    },
+    {
+      "epoch": 20.96,
+      "grad_norm": 1.1652100086212158,
+      "learning_rate": 2e-05,
+      "loss": 0.03687986,
+      "step": 10480
+    },
+    {
+      "epoch": 20.962,
+      "grad_norm": 1.4179235696792603,
+      "learning_rate": 2e-05,
+      "loss": 0.03052982,
+      "step": 10481
+    },
+    {
+      "epoch": 20.964,
+      "grad_norm": 0.9714958071708679,
+      "learning_rate": 2e-05,
+      "loss": 0.02744197,
+      "step": 10482
+    },
+    {
+      "epoch": 20.966,
+      "grad_norm": 1.1505253314971924,
+      "learning_rate": 2e-05,
+      "loss": 0.03075511,
+      "step": 10483
+    },
+    {
+      "epoch": 20.968,
+      "grad_norm": 1.3006511926651,
+      "learning_rate": 2e-05,
+      "loss": 0.04219224,
+      "step": 10484
+    },
+    {
+      "epoch": 20.97,
+      "grad_norm": 1.2317531108856201,
+      "learning_rate": 2e-05,
+      "loss": 0.0362198,
+      "step": 10485
+    },
+    {
+      "epoch": 20.972,
+      "grad_norm": 1.2480534315109253,
+      "learning_rate": 2e-05,
+      "loss": 0.04598153,
+      "step": 10486
+    },
+    {
+      "epoch": 20.974,
+      "grad_norm": 1.9260978698730469,
+      "learning_rate": 2e-05,
+      "loss": 0.05104748,
+      "step": 10487
+    },
+    {
+      "epoch": 20.976,
+      "grad_norm": 1.0552128553390503,
+      "learning_rate": 2e-05,
+      "loss": 0.02821925,
+      "step": 10488
+    },
+    {
+      "epoch": 20.978,
+      "grad_norm": 1.0819281339645386,
+      "learning_rate": 2e-05,
+      "loss": 0.03735295,
+      "step": 10489
+    },
+    {
+      "epoch": 20.98,
+      "grad_norm": 0.9959445595741272,
+      "learning_rate": 2e-05,
+      "loss": 0.03469016,
+      "step": 10490
+    },
+    {
+      "epoch": 20.982,
+      "grad_norm": 1.0732587575912476,
+      "learning_rate": 2e-05,
+      "loss": 0.02674491,
+      "step": 10491
+    },
+    {
+      "epoch": 20.984,
+      "grad_norm": 1.1409895420074463,
+      "learning_rate": 2e-05,
+      "loss": 0.02786115,
+      "step": 10492
+    },
+    {
+      "epoch": 20.986,
+      "grad_norm": 0.8526608943939209,
+      "learning_rate": 2e-05,
+      "loss": 0.02098497,
+      "step": 10493
+    },
+    {
+      "epoch": 20.988,
+      "grad_norm": 1.0358623266220093,
+      "learning_rate": 2e-05,
+      "loss": 0.02189691,
+      "step": 10494
+    },
+    {
+      "epoch": 20.99,
+      "grad_norm": 1.4046947956085205,
+      "learning_rate": 2e-05,
+      "loss": 0.0338182,
+      "step": 10495
+    },
+    {
+      "epoch": 20.992,
+      "grad_norm": 1.1783146858215332,
+      "learning_rate": 2e-05,
+      "loss": 0.03898209,
+      "step": 10496
+    },
+    {
+      "epoch": 20.994,
+      "grad_norm": 0.9742079973220825,
+      "learning_rate": 2e-05,
+      "loss": 0.02756248,
+      "step": 10497
+    },
+    {
+      "epoch": 20.996,
+      "grad_norm": 2.197392463684082,
+      "learning_rate": 2e-05,
+      "loss": 0.04526626,
+      "step": 10498
+    },
+    {
+      "epoch": 20.998,
+      "grad_norm": 1.057904601097107,
+      "learning_rate": 2e-05,
+      "loss": 0.04241625,
+      "step": 10499
+    },
+    {
+      "epoch": 21.0,
+      "grad_norm": 1.4337189197540283,
+      "learning_rate": 2e-05,
+      "loss": 0.03345023,
+      "step": 10500
+    },
+    {
+      "epoch": 21.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.998,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9740518962075848,
+        "Equal_1": 0.998,
+        "Equal_2": 0.9820359281437125,
+        "Equal_3": 0.9680638722554891,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.998003992015968,
+        "Parallel_1": 0.9879759519038076,
+        "Parallel_2": 0.9899799599198397,
+        "Parallel_3": 0.986,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.984,
+        "Perpendicular_3": 0.8186372745490982,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.994,
+        "PointLiesOnCircle_3": 0.988,
+        "PointLiesOnLine_1": 0.9959919839679359,
+        "PointLiesOnLine_2": 0.9899799599198397,
+        "PointLiesOnLine_3": 0.9840319361277445
+      },
+      "eval_runtime": 319.6633,
+      "eval_samples_per_second": 32.847,
+      "eval_steps_per_second": 0.657,
+      "step": 10500
+    },
+    {
+      "epoch": 21.002,
+      "grad_norm": 1.4172569513320923,
+      "learning_rate": 2e-05,
+      "loss": 0.04958025,
+      "step": 10501
+    },
+    {
+      "epoch": 21.004,
+      "grad_norm": 1.2641007900238037,
+      "learning_rate": 2e-05,
+      "loss": 0.04138844,
+      "step": 10502
+    },
+    {
+      "epoch": 21.006,
+      "grad_norm": 1.5297917127609253,
+      "learning_rate": 2e-05,
+      "loss": 0.05362614,
+      "step": 10503
+    },
+    {
+      "epoch": 21.008,
+      "grad_norm": 1.4040725231170654,
+      "learning_rate": 2e-05,
+      "loss": 0.02881721,
+      "step": 10504
+    },
+    {
+      "epoch": 21.01,
+      "grad_norm": 2.7434298992156982,
+      "learning_rate": 2e-05,
+      "loss": 0.05710435,
+      "step": 10505
+    },
+    {
+      "epoch": 21.012,
+      "grad_norm": 0.9089720845222473,
+      "learning_rate": 2e-05,
+      "loss": 0.02749285,
+      "step": 10506
+    },
+    {
+      "epoch": 21.014,
+      "grad_norm": 1.093634009361267,
+      "learning_rate": 2e-05,
+      "loss": 0.02694854,
+      "step": 10507
+    },
+    {
+      "epoch": 21.016,
+      "grad_norm": 0.8842881917953491,
+      "learning_rate": 2e-05,
+      "loss": 0.02866191,
+      "step": 10508
+    },
+    {
+      "epoch": 21.018,
+      "grad_norm": 1.1293853521347046,
+      "learning_rate": 2e-05,
+      "loss": 0.03138885,
+      "step": 10509
+    },
+    {
+      "epoch": 21.02,
+      "grad_norm": 2.5073678493499756,
+      "learning_rate": 2e-05,
+      "loss": 0.036281,
+      "step": 10510
+    },
+    {
+      "epoch": 21.022,
+      "grad_norm": 0.8775820136070251,
+      "learning_rate": 2e-05,
+      "loss": 0.0302633,
+      "step": 10511
+    },
+    {
+      "epoch": 21.024,
+      "grad_norm": 1.844066858291626,
+      "learning_rate": 2e-05,
+      "loss": 0.05106075,
+      "step": 10512
+    },
+    {
+      "epoch": 21.026,
+      "grad_norm": 1.2659056186676025,
+      "learning_rate": 2e-05,
+      "loss": 0.03020029,
+      "step": 10513
+    },
+    {
+      "epoch": 21.028,
+      "grad_norm": 1.132651686668396,
+      "learning_rate": 2e-05,
+      "loss": 0.02988936,
+      "step": 10514
+    },
+    {
+      "epoch": 21.03,
+      "grad_norm": 0.9240502119064331,
+      "learning_rate": 2e-05,
+      "loss": 0.02960332,
+      "step": 10515
+    },
+    {
+      "epoch": 21.032,
+      "grad_norm": 3.4578001499176025,
+      "learning_rate": 2e-05,
+      "loss": 0.03723548,
+      "step": 10516
+    },
+    {
+      "epoch": 21.034,
+      "grad_norm": 1.327440619468689,
+      "learning_rate": 2e-05,
+      "loss": 0.04209972,
+      "step": 10517
+    },
+    {
+      "epoch": 21.036,
+      "grad_norm": 0.7104504704475403,
+      "learning_rate": 2e-05,
+      "loss": 0.01660552,
+      "step": 10518
+    },
+    {
+      "epoch": 21.038,
+      "grad_norm": 1.9451689720153809,
+      "learning_rate": 2e-05,
+      "loss": 0.04870243,
+      "step": 10519
+    },
+    {
+      "epoch": 21.04,
+      "grad_norm": 1.5693484544754028,
+      "learning_rate": 2e-05,
+      "loss": 0.05037589,
+      "step": 10520
+    },
+    {
+      "epoch": 21.042,
+      "grad_norm": 1.0665810108184814,
+      "learning_rate": 2e-05,
+      "loss": 0.03087834,
+      "step": 10521
+    },
+    {
+      "epoch": 21.044,
+      "grad_norm": 1.309026837348938,
+      "learning_rate": 2e-05,
+      "loss": 0.03562339,
+      "step": 10522
+    },
+    {
+      "epoch": 21.046,
+      "grad_norm": 1.5604883432388306,
+      "learning_rate": 2e-05,
+      "loss": 0.044502,
+      "step": 10523
+    },
+    {
+      "epoch": 21.048,
+      "grad_norm": 0.9193902015686035,
+      "learning_rate": 2e-05,
+      "loss": 0.02742101,
+      "step": 10524
+    },
+    {
+      "epoch": 21.05,
+      "grad_norm": 1.5470359325408936,
+      "learning_rate": 2e-05,
+      "loss": 0.03719456,
+      "step": 10525
+    },
+    {
+      "epoch": 21.052,
+      "grad_norm": 0.9354279637336731,
+      "learning_rate": 2e-05,
+      "loss": 0.0307904,
+      "step": 10526
+    },
+    {
+      "epoch": 21.054,
+      "grad_norm": 0.9876614212989807,
+      "learning_rate": 2e-05,
+      "loss": 0.03531288,
+      "step": 10527
+    },
+    {
+      "epoch": 21.056,
+      "grad_norm": 1.2255337238311768,
+      "learning_rate": 2e-05,
+      "loss": 0.03369109,
+      "step": 10528
+    },
+    {
+      "epoch": 21.058,
+      "grad_norm": 1.4609296321868896,
+      "learning_rate": 2e-05,
+      "loss": 0.04581051,
+      "step": 10529
+    },
+    {
+      "epoch": 21.06,
+      "grad_norm": 1.1844290494918823,
+      "learning_rate": 2e-05,
+      "loss": 0.02935641,
+      "step": 10530
+    },
+    {
+      "epoch": 21.062,
+      "grad_norm": 1.3473209142684937,
+      "learning_rate": 2e-05,
+      "loss": 0.03874001,
+      "step": 10531
+    },
+    {
+      "epoch": 21.064,
+      "grad_norm": 1.2861710786819458,
+      "learning_rate": 2e-05,
+      "loss": 0.03785869,
+      "step": 10532
+    },
+    {
+      "epoch": 21.066,
+      "grad_norm": 1.2455552816390991,
+      "learning_rate": 2e-05,
+      "loss": 0.02974376,
+      "step": 10533
+    },
+    {
+      "epoch": 21.068,
+      "grad_norm": 1.0268170833587646,
+      "learning_rate": 2e-05,
+      "loss": 0.0280002,
+      "step": 10534
+    },
+    {
+      "epoch": 21.07,
+      "grad_norm": 2.149333953857422,
+      "learning_rate": 2e-05,
+      "loss": 0.0475785,
+      "step": 10535
+    },
+    {
+      "epoch": 21.072,
+      "grad_norm": 0.8871065378189087,
+      "learning_rate": 2e-05,
+      "loss": 0.02665732,
+      "step": 10536
+    },
+    {
+      "epoch": 21.074,
+      "grad_norm": 1.2057782411575317,
+      "learning_rate": 2e-05,
+      "loss": 0.04148353,
+      "step": 10537
+    },
+    {
+      "epoch": 21.076,
+      "grad_norm": 2.8635036945343018,
+      "learning_rate": 2e-05,
+      "loss": 0.04504614,
+      "step": 10538
+    },
+    {
+      "epoch": 21.078,
+      "grad_norm": 1.0987458229064941,
+      "learning_rate": 2e-05,
+      "loss": 0.02471746,
+      "step": 10539
+    },
+    {
+      "epoch": 21.08,
+      "grad_norm": 1.5803258419036865,
+      "learning_rate": 2e-05,
+      "loss": 0.04167273,
+      "step": 10540
+    },
+    {
+      "epoch": 21.082,
+      "grad_norm": 0.9525123834609985,
+      "learning_rate": 2e-05,
+      "loss": 0.02323587,
+      "step": 10541
+    },
+    {
+      "epoch": 21.084,
+      "grad_norm": 1.051931619644165,
+      "learning_rate": 2e-05,
+      "loss": 0.03668146,
+      "step": 10542
+    },
+    {
+      "epoch": 21.086,
+      "grad_norm": 1.020364761352539,
+      "learning_rate": 2e-05,
+      "loss": 0.03155644,
+      "step": 10543
+    },
+    {
+      "epoch": 21.088,
+      "grad_norm": 1.9535802602767944,
+      "learning_rate": 2e-05,
+      "loss": 0.04253891,
+      "step": 10544
+    },
+    {
+      "epoch": 21.09,
+      "grad_norm": 1.563684105873108,
+      "learning_rate": 2e-05,
+      "loss": 0.03040604,
+      "step": 10545
+    },
+    {
+      "epoch": 21.092,
+      "grad_norm": 1.5148497819900513,
+      "learning_rate": 2e-05,
+      "loss": 0.04719204,
+      "step": 10546
+    },
+    {
+      "epoch": 21.094,
+      "grad_norm": 1.0597929954528809,
+      "learning_rate": 2e-05,
+      "loss": 0.03394167,
+      "step": 10547
+    },
+    {
+      "epoch": 21.096,
+      "grad_norm": 0.9946009516716003,
+      "learning_rate": 2e-05,
+      "loss": 0.02729562,
+      "step": 10548
+    },
+    {
+      "epoch": 21.098,
+      "grad_norm": 1.0386313199996948,
+      "learning_rate": 2e-05,
+      "loss": 0.02707431,
+      "step": 10549
+    },
+    {
+      "epoch": 21.1,
+      "grad_norm": 1.930969476699829,
+      "learning_rate": 2e-05,
+      "loss": 0.03724827,
+      "step": 10550
+    },
+    {
+      "epoch": 21.102,
+      "grad_norm": 0.8346301913261414,
+      "learning_rate": 2e-05,
+      "loss": 0.02298054,
+      "step": 10551
+    },
+    {
+      "epoch": 21.104,
+      "grad_norm": 1.1870561838150024,
+      "learning_rate": 2e-05,
+      "loss": 0.03454594,
+      "step": 10552
+    },
+    {
+      "epoch": 21.106,
+      "grad_norm": 1.3278943300247192,
+      "learning_rate": 2e-05,
+      "loss": 0.03413796,
+      "step": 10553
+    },
+    {
+      "epoch": 21.108,
+      "grad_norm": 0.9806317090988159,
+      "learning_rate": 2e-05,
+      "loss": 0.03050506,
+      "step": 10554
+    },
+    {
+      "epoch": 21.11,
+      "grad_norm": 1.6823358535766602,
+      "learning_rate": 2e-05,
+      "loss": 0.04869309,
+      "step": 10555
+    },
+    {
+      "epoch": 21.112,
+      "grad_norm": 0.9092576503753662,
+      "learning_rate": 2e-05,
+      "loss": 0.02638655,
+      "step": 10556
+    },
+    {
+      "epoch": 21.114,
+      "grad_norm": 1.2272701263427734,
+      "learning_rate": 2e-05,
+      "loss": 0.03748032,
+      "step": 10557
+    },
+    {
+      "epoch": 21.116,
+      "grad_norm": 1.1986057758331299,
+      "learning_rate": 2e-05,
+      "loss": 0.02931621,
+      "step": 10558
+    },
+    {
+      "epoch": 21.118,
+      "grad_norm": 1.966601014137268,
+      "learning_rate": 2e-05,
+      "loss": 0.04338966,
+      "step": 10559
+    },
+    {
+      "epoch": 21.12,
+      "grad_norm": 2.9418132305145264,
+      "learning_rate": 2e-05,
+      "loss": 0.05720629,
+      "step": 10560
+    },
+    {
+      "epoch": 21.122,
+      "grad_norm": 1.3184374570846558,
+      "learning_rate": 2e-05,
+      "loss": 0.03051766,
+      "step": 10561
+    },
+    {
+      "epoch": 21.124,
+      "grad_norm": 1.2324519157409668,
+      "learning_rate": 2e-05,
+      "loss": 0.04240867,
+      "step": 10562
+    },
+    {
+      "epoch": 21.126,
+      "grad_norm": 1.0354188680648804,
+      "learning_rate": 2e-05,
+      "loss": 0.03184131,
+      "step": 10563
+    },
+    {
+      "epoch": 21.128,
+      "grad_norm": 1.3084588050842285,
+      "learning_rate": 2e-05,
+      "loss": 0.04317082,
+      "step": 10564
+    },
+    {
+      "epoch": 21.13,
+      "grad_norm": 1.621947169303894,
+      "learning_rate": 2e-05,
+      "loss": 0.04106213,
+      "step": 10565
+    },
+    {
+      "epoch": 21.132,
+      "grad_norm": 2.01584529876709,
+      "learning_rate": 2e-05,
+      "loss": 0.04236465,
+      "step": 10566
+    },
+    {
+      "epoch": 21.134,
+      "grad_norm": 1.481381893157959,
+      "learning_rate": 2e-05,
+      "loss": 0.02457852,
+      "step": 10567
+    },
+    {
+      "epoch": 21.136,
+      "grad_norm": 1.3118445873260498,
+      "learning_rate": 2e-05,
+      "loss": 0.04216508,
+      "step": 10568
+    },
+    {
+      "epoch": 21.138,
+      "grad_norm": 1.3927664756774902,
+      "learning_rate": 2e-05,
+      "loss": 0.04639062,
+      "step": 10569
+    },
+    {
+      "epoch": 21.14,
+      "grad_norm": 1.2951942682266235,
+      "learning_rate": 2e-05,
+      "loss": 0.03051125,
+      "step": 10570
+    },
+    {
+      "epoch": 21.142,
+      "grad_norm": 3.7828381061553955,
+      "learning_rate": 2e-05,
+      "loss": 0.03892427,
+      "step": 10571
+    },
+    {
+      "epoch": 21.144,
+      "grad_norm": 1.8915671110153198,
+      "learning_rate": 2e-05,
+      "loss": 0.03912808,
+      "step": 10572
+    },
+    {
+      "epoch": 21.146,
+      "grad_norm": 1.7816556692123413,
+      "learning_rate": 2e-05,
+      "loss": 0.03675058,
+      "step": 10573
+    },
+    {
+      "epoch": 21.148,
+      "grad_norm": 2.443760395050049,
+      "learning_rate": 2e-05,
+      "loss": 0.03171434,
+      "step": 10574
+    },
+    {
+      "epoch": 21.15,
+      "grad_norm": 1.0887539386749268,
+      "learning_rate": 2e-05,
+      "loss": 0.03385027,
+      "step": 10575
+    },
+    {
+      "epoch": 21.152,
+      "grad_norm": 1.2952345609664917,
+      "learning_rate": 2e-05,
+      "loss": 0.04195005,
+      "step": 10576
+    },
+    {
+      "epoch": 21.154,
+      "grad_norm": 1.1885733604431152,
+      "learning_rate": 2e-05,
+      "loss": 0.02599936,
+      "step": 10577
+    },
+    {
+      "epoch": 21.156,
+      "grad_norm": 1.1241782903671265,
+      "learning_rate": 2e-05,
+      "loss": 0.03315149,
+      "step": 10578
+    },
+    {
+      "epoch": 21.158,
+      "grad_norm": 1.3803486824035645,
+      "learning_rate": 2e-05,
+      "loss": 0.03571745,
+      "step": 10579
+    },
+    {
+      "epoch": 21.16,
+      "grad_norm": 1.2156238555908203,
+      "learning_rate": 2e-05,
+      "loss": 0.04436778,
+      "step": 10580
+    },
+    {
+      "epoch": 21.162,
+      "grad_norm": 1.586046576499939,
+      "learning_rate": 2e-05,
+      "loss": 0.04987106,
+      "step": 10581
+    },
+    {
+      "epoch": 21.164,
+      "grad_norm": 1.261434555053711,
+      "learning_rate": 2e-05,
+      "loss": 0.03705215,
+      "step": 10582
+    },
+    {
+      "epoch": 21.166,
+      "grad_norm": 0.886199414730072,
+      "learning_rate": 2e-05,
+      "loss": 0.02416335,
+      "step": 10583
+    },
+    {
+      "epoch": 21.168,
+      "grad_norm": 1.2906107902526855,
+      "learning_rate": 2e-05,
+      "loss": 0.04419378,
+      "step": 10584
+    },
+    {
+      "epoch": 21.17,
+      "grad_norm": 1.1108649969100952,
+      "learning_rate": 2e-05,
+      "loss": 0.03079815,
+      "step": 10585
+    },
+    {
+      "epoch": 21.172,
+      "grad_norm": 2.070798635482788,
+      "learning_rate": 2e-05,
+      "loss": 0.0433712,
+      "step": 10586
+    },
+    {
+      "epoch": 21.174,
+      "grad_norm": 1.1178715229034424,
+      "learning_rate": 2e-05,
+      "loss": 0.03338214,
+      "step": 10587
+    },
+    {
+      "epoch": 21.176,
+      "grad_norm": 1.2780183553695679,
+      "learning_rate": 2e-05,
+      "loss": 0.04754562,
+      "step": 10588
+    },
+    {
+      "epoch": 21.178,
+      "grad_norm": 1.154023289680481,
+      "learning_rate": 2e-05,
+      "loss": 0.03899136,
+      "step": 10589
+    },
+    {
+      "epoch": 21.18,
+      "grad_norm": 0.9963793158531189,
+      "learning_rate": 2e-05,
+      "loss": 0.02338295,
+      "step": 10590
+    },
+    {
+      "epoch": 21.182,
+      "grad_norm": 1.49871027469635,
+      "learning_rate": 2e-05,
+      "loss": 0.04136465,
+      "step": 10591
+    },
+    {
+      "epoch": 21.184,
+      "grad_norm": 1.006020426750183,
+      "learning_rate": 2e-05,
+      "loss": 0.02428103,
+      "step": 10592
+    },
+    {
+      "epoch": 21.186,
+      "grad_norm": 1.1841108798980713,
+      "learning_rate": 2e-05,
+      "loss": 0.03439215,
+      "step": 10593
+    },
+    {
+      "epoch": 21.188,
+      "grad_norm": 1.2510030269622803,
+      "learning_rate": 2e-05,
+      "loss": 0.03146035,
+      "step": 10594
+    },
+    {
+      "epoch": 21.19,
+      "grad_norm": 0.7315729260444641,
+      "learning_rate": 2e-05,
+      "loss": 0.01793394,
+      "step": 10595
+    },
+    {
+      "epoch": 21.192,
+      "grad_norm": 1.2097729444503784,
+      "learning_rate": 2e-05,
+      "loss": 0.03397623,
+      "step": 10596
+    },
+    {
+      "epoch": 21.194,
+      "grad_norm": 1.0274741649627686,
+      "learning_rate": 2e-05,
+      "loss": 0.02825071,
+      "step": 10597
+    },
+    {
+      "epoch": 21.196,
+      "grad_norm": 0.9847630262374878,
+      "learning_rate": 2e-05,
+      "loss": 0.03145049,
+      "step": 10598
+    },
+    {
+      "epoch": 21.198,
+      "grad_norm": 1.3303278684616089,
+      "learning_rate": 2e-05,
+      "loss": 0.04809329,
+      "step": 10599
+    },
+    {
+      "epoch": 21.2,
+      "grad_norm": 1.0973892211914062,
+      "learning_rate": 2e-05,
+      "loss": 0.03610747,
+      "step": 10600
+    },
+    {
+      "epoch": 21.202,
+      "grad_norm": 1.8867298364639282,
+      "learning_rate": 2e-05,
+      "loss": 0.04427704,
+      "step": 10601
+    },
+    {
+      "epoch": 21.204,
+      "grad_norm": 1.922042965888977,
+      "learning_rate": 2e-05,
+      "loss": 0.0379353,
+      "step": 10602
+    },
+    {
+      "epoch": 21.206,
+      "grad_norm": 1.1012451648712158,
+      "learning_rate": 2e-05,
+      "loss": 0.03171213,
+      "step": 10603
+    },
+    {
+      "epoch": 21.208,
+      "grad_norm": 1.0954500436782837,
+      "learning_rate": 2e-05,
+      "loss": 0.02412405,
+      "step": 10604
+    },
+    {
+      "epoch": 21.21,
+      "grad_norm": 1.3452584743499756,
+      "learning_rate": 2e-05,
+      "loss": 0.0347417,
+      "step": 10605
+    },
+    {
+      "epoch": 21.212,
+      "grad_norm": 1.3375858068466187,
+      "learning_rate": 2e-05,
+      "loss": 0.03575499,
+      "step": 10606
+    },
+    {
+      "epoch": 21.214,
+      "grad_norm": 1.139094352722168,
+      "learning_rate": 2e-05,
+      "loss": 0.0376694,
+      "step": 10607
+    },
+    {
+      "epoch": 21.216,
+      "grad_norm": 1.5566028356552124,
+      "learning_rate": 2e-05,
+      "loss": 0.03238266,
+      "step": 10608
+    },
+    {
+      "epoch": 21.218,
+      "grad_norm": 1.9407174587249756,
+      "learning_rate": 2e-05,
+      "loss": 0.04087084,
+      "step": 10609
+    },
+    {
+      "epoch": 21.22,
+      "grad_norm": 5.404819011688232,
+      "learning_rate": 2e-05,
+      "loss": 0.04306193,
+      "step": 10610
+    },
+    {
+      "epoch": 21.222,
+      "grad_norm": 1.1310769319534302,
+      "learning_rate": 2e-05,
+      "loss": 0.03149523,
+      "step": 10611
+    },
+    {
+      "epoch": 21.224,
+      "grad_norm": 2.3743486404418945,
+      "learning_rate": 2e-05,
+      "loss": 0.0347545,
+      "step": 10612
+    },
+    {
+      "epoch": 21.226,
+      "grad_norm": 1.9811265468597412,
+      "learning_rate": 2e-05,
+      "loss": 0.04471551,
+      "step": 10613
+    },
+    {
+      "epoch": 21.228,
+      "grad_norm": 1.034940242767334,
+      "learning_rate": 2e-05,
+      "loss": 0.02823063,
+      "step": 10614
+    },
+    {
+      "epoch": 21.23,
+      "grad_norm": 1.0800728797912598,
+      "learning_rate": 2e-05,
+      "loss": 0.04498807,
+      "step": 10615
+    },
+    {
+      "epoch": 21.232,
+      "grad_norm": 1.2216445207595825,
+      "learning_rate": 2e-05,
+      "loss": 0.03705755,
+      "step": 10616
+    },
+    {
+      "epoch": 21.234,
+      "grad_norm": 1.5470830202102661,
+      "learning_rate": 2e-05,
+      "loss": 0.03225365,
+      "step": 10617
+    },
+    {
+      "epoch": 21.236,
+      "grad_norm": 2.094444990158081,
+      "learning_rate": 2e-05,
+      "loss": 0.04915155,
+      "step": 10618
+    },
+    {
+      "epoch": 21.238,
+      "grad_norm": 1.5187581777572632,
+      "learning_rate": 2e-05,
+      "loss": 0.04350504,
+      "step": 10619
+    },
+    {
+      "epoch": 21.24,
+      "grad_norm": 1.3666698932647705,
+      "learning_rate": 2e-05,
+      "loss": 0.05134146,
+      "step": 10620
+    },
+    {
+      "epoch": 21.242,
+      "grad_norm": 0.9730545282363892,
+      "learning_rate": 2e-05,
+      "loss": 0.0246509,
+      "step": 10621
+    },
+    {
+      "epoch": 21.244,
+      "grad_norm": 1.3245638608932495,
+      "learning_rate": 2e-05,
+      "loss": 0.03410356,
+      "step": 10622
+    },
+    {
+      "epoch": 21.246,
+      "grad_norm": 1.0870567560195923,
+      "learning_rate": 2e-05,
+      "loss": 0.03783543,
+      "step": 10623
+    },
+    {
+      "epoch": 21.248,
+      "grad_norm": 1.4121298789978027,
+      "learning_rate": 2e-05,
+      "loss": 0.0571961,
+      "step": 10624
+    },
+    {
+      "epoch": 21.25,
+      "grad_norm": 1.5266447067260742,
+      "learning_rate": 2e-05,
+      "loss": 0.04245551,
+      "step": 10625
+    },
+    {
+      "epoch": 21.252,
+      "grad_norm": 1.5656712055206299,
+      "learning_rate": 2e-05,
+      "loss": 0.04033789,
+      "step": 10626
+    },
+    {
+      "epoch": 21.254,
+      "grad_norm": 1.1030497550964355,
+      "learning_rate": 2e-05,
+      "loss": 0.03589434,
+      "step": 10627
+    },
+    {
+      "epoch": 21.256,
+      "grad_norm": 1.4850778579711914,
+      "learning_rate": 2e-05,
+      "loss": 0.03722943,
+      "step": 10628
+    },
+    {
+      "epoch": 21.258,
+      "grad_norm": 1.0993595123291016,
+      "learning_rate": 2e-05,
+      "loss": 0.02807938,
+      "step": 10629
+    },
+    {
+      "epoch": 21.26,
+      "grad_norm": 1.2334455251693726,
+      "learning_rate": 2e-05,
+      "loss": 0.03748228,
+      "step": 10630
+    },
+    {
+      "epoch": 21.262,
+      "grad_norm": 1.5018088817596436,
+      "learning_rate": 2e-05,
+      "loss": 0.04651701,
+      "step": 10631
+    },
+    {
+      "epoch": 21.264,
+      "grad_norm": 1.5770760774612427,
+      "learning_rate": 2e-05,
+      "loss": 0.03737814,
+      "step": 10632
+    },
+    {
+      "epoch": 21.266,
+      "grad_norm": 1.3698883056640625,
+      "learning_rate": 2e-05,
+      "loss": 0.03702818,
+      "step": 10633
+    },
+    {
+      "epoch": 21.268,
+      "grad_norm": 1.316947340965271,
+      "learning_rate": 2e-05,
+      "loss": 0.0374598,
+      "step": 10634
+    },
+    {
+      "epoch": 21.27,
+      "grad_norm": 1.1345027685165405,
+      "learning_rate": 2e-05,
+      "loss": 0.03323581,
+      "step": 10635
+    },
+    {
+      "epoch": 21.272,
+      "grad_norm": 1.6421146392822266,
+      "learning_rate": 2e-05,
+      "loss": 0.03895211,
+      "step": 10636
+    },
+    {
+      "epoch": 21.274,
+      "grad_norm": 0.812617301940918,
+      "learning_rate": 2e-05,
+      "loss": 0.02046642,
+      "step": 10637
+    },
+    {
+      "epoch": 21.276,
+      "grad_norm": 1.6946412324905396,
+      "learning_rate": 2e-05,
+      "loss": 0.03659395,
+      "step": 10638
+    },
+    {
+      "epoch": 21.278,
+      "grad_norm": 2.288081645965576,
+      "learning_rate": 2e-05,
+      "loss": 0.05296513,
+      "step": 10639
+    },
+    {
+      "epoch": 21.28,
+      "grad_norm": 1.681496500968933,
+      "learning_rate": 2e-05,
+      "loss": 0.03126854,
+      "step": 10640
+    },
+    {
+      "epoch": 21.282,
+      "grad_norm": 3.044177532196045,
+      "learning_rate": 2e-05,
+      "loss": 0.0360253,
+      "step": 10641
+    },
+    {
+      "epoch": 21.284,
+      "grad_norm": 1.5990822315216064,
+      "learning_rate": 2e-05,
+      "loss": 0.03093606,
+      "step": 10642
+    },
+    {
+      "epoch": 21.286,
+      "grad_norm": 1.1280889511108398,
+      "learning_rate": 2e-05,
+      "loss": 0.0305376,
+      "step": 10643
+    },
+    {
+      "epoch": 21.288,
+      "grad_norm": 0.7883005738258362,
+      "learning_rate": 2e-05,
+      "loss": 0.02066907,
+      "step": 10644
+    },
+    {
+      "epoch": 21.29,
+      "grad_norm": 0.9310852885246277,
+      "learning_rate": 2e-05,
+      "loss": 0.02677006,
+      "step": 10645
+    },
+    {
+      "epoch": 21.292,
+      "grad_norm": 1.0146135091781616,
+      "learning_rate": 2e-05,
+      "loss": 0.03050572,
+      "step": 10646
+    },
+    {
+      "epoch": 21.294,
+      "grad_norm": 1.7839759588241577,
+      "learning_rate": 2e-05,
+      "loss": 0.03637395,
+      "step": 10647
+    },
+    {
+      "epoch": 21.296,
+      "grad_norm": 1.1365673542022705,
+      "learning_rate": 2e-05,
+      "loss": 0.03653955,
+      "step": 10648
+    },
+    {
+      "epoch": 21.298,
+      "grad_norm": 1.2921481132507324,
+      "learning_rate": 2e-05,
+      "loss": 0.04410624,
+      "step": 10649
+    },
+    {
+      "epoch": 21.3,
+      "grad_norm": 1.155529499053955,
+      "learning_rate": 2e-05,
+      "loss": 0.03705712,
+      "step": 10650
+    },
+    {
+      "epoch": 21.302,
+      "grad_norm": 1.44944167137146,
+      "learning_rate": 2e-05,
+      "loss": 0.04845714,
+      "step": 10651
+    },
+    {
+      "epoch": 21.304,
+      "grad_norm": 1.276952862739563,
+      "learning_rate": 2e-05,
+      "loss": 0.04431798,
+      "step": 10652
+    },
+    {
+      "epoch": 21.306,
+      "grad_norm": 1.0401476621627808,
+      "learning_rate": 2e-05,
+      "loss": 0.02794194,
+      "step": 10653
+    },
+    {
+      "epoch": 21.308,
+      "grad_norm": 1.4874932765960693,
+      "learning_rate": 2e-05,
+      "loss": 0.04844416,
+      "step": 10654
+    },
+    {
+      "epoch": 21.31,
+      "grad_norm": 1.2578763961791992,
+      "learning_rate": 2e-05,
+      "loss": 0.02141695,
+      "step": 10655
+    },
+    {
+      "epoch": 21.312,
+      "grad_norm": 1.0979886054992676,
+      "learning_rate": 2e-05,
+      "loss": 0.03747806,
+      "step": 10656
+    },
+    {
+      "epoch": 21.314,
+      "grad_norm": 1.5013819932937622,
+      "learning_rate": 2e-05,
+      "loss": 0.03525694,
+      "step": 10657
+    },
+    {
+      "epoch": 21.316,
+      "grad_norm": 2.042410373687744,
+      "learning_rate": 2e-05,
+      "loss": 0.04237705,
+      "step": 10658
+    },
+    {
+      "epoch": 21.318,
+      "grad_norm": 0.8237473964691162,
+      "learning_rate": 2e-05,
+      "loss": 0.02557302,
+      "step": 10659
+    },
+    {
+      "epoch": 21.32,
+      "grad_norm": 1.2069071531295776,
+      "learning_rate": 2e-05,
+      "loss": 0.03492039,
+      "step": 10660
+    },
+    {
+      "epoch": 21.322,
+      "grad_norm": 2.445582866668701,
+      "learning_rate": 2e-05,
+      "loss": 0.04500083,
+      "step": 10661
+    },
+    {
+      "epoch": 21.324,
+      "grad_norm": 1.7285526990890503,
+      "learning_rate": 2e-05,
+      "loss": 0.0372141,
+      "step": 10662
+    },
+    {
+      "epoch": 21.326,
+      "grad_norm": 1.2738267183303833,
+      "learning_rate": 2e-05,
+      "loss": 0.04208554,
+      "step": 10663
+    },
+    {
+      "epoch": 21.328,
+      "grad_norm": 1.3018332719802856,
+      "learning_rate": 2e-05,
+      "loss": 0.02245201,
+      "step": 10664
+    },
+    {
+      "epoch": 21.33,
+      "grad_norm": 1.3849666118621826,
+      "learning_rate": 2e-05,
+      "loss": 0.04083522,
+      "step": 10665
+    },
+    {
+      "epoch": 21.332,
+      "grad_norm": 1.0650460720062256,
+      "learning_rate": 2e-05,
+      "loss": 0.03396599,
+      "step": 10666
+    },
+    {
+      "epoch": 21.334,
+      "grad_norm": 1.0239198207855225,
+      "learning_rate": 2e-05,
+      "loss": 0.02632653,
+      "step": 10667
+    },
+    {
+      "epoch": 21.336,
+      "grad_norm": 1.1510930061340332,
+      "learning_rate": 2e-05,
+      "loss": 0.03786356,
+      "step": 10668
+    },
+    {
+      "epoch": 21.338,
+      "grad_norm": 1.134979486465454,
+      "learning_rate": 2e-05,
+      "loss": 0.03500597,
+      "step": 10669
+    },
+    {
+      "epoch": 21.34,
+      "grad_norm": 1.7041795253753662,
+      "learning_rate": 2e-05,
+      "loss": 0.04310299,
+      "step": 10670
+    },
+    {
+      "epoch": 21.342,
+      "grad_norm": 1.080082893371582,
+      "learning_rate": 2e-05,
+      "loss": 0.02889851,
+      "step": 10671
+    },
+    {
+      "epoch": 21.344,
+      "grad_norm": 1.2668462991714478,
+      "learning_rate": 2e-05,
+      "loss": 0.06208187,
+      "step": 10672
+    },
+    {
+      "epoch": 21.346,
+      "grad_norm": 0.9316691160202026,
+      "learning_rate": 2e-05,
+      "loss": 0.02679442,
+      "step": 10673
+    },
+    {
+      "epoch": 21.348,
+      "grad_norm": 1.3193854093551636,
+      "learning_rate": 2e-05,
+      "loss": 0.03310707,
+      "step": 10674
+    },
+    {
+      "epoch": 21.35,
+      "grad_norm": 1.5377140045166016,
+      "learning_rate": 2e-05,
+      "loss": 0.03435023,
+      "step": 10675
+    },
+    {
+      "epoch": 21.352,
+      "grad_norm": 1.0346852540969849,
+      "learning_rate": 2e-05,
+      "loss": 0.02915202,
+      "step": 10676
+    },
+    {
+      "epoch": 21.354,
+      "grad_norm": 1.0699576139450073,
+      "learning_rate": 2e-05,
+      "loss": 0.03234834,
+      "step": 10677
+    },
+    {
+      "epoch": 21.356,
+      "grad_norm": 4.232156753540039,
+      "learning_rate": 2e-05,
+      "loss": 0.04847697,
+      "step": 10678
+    },
+    {
+      "epoch": 21.358,
+      "grad_norm": 0.8950043320655823,
+      "learning_rate": 2e-05,
+      "loss": 0.02786786,
+      "step": 10679
+    },
+    {
+      "epoch": 21.36,
+      "grad_norm": 1.1412599086761475,
+      "learning_rate": 2e-05,
+      "loss": 0.02532648,
+      "step": 10680
+    },
+    {
+      "epoch": 21.362,
+      "grad_norm": 1.9713459014892578,
+      "learning_rate": 2e-05,
+      "loss": 0.04944416,
+      "step": 10681
+    },
+    {
+      "epoch": 21.364,
+      "grad_norm": 1.2647299766540527,
+      "learning_rate": 2e-05,
+      "loss": 0.04223265,
+      "step": 10682
+    },
+    {
+      "epoch": 21.366,
+      "grad_norm": 2.940858840942383,
+      "learning_rate": 2e-05,
+      "loss": 0.03530517,
+      "step": 10683
+    },
+    {
+      "epoch": 21.368,
+      "grad_norm": 0.9534826874732971,
+      "learning_rate": 2e-05,
+      "loss": 0.02479281,
+      "step": 10684
+    },
+    {
+      "epoch": 21.37,
+      "grad_norm": 1.3853439092636108,
+      "learning_rate": 2e-05,
+      "loss": 0.03295173,
+      "step": 10685
+    },
+    {
+      "epoch": 21.372,
+      "grad_norm": 1.2453714609146118,
+      "learning_rate": 2e-05,
+      "loss": 0.02070951,
+      "step": 10686
+    },
+    {
+      "epoch": 21.374,
+      "grad_norm": 1.1567037105560303,
+      "learning_rate": 2e-05,
+      "loss": 0.03087223,
+      "step": 10687
+    },
+    {
+      "epoch": 21.376,
+      "grad_norm": 0.9362542629241943,
+      "learning_rate": 2e-05,
+      "loss": 0.01926921,
+      "step": 10688
+    },
+    {
+      "epoch": 21.378,
+      "grad_norm": 0.9018503427505493,
+      "learning_rate": 2e-05,
+      "loss": 0.02120415,
+      "step": 10689
+    },
+    {
+      "epoch": 21.38,
+      "grad_norm": 1.236549735069275,
+      "learning_rate": 2e-05,
+      "loss": 0.04043602,
+      "step": 10690
+    },
+    {
+      "epoch": 21.382,
+      "grad_norm": 1.1474121809005737,
+      "learning_rate": 2e-05,
+      "loss": 0.03565197,
+      "step": 10691
+    },
+    {
+      "epoch": 21.384,
+      "grad_norm": 1.0585695505142212,
+      "learning_rate": 2e-05,
+      "loss": 0.03018987,
+      "step": 10692
+    },
+    {
+      "epoch": 21.386,
+      "grad_norm": 1.0328543186187744,
+      "learning_rate": 2e-05,
+      "loss": 0.03397262,
+      "step": 10693
+    },
+    {
+      "epoch": 21.388,
+      "grad_norm": 1.446506381034851,
+      "learning_rate": 2e-05,
+      "loss": 0.02783697,
+      "step": 10694
+    },
+    {
+      "epoch": 21.39,
+      "grad_norm": 1.6255260705947876,
+      "learning_rate": 2e-05,
+      "loss": 0.03847382,
+      "step": 10695
+    },
+    {
+      "epoch": 21.392,
+      "grad_norm": 1.0737837553024292,
+      "learning_rate": 2e-05,
+      "loss": 0.02603565,
+      "step": 10696
+    },
+    {
+      "epoch": 21.394,
+      "grad_norm": 1.392006754875183,
+      "learning_rate": 2e-05,
+      "loss": 0.03012003,
+      "step": 10697
+    },
+    {
+      "epoch": 21.396,
+      "grad_norm": 2.5857255458831787,
+      "learning_rate": 2e-05,
+      "loss": 0.03107409,
+      "step": 10698
+    },
+    {
+      "epoch": 21.398,
+      "grad_norm": 1.2561590671539307,
+      "learning_rate": 2e-05,
+      "loss": 0.03176247,
+      "step": 10699
+    },
+    {
+      "epoch": 21.4,
+      "grad_norm": 1.0437184572219849,
+      "learning_rate": 2e-05,
+      "loss": 0.03001492,
+      "step": 10700
+    },
+    {
+      "epoch": 21.402,
+      "grad_norm": 1.9724184274673462,
+      "learning_rate": 2e-05,
+      "loss": 0.02830181,
+      "step": 10701
+    },
+    {
+      "epoch": 21.404,
+      "grad_norm": 1.181110143661499,
+      "learning_rate": 2e-05,
+      "loss": 0.0300355,
+      "step": 10702
+    },
+    {
+      "epoch": 21.406,
+      "grad_norm": 2.5362696647644043,
+      "learning_rate": 2e-05,
+      "loss": 0.04556922,
+      "step": 10703
+    },
+    {
+      "epoch": 21.408,
+      "grad_norm": 1.9228975772857666,
+      "learning_rate": 2e-05,
+      "loss": 0.0374077,
+      "step": 10704
+    },
+    {
+      "epoch": 21.41,
+      "grad_norm": 1.4400721788406372,
+      "learning_rate": 2e-05,
+      "loss": 0.0367013,
+      "step": 10705
+    },
+    {
+      "epoch": 21.412,
+      "grad_norm": 1.2807708978652954,
+      "learning_rate": 2e-05,
+      "loss": 0.045549,
+      "step": 10706
+    },
+    {
+      "epoch": 21.414,
+      "grad_norm": 2.0344345569610596,
+      "learning_rate": 2e-05,
+      "loss": 0.04431529,
+      "step": 10707
+    },
+    {
+      "epoch": 21.416,
+      "grad_norm": 3.9442005157470703,
+      "learning_rate": 2e-05,
+      "loss": 0.04201176,
+      "step": 10708
+    },
+    {
+      "epoch": 21.418,
+      "grad_norm": 0.9622934460639954,
+      "learning_rate": 2e-05,
+      "loss": 0.0284987,
+      "step": 10709
+    },
+    {
+      "epoch": 21.42,
+      "grad_norm": 1.0971020460128784,
+      "learning_rate": 2e-05,
+      "loss": 0.036263,
+      "step": 10710
+    },
+    {
+      "epoch": 21.422,
+      "grad_norm": 0.9987945556640625,
+      "learning_rate": 2e-05,
+      "loss": 0.02871657,
+      "step": 10711
+    },
+    {
+      "epoch": 21.424,
+      "grad_norm": 0.8595823645591736,
+      "learning_rate": 2e-05,
+      "loss": 0.02600442,
+      "step": 10712
+    },
+    {
+      "epoch": 21.426,
+      "grad_norm": 0.8434016108512878,
+      "learning_rate": 2e-05,
+      "loss": 0.02094415,
+      "step": 10713
+    },
+    {
+      "epoch": 21.428,
+      "grad_norm": 1.313782811164856,
+      "learning_rate": 2e-05,
+      "loss": 0.04309317,
+      "step": 10714
+    },
+    {
+      "epoch": 21.43,
+      "grad_norm": 1.0133094787597656,
+      "learning_rate": 2e-05,
+      "loss": 0.03012585,
+      "step": 10715
+    },
+    {
+      "epoch": 21.432,
+      "grad_norm": 1.157460331916809,
+      "learning_rate": 2e-05,
+      "loss": 0.03752935,
+      "step": 10716
+    },
+    {
+      "epoch": 21.434,
+      "grad_norm": 1.142626166343689,
+      "learning_rate": 2e-05,
+      "loss": 0.02614125,
+      "step": 10717
+    },
+    {
+      "epoch": 21.436,
+      "grad_norm": 1.3068962097167969,
+      "learning_rate": 2e-05,
+      "loss": 0.04685502,
+      "step": 10718
+    },
+    {
+      "epoch": 21.438,
+      "grad_norm": 1.4647998809814453,
+      "learning_rate": 2e-05,
+      "loss": 0.04809128,
+      "step": 10719
+    },
+    {
+      "epoch": 21.44,
+      "grad_norm": 0.9506215453147888,
+      "learning_rate": 2e-05,
+      "loss": 0.02731987,
+      "step": 10720
+    },
+    {
+      "epoch": 21.442,
+      "grad_norm": 1.687732458114624,
+      "learning_rate": 2e-05,
+      "loss": 0.0411804,
+      "step": 10721
+    },
+    {
+      "epoch": 21.444,
+      "grad_norm": 1.8168407678604126,
+      "learning_rate": 2e-05,
+      "loss": 0.04425385,
+      "step": 10722
+    },
+    {
+      "epoch": 21.446,
+      "grad_norm": 1.2137858867645264,
+      "learning_rate": 2e-05,
+      "loss": 0.03381842,
+      "step": 10723
+    },
+    {
+      "epoch": 21.448,
+      "grad_norm": 1.0868464708328247,
+      "learning_rate": 2e-05,
+      "loss": 0.0301272,
+      "step": 10724
+    },
+    {
+      "epoch": 21.45,
+      "grad_norm": 2.293076753616333,
+      "learning_rate": 2e-05,
+      "loss": 0.05272197,
+      "step": 10725
+    },
+    {
+      "epoch": 21.452,
+      "grad_norm": 1.2076009511947632,
+      "learning_rate": 2e-05,
+      "loss": 0.04387645,
+      "step": 10726
+    },
+    {
+      "epoch": 21.454,
+      "grad_norm": 1.757698655128479,
+      "learning_rate": 2e-05,
+      "loss": 0.02590645,
+      "step": 10727
+    },
+    {
+      "epoch": 21.456,
+      "grad_norm": 1.1657187938690186,
+      "learning_rate": 2e-05,
+      "loss": 0.03529991,
+      "step": 10728
+    },
+    {
+      "epoch": 21.458,
+      "grad_norm": 1.177914023399353,
+      "learning_rate": 2e-05,
+      "loss": 0.03198649,
+      "step": 10729
+    },
+    {
+      "epoch": 21.46,
+      "grad_norm": 2.1333088874816895,
+      "learning_rate": 2e-05,
+      "loss": 0.05930211,
+      "step": 10730
+    },
+    {
+      "epoch": 21.462,
+      "grad_norm": 1.0726081132888794,
+      "learning_rate": 2e-05,
+      "loss": 0.02519402,
+      "step": 10731
+    },
+    {
+      "epoch": 21.464,
+      "grad_norm": 1.8570704460144043,
+      "learning_rate": 2e-05,
+      "loss": 0.04541021,
+      "step": 10732
+    },
+    {
+      "epoch": 21.466,
+      "grad_norm": 1.4362432956695557,
+      "learning_rate": 2e-05,
+      "loss": 0.03639318,
+      "step": 10733
+    },
+    {
+      "epoch": 21.468,
+      "grad_norm": 1.3196659088134766,
+      "learning_rate": 2e-05,
+      "loss": 0.03501175,
+      "step": 10734
+    },
+    {
+      "epoch": 21.47,
+      "grad_norm": 1.788077712059021,
+      "learning_rate": 2e-05,
+      "loss": 0.04476755,
+      "step": 10735
+    },
+    {
+      "epoch": 21.472,
+      "grad_norm": 2.1154396533966064,
+      "learning_rate": 2e-05,
+      "loss": 0.04601542,
+      "step": 10736
+    },
+    {
+      "epoch": 21.474,
+      "grad_norm": 0.9764267206192017,
+      "learning_rate": 2e-05,
+      "loss": 0.03156991,
+      "step": 10737
+    },
+    {
+      "epoch": 21.476,
+      "grad_norm": 0.862126886844635,
+      "learning_rate": 2e-05,
+      "loss": 0.02627644,
+      "step": 10738
+    },
+    {
+      "epoch": 21.478,
+      "grad_norm": 2.1665496826171875,
+      "learning_rate": 2e-05,
+      "loss": 0.03621161,
+      "step": 10739
+    },
+    {
+      "epoch": 21.48,
+      "grad_norm": 0.9720172882080078,
+      "learning_rate": 2e-05,
+      "loss": 0.02424107,
+      "step": 10740
+    },
+    {
+      "epoch": 21.482,
+      "grad_norm": 1.1352647542953491,
+      "learning_rate": 2e-05,
+      "loss": 0.0395835,
+      "step": 10741
+    },
+    {
+      "epoch": 21.484,
+      "grad_norm": 1.2119773626327515,
+      "learning_rate": 2e-05,
+      "loss": 0.03099651,
+      "step": 10742
+    },
+    {
+      "epoch": 21.486,
+      "grad_norm": 1.091362476348877,
+      "learning_rate": 2e-05,
+      "loss": 0.03427457,
+      "step": 10743
+    },
+    {
+      "epoch": 21.488,
+      "grad_norm": 1.5212750434875488,
+      "learning_rate": 2e-05,
+      "loss": 0.03149851,
+      "step": 10744
+    },
+    {
+      "epoch": 21.49,
+      "grad_norm": 2.698265552520752,
+      "learning_rate": 2e-05,
+      "loss": 0.03275203,
+      "step": 10745
+    },
+    {
+      "epoch": 21.492,
+      "grad_norm": 0.9697499871253967,
+      "learning_rate": 2e-05,
+      "loss": 0.02803354,
+      "step": 10746
+    },
+    {
+      "epoch": 21.494,
+      "grad_norm": 1.4407355785369873,
+      "learning_rate": 2e-05,
+      "loss": 0.04102314,
+      "step": 10747
+    },
+    {
+      "epoch": 21.496,
+      "grad_norm": 0.9825798869132996,
+      "learning_rate": 2e-05,
+      "loss": 0.02902812,
+      "step": 10748
+    },
+    {
+      "epoch": 21.498,
+      "grad_norm": 0.9173822999000549,
+      "learning_rate": 2e-05,
+      "loss": 0.02705132,
+      "step": 10749
+    },
+    {
+      "epoch": 21.5,
+      "grad_norm": 1.5008710622787476,
+      "learning_rate": 2e-05,
+      "loss": 0.03890024,
+      "step": 10750
+    },
+    {
+      "epoch": 21.502,
+      "grad_norm": 2.323859214782715,
+      "learning_rate": 2e-05,
+      "loss": 0.03871877,
+      "step": 10751
+    },
+    {
+      "epoch": 21.504,
+      "grad_norm": 0.9519139528274536,
+      "learning_rate": 2e-05,
+      "loss": 0.02205279,
+      "step": 10752
+    },
+    {
+      "epoch": 21.506,
+      "grad_norm": 1.150467038154602,
+      "learning_rate": 2e-05,
+      "loss": 0.03497121,
+      "step": 10753
+    },
+    {
+      "epoch": 21.508,
+      "grad_norm": 1.5614557266235352,
+      "learning_rate": 2e-05,
+      "loss": 0.03725933,
+      "step": 10754
+    },
+    {
+      "epoch": 21.51,
+      "grad_norm": 0.8480432629585266,
+      "learning_rate": 2e-05,
+      "loss": 0.02463213,
+      "step": 10755
+    },
+    {
+      "epoch": 21.512,
+      "grad_norm": 1.0168507099151611,
+      "learning_rate": 2e-05,
+      "loss": 0.03348152,
+      "step": 10756
+    },
+    {
+      "epoch": 21.514,
+      "grad_norm": 1.9600995779037476,
+      "learning_rate": 2e-05,
+      "loss": 0.04116196,
+      "step": 10757
+    },
+    {
+      "epoch": 21.516,
+      "grad_norm": 2.4078519344329834,
+      "learning_rate": 2e-05,
+      "loss": 0.05193283,
+      "step": 10758
+    },
+    {
+      "epoch": 21.518,
+      "grad_norm": 1.9011753797531128,
+      "learning_rate": 2e-05,
+      "loss": 0.04391605,
+      "step": 10759
+    },
+    {
+      "epoch": 21.52,
+      "grad_norm": 1.0992064476013184,
+      "learning_rate": 2e-05,
+      "loss": 0.03505319,
+      "step": 10760
+    },
+    {
+      "epoch": 21.522,
+      "grad_norm": 1.1341238021850586,
+      "learning_rate": 2e-05,
+      "loss": 0.04327236,
+      "step": 10761
+    },
+    {
+      "epoch": 21.524,
+      "grad_norm": 0.8820840716362,
+      "learning_rate": 2e-05,
+      "loss": 0.02557,
+      "step": 10762
+    },
+    {
+      "epoch": 21.526,
+      "grad_norm": 1.7110241651535034,
+      "learning_rate": 2e-05,
+      "loss": 0.02929918,
+      "step": 10763
+    },
+    {
+      "epoch": 21.528,
+      "grad_norm": 2.175755262374878,
+      "learning_rate": 2e-05,
+      "loss": 0.04092379,
+      "step": 10764
+    },
+    {
+      "epoch": 21.53,
+      "grad_norm": 0.9964827299118042,
+      "learning_rate": 2e-05,
+      "loss": 0.02928205,
+      "step": 10765
+    },
+    {
+      "epoch": 21.532,
+      "grad_norm": 1.0591028928756714,
+      "learning_rate": 2e-05,
+      "loss": 0.03802252,
+      "step": 10766
+    },
+    {
+      "epoch": 21.534,
+      "grad_norm": 1.0037879943847656,
+      "learning_rate": 2e-05,
+      "loss": 0.02184133,
+      "step": 10767
+    },
+    {
+      "epoch": 21.536,
+      "grad_norm": 1.0294376611709595,
+      "learning_rate": 2e-05,
+      "loss": 0.03275803,
+      "step": 10768
+    },
+    {
+      "epoch": 21.538,
+      "grad_norm": 1.2927500009536743,
+      "learning_rate": 2e-05,
+      "loss": 0.05052555,
+      "step": 10769
+    },
+    {
+      "epoch": 21.54,
+      "grad_norm": 1.269002079963684,
+      "learning_rate": 2e-05,
+      "loss": 0.04041898,
+      "step": 10770
+    },
+    {
+      "epoch": 21.542,
+      "grad_norm": 1.3406201601028442,
+      "learning_rate": 2e-05,
+      "loss": 0.0340461,
+      "step": 10771
+    },
+    {
+      "epoch": 21.544,
+      "grad_norm": 1.2689399719238281,
+      "learning_rate": 2e-05,
+      "loss": 0.02636706,
+      "step": 10772
+    },
+    {
+      "epoch": 21.546,
+      "grad_norm": 1.7005540132522583,
+      "learning_rate": 2e-05,
+      "loss": 0.03229015,
+      "step": 10773
+    },
+    {
+      "epoch": 21.548000000000002,
+      "grad_norm": 1.281423568725586,
+      "learning_rate": 2e-05,
+      "loss": 0.03840701,
+      "step": 10774
+    },
+    {
+      "epoch": 21.55,
+      "grad_norm": 1.9321891069412231,
+      "learning_rate": 2e-05,
+      "loss": 0.03856434,
+      "step": 10775
+    },
+    {
+      "epoch": 21.552,
+      "grad_norm": 2.3012235164642334,
+      "learning_rate": 2e-05,
+      "loss": 0.04851568,
+      "step": 10776
+    },
+    {
+      "epoch": 21.554,
+      "grad_norm": 0.8743276000022888,
+      "learning_rate": 2e-05,
+      "loss": 0.0196914,
+      "step": 10777
+    },
+    {
+      "epoch": 21.556,
+      "grad_norm": 1.0524275302886963,
+      "learning_rate": 2e-05,
+      "loss": 0.038031,
+      "step": 10778
+    },
+    {
+      "epoch": 21.558,
+      "grad_norm": 1.160980463027954,
+      "learning_rate": 2e-05,
+      "loss": 0.04083832,
+      "step": 10779
+    },
+    {
+      "epoch": 21.56,
+      "grad_norm": 3.0021450519561768,
+      "learning_rate": 2e-05,
+      "loss": 0.03252212,
+      "step": 10780
+    },
+    {
+      "epoch": 21.562,
+      "grad_norm": 2.0682501792907715,
+      "learning_rate": 2e-05,
+      "loss": 0.03913051,
+      "step": 10781
+    },
+    {
+      "epoch": 21.564,
+      "grad_norm": 1.703480839729309,
+      "learning_rate": 2e-05,
+      "loss": 0.04668419,
+      "step": 10782
+    },
+    {
+      "epoch": 21.566,
+      "grad_norm": 1.4658693075180054,
+      "learning_rate": 2e-05,
+      "loss": 0.03443877,
+      "step": 10783
+    },
+    {
+      "epoch": 21.568,
+      "grad_norm": 1.1543418169021606,
+      "learning_rate": 2e-05,
+      "loss": 0.02747395,
+      "step": 10784
+    },
+    {
+      "epoch": 21.57,
+      "grad_norm": 3.2979228496551514,
+      "learning_rate": 2e-05,
+      "loss": 0.05042007,
+      "step": 10785
+    },
+    {
+      "epoch": 21.572,
+      "grad_norm": 1.8336671590805054,
+      "learning_rate": 2e-05,
+      "loss": 0.02918666,
+      "step": 10786
+    },
+    {
+      "epoch": 21.574,
+      "grad_norm": 1.3803110122680664,
+      "learning_rate": 2e-05,
+      "loss": 0.03612125,
+      "step": 10787
+    },
+    {
+      "epoch": 21.576,
+      "grad_norm": 1.217238426208496,
+      "learning_rate": 2e-05,
+      "loss": 0.03045898,
+      "step": 10788
+    },
+    {
+      "epoch": 21.578,
+      "grad_norm": 0.8724953532218933,
+      "learning_rate": 2e-05,
+      "loss": 0.02683663,
+      "step": 10789
+    },
+    {
+      "epoch": 21.58,
+      "grad_norm": 1.0224673748016357,
+      "learning_rate": 2e-05,
+      "loss": 0.03132909,
+      "step": 10790
+    },
+    {
+      "epoch": 21.582,
+      "grad_norm": 1.0978913307189941,
+      "learning_rate": 2e-05,
+      "loss": 0.03923561,
+      "step": 10791
+    },
+    {
+      "epoch": 21.584,
+      "grad_norm": 1.336134672164917,
+      "learning_rate": 2e-05,
+      "loss": 0.03779566,
+      "step": 10792
+    },
+    {
+      "epoch": 21.586,
+      "grad_norm": 2.5332863330841064,
+      "learning_rate": 2e-05,
+      "loss": 0.04706411,
+      "step": 10793
+    },
+    {
+      "epoch": 21.588,
+      "grad_norm": 0.9782661199569702,
+      "learning_rate": 2e-05,
+      "loss": 0.02816963,
+      "step": 10794
+    },
+    {
+      "epoch": 21.59,
+      "grad_norm": 0.963988184928894,
+      "learning_rate": 2e-05,
+      "loss": 0.03094437,
+      "step": 10795
+    },
+    {
+      "epoch": 21.592,
+      "grad_norm": 2.6186649799346924,
+      "learning_rate": 2e-05,
+      "loss": 0.02780627,
+      "step": 10796
+    },
+    {
+      "epoch": 21.594,
+      "grad_norm": 1.1740295886993408,
+      "learning_rate": 2e-05,
+      "loss": 0.03014258,
+      "step": 10797
+    },
+    {
+      "epoch": 21.596,
+      "grad_norm": 1.2682572603225708,
+      "learning_rate": 2e-05,
+      "loss": 0.03773146,
+      "step": 10798
+    },
+    {
+      "epoch": 21.598,
+      "grad_norm": 1.0841542482376099,
+      "learning_rate": 2e-05,
+      "loss": 0.02988249,
+      "step": 10799
+    },
+    {
+      "epoch": 21.6,
+      "grad_norm": 1.3508737087249756,
+      "learning_rate": 2e-05,
+      "loss": 0.0372295,
+      "step": 10800
+    },
+    {
+      "epoch": 21.602,
+      "grad_norm": 1.229504942893982,
+      "learning_rate": 2e-05,
+      "loss": 0.03721376,
+      "step": 10801
+    },
+    {
+      "epoch": 21.604,
+      "grad_norm": 1.0140235424041748,
+      "learning_rate": 2e-05,
+      "loss": 0.02006028,
+      "step": 10802
+    },
+    {
+      "epoch": 21.606,
+      "grad_norm": 2.93935489654541,
+      "learning_rate": 2e-05,
+      "loss": 0.04623206,
+      "step": 10803
+    },
+    {
+      "epoch": 21.608,
+      "grad_norm": 1.6117465496063232,
+      "learning_rate": 2e-05,
+      "loss": 0.02327195,
+      "step": 10804
+    },
+    {
+      "epoch": 21.61,
+      "grad_norm": 1.7175722122192383,
+      "learning_rate": 2e-05,
+      "loss": 0.03024486,
+      "step": 10805
+    },
+    {
+      "epoch": 21.612,
+      "grad_norm": 2.5179755687713623,
+      "learning_rate": 2e-05,
+      "loss": 0.03515901,
+      "step": 10806
+    },
+    {
+      "epoch": 21.614,
+      "grad_norm": 1.4241502285003662,
+      "learning_rate": 2e-05,
+      "loss": 0.03859766,
+      "step": 10807
+    },
+    {
+      "epoch": 21.616,
+      "grad_norm": 2.0011253356933594,
+      "learning_rate": 2e-05,
+      "loss": 0.02655087,
+      "step": 10808
+    },
+    {
+      "epoch": 21.618,
+      "grad_norm": 1.9893662929534912,
+      "learning_rate": 2e-05,
+      "loss": 0.03004796,
+      "step": 10809
+    },
+    {
+      "epoch": 21.62,
+      "grad_norm": 1.3463841676712036,
+      "learning_rate": 2e-05,
+      "loss": 0.02972662,
+      "step": 10810
+    },
+    {
+      "epoch": 21.622,
+      "grad_norm": 1.1787477731704712,
+      "learning_rate": 2e-05,
+      "loss": 0.0288261,
+      "step": 10811
+    },
+    {
+      "epoch": 21.624,
+      "grad_norm": 1.237833857536316,
+      "learning_rate": 2e-05,
+      "loss": 0.03204871,
+      "step": 10812
+    },
+    {
+      "epoch": 21.626,
+      "grad_norm": 3.5159149169921875,
+      "learning_rate": 2e-05,
+      "loss": 0.04790648,
+      "step": 10813
+    },
+    {
+      "epoch": 21.628,
+      "grad_norm": 0.845460057258606,
+      "learning_rate": 2e-05,
+      "loss": 0.02701693,
+      "step": 10814
+    },
+    {
+      "epoch": 21.63,
+      "grad_norm": 1.7467832565307617,
+      "learning_rate": 2e-05,
+      "loss": 0.02598082,
+      "step": 10815
+    },
+    {
+      "epoch": 21.632,
+      "grad_norm": 2.1213173866271973,
+      "learning_rate": 2e-05,
+      "loss": 0.04864348,
+      "step": 10816
+    },
+    {
+      "epoch": 21.634,
+      "grad_norm": 1.6620383262634277,
+      "learning_rate": 2e-05,
+      "loss": 0.05002079,
+      "step": 10817
+    },
+    {
+      "epoch": 21.636,
+      "grad_norm": 1.2031606435775757,
+      "learning_rate": 2e-05,
+      "loss": 0.02948435,
+      "step": 10818
+    },
+    {
+      "epoch": 21.638,
+      "grad_norm": 3.743027687072754,
+      "learning_rate": 2e-05,
+      "loss": 0.05274049,
+      "step": 10819
+    },
+    {
+      "epoch": 21.64,
+      "grad_norm": 2.317620038986206,
+      "learning_rate": 2e-05,
+      "loss": 0.03110951,
+      "step": 10820
+    },
+    {
+      "epoch": 21.642,
+      "grad_norm": 1.2688599824905396,
+      "learning_rate": 2e-05,
+      "loss": 0.03901623,
+      "step": 10821
+    },
+    {
+      "epoch": 21.644,
+      "grad_norm": 1.1904910802841187,
+      "learning_rate": 2e-05,
+      "loss": 0.03599124,
+      "step": 10822
+    },
+    {
+      "epoch": 21.646,
+      "grad_norm": 1.423413634300232,
+      "learning_rate": 2e-05,
+      "loss": 0.04643358,
+      "step": 10823
+    },
+    {
+      "epoch": 21.648,
+      "grad_norm": 1.1683127880096436,
+      "learning_rate": 2e-05,
+      "loss": 0.03427366,
+      "step": 10824
+    },
+    {
+      "epoch": 21.65,
+      "grad_norm": 0.9204855561256409,
+      "learning_rate": 2e-05,
+      "loss": 0.03122223,
+      "step": 10825
+    },
+    {
+      "epoch": 21.652,
+      "grad_norm": 1.1113882064819336,
+      "learning_rate": 2e-05,
+      "loss": 0.0381987,
+      "step": 10826
+    },
+    {
+      "epoch": 21.654,
+      "grad_norm": 1.2636103630065918,
+      "learning_rate": 2e-05,
+      "loss": 0.03268664,
+      "step": 10827
+    },
+    {
+      "epoch": 21.656,
+      "grad_norm": 0.9222037196159363,
+      "learning_rate": 2e-05,
+      "loss": 0.03147195,
+      "step": 10828
+    },
+    {
+      "epoch": 21.658,
+      "grad_norm": 1.5140608549118042,
+      "learning_rate": 2e-05,
+      "loss": 0.03292232,
+      "step": 10829
+    },
+    {
+      "epoch": 21.66,
+      "grad_norm": 1.063954472541809,
+      "learning_rate": 2e-05,
+      "loss": 0.03509793,
+      "step": 10830
+    },
+    {
+      "epoch": 21.662,
+      "grad_norm": 1.110836148262024,
+      "learning_rate": 2e-05,
+      "loss": 0.03393824,
+      "step": 10831
+    },
+    {
+      "epoch": 21.664,
+      "grad_norm": 0.9153269529342651,
+      "learning_rate": 2e-05,
+      "loss": 0.02850673,
+      "step": 10832
+    },
+    {
+      "epoch": 21.666,
+      "grad_norm": 1.3513257503509521,
+      "learning_rate": 2e-05,
+      "loss": 0.03400831,
+      "step": 10833
+    },
+    {
+      "epoch": 21.668,
+      "grad_norm": 0.828287661075592,
+      "learning_rate": 2e-05,
+      "loss": 0.02091564,
+      "step": 10834
+    },
+    {
+      "epoch": 21.67,
+      "grad_norm": 1.5664379596710205,
+      "learning_rate": 2e-05,
+      "loss": 0.03538562,
+      "step": 10835
+    },
+    {
+      "epoch": 21.672,
+      "grad_norm": 2.088440179824829,
+      "learning_rate": 2e-05,
+      "loss": 0.03902008,
+      "step": 10836
+    },
+    {
+      "epoch": 21.674,
+      "grad_norm": 0.9757646322250366,
+      "learning_rate": 2e-05,
+      "loss": 0.03331909,
+      "step": 10837
+    },
+    {
+      "epoch": 21.676,
+      "grad_norm": 1.3544541597366333,
+      "learning_rate": 2e-05,
+      "loss": 0.03964958,
+      "step": 10838
+    },
+    {
+      "epoch": 21.678,
+      "grad_norm": 1.8273234367370605,
+      "learning_rate": 2e-05,
+      "loss": 0.03560327,
+      "step": 10839
+    },
+    {
+      "epoch": 21.68,
+      "grad_norm": 0.8851908445358276,
+      "learning_rate": 2e-05,
+      "loss": 0.02724171,
+      "step": 10840
+    },
+    {
+      "epoch": 21.682,
+      "grad_norm": 1.3723584413528442,
+      "learning_rate": 2e-05,
+      "loss": 0.03260377,
+      "step": 10841
+    },
+    {
+      "epoch": 21.684,
+      "grad_norm": 1.7293577194213867,
+      "learning_rate": 2e-05,
+      "loss": 0.04087131,
+      "step": 10842
+    },
+    {
+      "epoch": 21.686,
+      "grad_norm": 1.4376215934753418,
+      "learning_rate": 2e-05,
+      "loss": 0.03094893,
+      "step": 10843
+    },
+    {
+      "epoch": 21.688,
+      "grad_norm": 2.080545663833618,
+      "learning_rate": 2e-05,
+      "loss": 0.04254334,
+      "step": 10844
+    },
+    {
+      "epoch": 21.69,
+      "grad_norm": 1.221475601196289,
+      "learning_rate": 2e-05,
+      "loss": 0.042444,
+      "step": 10845
+    },
+    {
+      "epoch": 21.692,
+      "grad_norm": 1.0191986560821533,
+      "learning_rate": 2e-05,
+      "loss": 0.03012396,
+      "step": 10846
+    },
+    {
+      "epoch": 21.694,
+      "grad_norm": 1.5199073553085327,
+      "learning_rate": 2e-05,
+      "loss": 0.04867528,
+      "step": 10847
+    },
+    {
+      "epoch": 21.696,
+      "grad_norm": 1.0538051128387451,
+      "learning_rate": 2e-05,
+      "loss": 0.02168122,
+      "step": 10848
+    },
+    {
+      "epoch": 21.698,
+      "grad_norm": 1.2353911399841309,
+      "learning_rate": 2e-05,
+      "loss": 0.04300802,
+      "step": 10849
+    },
+    {
+      "epoch": 21.7,
+      "grad_norm": 1.1052453517913818,
+      "learning_rate": 2e-05,
+      "loss": 0.01482047,
+      "step": 10850
+    },
+    {
+      "epoch": 21.701999999999998,
+      "grad_norm": 1.7022722959518433,
+      "learning_rate": 2e-05,
+      "loss": 0.04935133,
+      "step": 10851
+    },
+    {
+      "epoch": 21.704,
+      "grad_norm": 1.097477674484253,
+      "learning_rate": 2e-05,
+      "loss": 0.03660508,
+      "step": 10852
+    },
+    {
+      "epoch": 21.706,
+      "grad_norm": 2.1557564735412598,
+      "learning_rate": 2e-05,
+      "loss": 0.03016577,
+      "step": 10853
+    },
+    {
+      "epoch": 21.708,
+      "grad_norm": 2.0594379901885986,
+      "learning_rate": 2e-05,
+      "loss": 0.02772827,
+      "step": 10854
+    },
+    {
+      "epoch": 21.71,
+      "grad_norm": 1.2176257371902466,
+      "learning_rate": 2e-05,
+      "loss": 0.04331783,
+      "step": 10855
+    },
+    {
+      "epoch": 21.712,
+      "grad_norm": 1.0477436780929565,
+      "learning_rate": 2e-05,
+      "loss": 0.03022527,
+      "step": 10856
+    },
+    {
+      "epoch": 21.714,
+      "grad_norm": 0.903574526309967,
+      "learning_rate": 2e-05,
+      "loss": 0.02697276,
+      "step": 10857
+    },
+    {
+      "epoch": 21.716,
+      "grad_norm": 1.0498381853103638,
+      "learning_rate": 2e-05,
+      "loss": 0.03695409,
+      "step": 10858
+    },
+    {
+      "epoch": 21.718,
+      "grad_norm": 1.2936769723892212,
+      "learning_rate": 2e-05,
+      "loss": 0.02531984,
+      "step": 10859
+    },
+    {
+      "epoch": 21.72,
+      "grad_norm": 1.3947609663009644,
+      "learning_rate": 2e-05,
+      "loss": 0.04383749,
+      "step": 10860
+    },
+    {
+      "epoch": 21.722,
+      "grad_norm": 1.824804425239563,
+      "learning_rate": 2e-05,
+      "loss": 0.05057254,
+      "step": 10861
+    },
+    {
+      "epoch": 21.724,
+      "grad_norm": 0.9313336610794067,
+      "learning_rate": 2e-05,
+      "loss": 0.02528257,
+      "step": 10862
+    },
+    {
+      "epoch": 21.726,
+      "grad_norm": 1.0199240446090698,
+      "learning_rate": 2e-05,
+      "loss": 0.02904977,
+      "step": 10863
+    },
+    {
+      "epoch": 21.728,
+      "grad_norm": 1.1817435026168823,
+      "learning_rate": 2e-05,
+      "loss": 0.04280265,
+      "step": 10864
+    },
+    {
+      "epoch": 21.73,
+      "grad_norm": 1.3964251279830933,
+      "learning_rate": 2e-05,
+      "loss": 0.03219439,
+      "step": 10865
+    },
+    {
+      "epoch": 21.732,
+      "grad_norm": 1.4096508026123047,
+      "learning_rate": 2e-05,
+      "loss": 0.03554904,
+      "step": 10866
+    },
+    {
+      "epoch": 21.734,
+      "grad_norm": 1.8143075704574585,
+      "learning_rate": 2e-05,
+      "loss": 0.03615258,
+      "step": 10867
+    },
+    {
+      "epoch": 21.736,
+      "grad_norm": 0.7444595694541931,
+      "learning_rate": 2e-05,
+      "loss": 0.01764659,
+      "step": 10868
+    },
+    {
+      "epoch": 21.738,
+      "grad_norm": 1.0963743925094604,
+      "learning_rate": 2e-05,
+      "loss": 0.03289765,
+      "step": 10869
+    },
+    {
+      "epoch": 21.74,
+      "grad_norm": 1.7623260021209717,
+      "learning_rate": 2e-05,
+      "loss": 0.03723545,
+      "step": 10870
+    },
+    {
+      "epoch": 21.742,
+      "grad_norm": 1.765259861946106,
+      "learning_rate": 2e-05,
+      "loss": 0.03898823,
+      "step": 10871
+    },
+    {
+      "epoch": 21.744,
+      "grad_norm": 1.039548635482788,
+      "learning_rate": 2e-05,
+      "loss": 0.02444161,
+      "step": 10872
+    },
+    {
+      "epoch": 21.746,
+      "grad_norm": 1.2688932418823242,
+      "learning_rate": 2e-05,
+      "loss": 0.0317545,
+      "step": 10873
+    },
+    {
+      "epoch": 21.748,
+      "grad_norm": 1.3791906833648682,
+      "learning_rate": 2e-05,
+      "loss": 0.03312441,
+      "step": 10874
+    },
+    {
+      "epoch": 21.75,
+      "grad_norm": 0.9484188556671143,
+      "learning_rate": 2e-05,
+      "loss": 0.0266212,
+      "step": 10875
+    },
+    {
+      "epoch": 21.752,
+      "grad_norm": 0.8790581226348877,
+      "learning_rate": 2e-05,
+      "loss": 0.0297839,
+      "step": 10876
+    },
+    {
+      "epoch": 21.754,
+      "grad_norm": 1.943943738937378,
+      "learning_rate": 2e-05,
+      "loss": 0.04509932,
+      "step": 10877
+    },
+    {
+      "epoch": 21.756,
+      "grad_norm": 1.0038764476776123,
+      "learning_rate": 2e-05,
+      "loss": 0.03279506,
+      "step": 10878
+    },
+    {
+      "epoch": 21.758,
+      "grad_norm": 1.6507289409637451,
+      "learning_rate": 2e-05,
+      "loss": 0.04125448,
+      "step": 10879
+    },
+    {
+      "epoch": 21.76,
+      "grad_norm": 1.6259219646453857,
+      "learning_rate": 2e-05,
+      "loss": 0.04601936,
+      "step": 10880
+    },
+    {
+      "epoch": 21.762,
+      "grad_norm": 1.209537386894226,
+      "learning_rate": 2e-05,
+      "loss": 0.03919362,
+      "step": 10881
+    },
+    {
+      "epoch": 21.764,
+      "grad_norm": 1.0259144306182861,
+      "learning_rate": 2e-05,
+      "loss": 0.03256726,
+      "step": 10882
+    },
+    {
+      "epoch": 21.766,
+      "grad_norm": 1.5114238262176514,
+      "learning_rate": 2e-05,
+      "loss": 0.04673877,
+      "step": 10883
+    },
+    {
+      "epoch": 21.768,
+      "grad_norm": 1.0645627975463867,
+      "learning_rate": 2e-05,
+      "loss": 0.03577953,
+      "step": 10884
+    },
+    {
+      "epoch": 21.77,
+      "grad_norm": 2.1890196800231934,
+      "learning_rate": 2e-05,
+      "loss": 0.0552833,
+      "step": 10885
+    },
+    {
+      "epoch": 21.772,
+      "grad_norm": 1.5644853115081787,
+      "learning_rate": 2e-05,
+      "loss": 0.05350343,
+      "step": 10886
+    },
+    {
+      "epoch": 21.774,
+      "grad_norm": 0.8159758448600769,
+      "learning_rate": 2e-05,
+      "loss": 0.0221364,
+      "step": 10887
+    },
+    {
+      "epoch": 21.776,
+      "grad_norm": 1.310706615447998,
+      "learning_rate": 2e-05,
+      "loss": 0.02867308,
+      "step": 10888
+    },
+    {
+      "epoch": 21.778,
+      "grad_norm": 1.3240361213684082,
+      "learning_rate": 2e-05,
+      "loss": 0.04336674,
+      "step": 10889
+    },
+    {
+      "epoch": 21.78,
+      "grad_norm": 1.7382214069366455,
+      "learning_rate": 2e-05,
+      "loss": 0.04056187,
+      "step": 10890
+    },
+    {
+      "epoch": 21.782,
+      "grad_norm": 1.1201266050338745,
+      "learning_rate": 2e-05,
+      "loss": 0.0285767,
+      "step": 10891
+    },
+    {
+      "epoch": 21.784,
+      "grad_norm": 1.0340373516082764,
+      "learning_rate": 2e-05,
+      "loss": 0.0394372,
+      "step": 10892
+    },
+    {
+      "epoch": 21.786,
+      "grad_norm": 1.3631528615951538,
+      "learning_rate": 2e-05,
+      "loss": 0.04386802,
+      "step": 10893
+    },
+    {
+      "epoch": 21.788,
+      "grad_norm": 1.7087925672531128,
+      "learning_rate": 2e-05,
+      "loss": 0.05740018,
+      "step": 10894
+    },
+    {
+      "epoch": 21.79,
+      "grad_norm": 1.2946619987487793,
+      "learning_rate": 2e-05,
+      "loss": 0.04351215,
+      "step": 10895
+    },
+    {
+      "epoch": 21.792,
+      "grad_norm": 1.9575139284133911,
+      "learning_rate": 2e-05,
+      "loss": 0.04059194,
+      "step": 10896
+    },
+    {
+      "epoch": 21.794,
+      "grad_norm": 1.6252760887145996,
+      "learning_rate": 2e-05,
+      "loss": 0.03451142,
+      "step": 10897
+    },
+    {
+      "epoch": 21.796,
+      "grad_norm": 1.5331840515136719,
+      "learning_rate": 2e-05,
+      "loss": 0.03251875,
+      "step": 10898
+    },
+    {
+      "epoch": 21.798000000000002,
+      "grad_norm": 1.0864592790603638,
+      "learning_rate": 2e-05,
+      "loss": 0.03452892,
+      "step": 10899
+    },
+    {
+      "epoch": 21.8,
+      "grad_norm": 0.8250343799591064,
+      "learning_rate": 2e-05,
+      "loss": 0.02099156,
+      "step": 10900
+    },
+    {
+      "epoch": 21.802,
+      "grad_norm": 1.166359782218933,
+      "learning_rate": 2e-05,
+      "loss": 0.04394403,
+      "step": 10901
+    },
+    {
+      "epoch": 21.804,
+      "grad_norm": 1.5257127285003662,
+      "learning_rate": 2e-05,
+      "loss": 0.03211277,
+      "step": 10902
+    },
+    {
+      "epoch": 21.806,
+      "grad_norm": 1.0145310163497925,
+      "learning_rate": 2e-05,
+      "loss": 0.02621819,
+      "step": 10903
+    },
+    {
+      "epoch": 21.808,
+      "grad_norm": 2.11439847946167,
+      "learning_rate": 2e-05,
+      "loss": 0.04509841,
+      "step": 10904
+    },
+    {
+      "epoch": 21.81,
+      "grad_norm": 1.0672552585601807,
+      "learning_rate": 2e-05,
+      "loss": 0.03072799,
+      "step": 10905
+    },
+    {
+      "epoch": 21.812,
+      "grad_norm": 1.1568739414215088,
+      "learning_rate": 2e-05,
+      "loss": 0.04023495,
+      "step": 10906
+    },
+    {
+      "epoch": 21.814,
+      "grad_norm": 1.9787969589233398,
+      "learning_rate": 2e-05,
+      "loss": 0.04461158,
+      "step": 10907
+    },
+    {
+      "epoch": 21.816,
+      "grad_norm": 1.268618106842041,
+      "learning_rate": 2e-05,
+      "loss": 0.04039661,
+      "step": 10908
+    },
+    {
+      "epoch": 21.818,
+      "grad_norm": 1.030121088027954,
+      "learning_rate": 2e-05,
+      "loss": 0.03256563,
+      "step": 10909
+    },
+    {
+      "epoch": 21.82,
+      "grad_norm": 1.3051402568817139,
+      "learning_rate": 2e-05,
+      "loss": 0.03315262,
+      "step": 10910
+    },
+    {
+      "epoch": 21.822,
+      "grad_norm": 1.599260687828064,
+      "learning_rate": 2e-05,
+      "loss": 0.05178364,
+      "step": 10911
+    },
+    {
+      "epoch": 21.824,
+      "grad_norm": 1.3491737842559814,
+      "learning_rate": 2e-05,
+      "loss": 0.03285315,
+      "step": 10912
+    },
+    {
+      "epoch": 21.826,
+      "grad_norm": 0.9857081770896912,
+      "learning_rate": 2e-05,
+      "loss": 0.02820598,
+      "step": 10913
+    },
+    {
+      "epoch": 21.828,
+      "grad_norm": 1.2068244218826294,
+      "learning_rate": 2e-05,
+      "loss": 0.02220411,
+      "step": 10914
+    },
+    {
+      "epoch": 21.83,
+      "grad_norm": 1.127752661705017,
+      "learning_rate": 2e-05,
+      "loss": 0.03646821,
+      "step": 10915
+    },
+    {
+      "epoch": 21.832,
+      "grad_norm": 1.1618894338607788,
+      "learning_rate": 2e-05,
+      "loss": 0.03601151,
+      "step": 10916
+    },
+    {
+      "epoch": 21.834,
+      "grad_norm": 2.112398624420166,
+      "learning_rate": 2e-05,
+      "loss": 0.03499029,
+      "step": 10917
+    },
+    {
+      "epoch": 21.836,
+      "grad_norm": 1.555198073387146,
+      "learning_rate": 2e-05,
+      "loss": 0.04573829,
+      "step": 10918
+    },
+    {
+      "epoch": 21.838,
+      "grad_norm": 0.9329383373260498,
+      "learning_rate": 2e-05,
+      "loss": 0.03361698,
+      "step": 10919
+    },
+    {
+      "epoch": 21.84,
+      "grad_norm": 1.7770673036575317,
+      "learning_rate": 2e-05,
+      "loss": 0.03091755,
+      "step": 10920
+    },
+    {
+      "epoch": 21.842,
+      "grad_norm": 1.157621145248413,
+      "learning_rate": 2e-05,
+      "loss": 0.03168878,
+      "step": 10921
+    },
+    {
+      "epoch": 21.844,
+      "grad_norm": 1.2026602029800415,
+      "learning_rate": 2e-05,
+      "loss": 0.02261108,
+      "step": 10922
+    },
+    {
+      "epoch": 21.846,
+      "grad_norm": 1.4362621307373047,
+      "learning_rate": 2e-05,
+      "loss": 0.02808312,
+      "step": 10923
+    },
+    {
+      "epoch": 21.848,
+      "grad_norm": 1.054166555404663,
+      "learning_rate": 2e-05,
+      "loss": 0.03865466,
+      "step": 10924
+    },
+    {
+      "epoch": 21.85,
+      "grad_norm": 1.0605528354644775,
+      "learning_rate": 2e-05,
+      "loss": 0.03541258,
+      "step": 10925
+    },
+    {
+      "epoch": 21.852,
+      "grad_norm": 1.4870340824127197,
+      "learning_rate": 2e-05,
+      "loss": 0.04259035,
+      "step": 10926
+    },
+    {
+      "epoch": 21.854,
+      "grad_norm": 1.6365232467651367,
+      "learning_rate": 2e-05,
+      "loss": 0.03734278,
+      "step": 10927
+    },
+    {
+      "epoch": 21.856,
+      "grad_norm": 0.7869289517402649,
+      "learning_rate": 2e-05,
+      "loss": 0.02219597,
+      "step": 10928
+    },
+    {
+      "epoch": 21.858,
+      "grad_norm": 1.1174979209899902,
+      "learning_rate": 2e-05,
+      "loss": 0.03317832,
+      "step": 10929
+    },
+    {
+      "epoch": 21.86,
+      "grad_norm": 1.7156922817230225,
+      "learning_rate": 2e-05,
+      "loss": 0.03578085,
+      "step": 10930
+    },
+    {
+      "epoch": 21.862,
+      "grad_norm": 1.5693000555038452,
+      "learning_rate": 2e-05,
+      "loss": 0.03383234,
+      "step": 10931
+    },
+    {
+      "epoch": 21.864,
+      "grad_norm": 2.124469041824341,
+      "learning_rate": 2e-05,
+      "loss": 0.029794,
+      "step": 10932
+    },
+    {
+      "epoch": 21.866,
+      "grad_norm": 1.151176929473877,
+      "learning_rate": 2e-05,
+      "loss": 0.04017279,
+      "step": 10933
+    },
+    {
+      "epoch": 21.868,
+      "grad_norm": 0.8996794819831848,
+      "learning_rate": 2e-05,
+      "loss": 0.02869819,
+      "step": 10934
+    },
+    {
+      "epoch": 21.87,
+      "grad_norm": 0.9696109294891357,
+      "learning_rate": 2e-05,
+      "loss": 0.032364,
+      "step": 10935
+    },
+    {
+      "epoch": 21.872,
+      "grad_norm": 1.4245364665985107,
+      "learning_rate": 2e-05,
+      "loss": 0.03265813,
+      "step": 10936
+    },
+    {
+      "epoch": 21.874,
+      "grad_norm": 1.5670595169067383,
+      "learning_rate": 2e-05,
+      "loss": 0.04186102,
+      "step": 10937
+    },
+    {
+      "epoch": 21.876,
+      "grad_norm": 1.4556249380111694,
+      "learning_rate": 2e-05,
+      "loss": 0.04338999,
+      "step": 10938
+    },
+    {
+      "epoch": 21.878,
+      "grad_norm": 1.3597520589828491,
+      "learning_rate": 2e-05,
+      "loss": 0.03609984,
+      "step": 10939
+    },
+    {
+      "epoch": 21.88,
+      "grad_norm": 1.7528231143951416,
+      "learning_rate": 2e-05,
+      "loss": 0.03558186,
+      "step": 10940
+    },
+    {
+      "epoch": 21.882,
+      "grad_norm": 1.0914018154144287,
+      "learning_rate": 2e-05,
+      "loss": 0.0334373,
+      "step": 10941
+    },
+    {
+      "epoch": 21.884,
+      "grad_norm": 1.0560919046401978,
+      "learning_rate": 2e-05,
+      "loss": 0.02701947,
+      "step": 10942
+    },
+    {
+      "epoch": 21.886,
+      "grad_norm": 1.0461790561676025,
+      "learning_rate": 2e-05,
+      "loss": 0.02477691,
+      "step": 10943
+    },
+    {
+      "epoch": 21.888,
+      "grad_norm": 1.0176409482955933,
+      "learning_rate": 2e-05,
+      "loss": 0.02632477,
+      "step": 10944
+    },
+    {
+      "epoch": 21.89,
+      "grad_norm": 1.188032627105713,
+      "learning_rate": 2e-05,
+      "loss": 0.03136832,
+      "step": 10945
+    },
+    {
+      "epoch": 21.892,
+      "grad_norm": 1.8192259073257446,
+      "learning_rate": 2e-05,
+      "loss": 0.03940712,
+      "step": 10946
+    },
+    {
+      "epoch": 21.894,
+      "grad_norm": 1.1689777374267578,
+      "learning_rate": 2e-05,
+      "loss": 0.04128121,
+      "step": 10947
+    },
+    {
+      "epoch": 21.896,
+      "grad_norm": 0.9614786505699158,
+      "learning_rate": 2e-05,
+      "loss": 0.03097624,
+      "step": 10948
+    },
+    {
+      "epoch": 21.898,
+      "grad_norm": 2.1277706623077393,
+      "learning_rate": 2e-05,
+      "loss": 0.04416885,
+      "step": 10949
+    },
+    {
+      "epoch": 21.9,
+      "grad_norm": 1.0847748517990112,
+      "learning_rate": 2e-05,
+      "loss": 0.047647,
+      "step": 10950
+    },
+    {
+      "epoch": 21.902,
+      "grad_norm": 1.4761464595794678,
+      "learning_rate": 2e-05,
+      "loss": 0.03700073,
+      "step": 10951
+    },
+    {
+      "epoch": 21.904,
+      "grad_norm": 1.335006833076477,
+      "learning_rate": 2e-05,
+      "loss": 0.02501204,
+      "step": 10952
+    },
+    {
+      "epoch": 21.906,
+      "grad_norm": 1.1310573816299438,
+      "learning_rate": 2e-05,
+      "loss": 0.03214348,
+      "step": 10953
+    },
+    {
+      "epoch": 21.908,
+      "grad_norm": 2.758718490600586,
+      "learning_rate": 2e-05,
+      "loss": 0.0428556,
+      "step": 10954
+    },
+    {
+      "epoch": 21.91,
+      "grad_norm": 1.922041654586792,
+      "learning_rate": 2e-05,
+      "loss": 0.04771857,
+      "step": 10955
+    },
+    {
+      "epoch": 21.912,
+      "grad_norm": 1.7065603733062744,
+      "learning_rate": 2e-05,
+      "loss": 0.0471417,
+      "step": 10956
+    },
+    {
+      "epoch": 21.914,
+      "grad_norm": 1.5573147535324097,
+      "learning_rate": 2e-05,
+      "loss": 0.03855333,
+      "step": 10957
+    },
+    {
+      "epoch": 21.916,
+      "grad_norm": 1.1030230522155762,
+      "learning_rate": 2e-05,
+      "loss": 0.03686146,
+      "step": 10958
+    },
+    {
+      "epoch": 21.918,
+      "grad_norm": 1.0365067720413208,
+      "learning_rate": 2e-05,
+      "loss": 0.02389233,
+      "step": 10959
+    },
+    {
+      "epoch": 21.92,
+      "grad_norm": 1.0167555809020996,
+      "learning_rate": 2e-05,
+      "loss": 0.03511351,
+      "step": 10960
+    },
+    {
+      "epoch": 21.922,
+      "grad_norm": 1.0623586177825928,
+      "learning_rate": 2e-05,
+      "loss": 0.0339056,
+      "step": 10961
+    },
+    {
+      "epoch": 21.924,
+      "grad_norm": 2.31209397315979,
+      "learning_rate": 2e-05,
+      "loss": 0.04047135,
+      "step": 10962
+    },
+    {
+      "epoch": 21.926,
+      "grad_norm": 1.1770862340927124,
+      "learning_rate": 2e-05,
+      "loss": 0.04121822,
+      "step": 10963
+    },
+    {
+      "epoch": 21.928,
+      "grad_norm": 0.8528270125389099,
+      "learning_rate": 2e-05,
+      "loss": 0.02435702,
+      "step": 10964
+    },
+    {
+      "epoch": 21.93,
+      "grad_norm": 1.0807077884674072,
+      "learning_rate": 2e-05,
+      "loss": 0.03478127,
+      "step": 10965
+    },
+    {
+      "epoch": 21.932,
+      "grad_norm": 1.1846781969070435,
+      "learning_rate": 2e-05,
+      "loss": 0.04354676,
+      "step": 10966
+    },
+    {
+      "epoch": 21.934,
+      "grad_norm": 1.334719181060791,
+      "learning_rate": 2e-05,
+      "loss": 0.02577735,
+      "step": 10967
+    },
+    {
+      "epoch": 21.936,
+      "grad_norm": 1.0082858800888062,
+      "learning_rate": 2e-05,
+      "loss": 0.0299839,
+      "step": 10968
+    },
+    {
+      "epoch": 21.938,
+      "grad_norm": 1.5717273950576782,
+      "learning_rate": 2e-05,
+      "loss": 0.02820227,
+      "step": 10969
+    },
+    {
+      "epoch": 21.94,
+      "grad_norm": 1.0491491556167603,
+      "learning_rate": 2e-05,
+      "loss": 0.02673104,
+      "step": 10970
+    },
+    {
+      "epoch": 21.942,
+      "grad_norm": 1.0635792016983032,
+      "learning_rate": 2e-05,
+      "loss": 0.02951121,
+      "step": 10971
+    },
+    {
+      "epoch": 21.944,
+      "grad_norm": 1.3966928720474243,
+      "learning_rate": 2e-05,
+      "loss": 0.04850317,
+      "step": 10972
+    },
+    {
+      "epoch": 21.946,
+      "grad_norm": 1.1007291078567505,
+      "learning_rate": 2e-05,
+      "loss": 0.02945408,
+      "step": 10973
+    },
+    {
+      "epoch": 21.948,
+      "grad_norm": 0.8906104564666748,
+      "learning_rate": 2e-05,
+      "loss": 0.03067725,
+      "step": 10974
+    },
+    {
+      "epoch": 21.95,
+      "grad_norm": 1.0863364934921265,
+      "learning_rate": 2e-05,
+      "loss": 0.02871832,
+      "step": 10975
+    },
+    {
+      "epoch": 21.951999999999998,
+      "grad_norm": 1.6591525077819824,
+      "learning_rate": 2e-05,
+      "loss": 0.04460109,
+      "step": 10976
+    },
+    {
+      "epoch": 21.954,
+      "grad_norm": 1.431624174118042,
+      "learning_rate": 2e-05,
+      "loss": 0.03387917,
+      "step": 10977
+    },
+    {
+      "epoch": 21.956,
+      "grad_norm": 1.215907335281372,
+      "learning_rate": 2e-05,
+      "loss": 0.05533048,
+      "step": 10978
+    },
+    {
+      "epoch": 21.958,
+      "grad_norm": 1.0936074256896973,
+      "learning_rate": 2e-05,
+      "loss": 0.02711234,
+      "step": 10979
+    },
+    {
+      "epoch": 21.96,
+      "grad_norm": 1.123211145401001,
+      "learning_rate": 2e-05,
+      "loss": 0.04708305,
+      "step": 10980
+    },
+    {
+      "epoch": 21.962,
+      "grad_norm": 1.28691828250885,
+      "learning_rate": 2e-05,
+      "loss": 0.03268384,
+      "step": 10981
+    },
+    {
+      "epoch": 21.964,
+      "grad_norm": 1.013840913772583,
+      "learning_rate": 2e-05,
+      "loss": 0.03659943,
+      "step": 10982
+    },
+    {
+      "epoch": 21.966,
+      "grad_norm": 1.2497045993804932,
+      "learning_rate": 2e-05,
+      "loss": 0.05112303,
+      "step": 10983
+    },
+    {
+      "epoch": 21.968,
+      "grad_norm": 0.9134369492530823,
+      "learning_rate": 2e-05,
+      "loss": 0.02654456,
+      "step": 10984
+    },
+    {
+      "epoch": 21.97,
+      "grad_norm": 1.2738003730773926,
+      "learning_rate": 2e-05,
+      "loss": 0.03234825,
+      "step": 10985
+    },
+    {
+      "epoch": 21.972,
+      "grad_norm": 0.8131256699562073,
+      "learning_rate": 2e-05,
+      "loss": 0.0200008,
+      "step": 10986
+    },
+    {
+      "epoch": 21.974,
+      "grad_norm": 1.4570480585098267,
+      "learning_rate": 2e-05,
+      "loss": 0.02533815,
+      "step": 10987
+    },
+    {
+      "epoch": 21.976,
+      "grad_norm": 0.9933975338935852,
+      "learning_rate": 2e-05,
+      "loss": 0.03682224,
+      "step": 10988
+    },
+    {
+      "epoch": 21.978,
+      "grad_norm": 2.5660200119018555,
+      "learning_rate": 2e-05,
+      "loss": 0.04893076,
+      "step": 10989
+    },
+    {
+      "epoch": 21.98,
+      "grad_norm": 1.247011661529541,
+      "learning_rate": 2e-05,
+      "loss": 0.03728838,
+      "step": 10990
+    },
+    {
+      "epoch": 21.982,
+      "grad_norm": 2.838491201400757,
+      "learning_rate": 2e-05,
+      "loss": 0.04673835,
+      "step": 10991
+    },
+    {
+      "epoch": 21.984,
+      "grad_norm": 1.4233250617980957,
+      "learning_rate": 2e-05,
+      "loss": 0.04645887,
+      "step": 10992
+    },
+    {
+      "epoch": 21.986,
+      "grad_norm": 1.8115314245224,
+      "learning_rate": 2e-05,
+      "loss": 0.03267002,
+      "step": 10993
+    },
+    {
+      "epoch": 21.988,
+      "grad_norm": 0.8721919655799866,
+      "learning_rate": 2e-05,
+      "loss": 0.03281524,
+      "step": 10994
+    },
+    {
+      "epoch": 21.99,
+      "grad_norm": 1.70564866065979,
+      "learning_rate": 2e-05,
+      "loss": 0.04003553,
+      "step": 10995
+    },
+    {
+      "epoch": 21.992,
+      "grad_norm": 0.9615834951400757,
+      "learning_rate": 2e-05,
+      "loss": 0.03332902,
+      "step": 10996
+    },
+    {
+      "epoch": 21.994,
+      "grad_norm": 1.3987194299697876,
+      "learning_rate": 2e-05,
+      "loss": 0.03447945,
+      "step": 10997
+    },
+    {
+      "epoch": 21.996,
+      "grad_norm": 0.9507851600646973,
+      "learning_rate": 2e-05,
+      "loss": 0.02853581,
+      "step": 10998
+    },
+    {
+      "epoch": 21.998,
+      "grad_norm": 0.9905051589012146,
+      "learning_rate": 2e-05,
+      "loss": 0.02947031,
+      "step": 10999
+    },
+    {
+      "epoch": 22.0,
+      "grad_norm": 1.9134247303009033,
+      "learning_rate": 2e-05,
+      "loss": 0.03030916,
+      "step": 11000
+    },
+    {
+      "epoch": 22.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9760479041916168,
+        "Equal_1": 0.998,
+        "Equal_2": 0.9800399201596807,
+        "Equal_3": 0.9740518962075848,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9960079840319361,
+        "Parallel_1": 0.9919839679358717,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 1.0,
+        "Perpendicular_2": 0.988,
+        "Perpendicular_3": 0.8296593186372746,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9956666666666667,
+        "PointLiesOnCircle_3": 0.988,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9939879759519038,
+        "PointLiesOnLine_3": 0.9800399201596807
+      },
+      "eval_runtime": 319.6303,
+      "eval_samples_per_second": 32.85,
+      "eval_steps_per_second": 0.657,
+      "step": 11000
+    },
+    {
+      "epoch": 22.002,
+      "grad_norm": 1.5013846158981323,
+      "learning_rate": 2e-05,
+      "loss": 0.04865558,
+      "step": 11001
+    },
+    {
+      "epoch": 22.004,
+      "grad_norm": 0.9657471179962158,
+      "learning_rate": 2e-05,
+      "loss": 0.03410444,
+      "step": 11002
+    },
+    {
+      "epoch": 22.006,
+      "grad_norm": 1.055569052696228,
+      "learning_rate": 2e-05,
+      "loss": 0.03405615,
+      "step": 11003
+    },
+    {
+      "epoch": 22.008,
+      "grad_norm": 1.843996524810791,
+      "learning_rate": 2e-05,
+      "loss": 0.05212854,
+      "step": 11004
+    },
+    {
+      "epoch": 22.01,
+      "grad_norm": 1.2965750694274902,
+      "learning_rate": 2e-05,
+      "loss": 0.04010913,
+      "step": 11005
+    },
+    {
+      "epoch": 22.012,
+      "grad_norm": 0.9425175189971924,
+      "learning_rate": 2e-05,
+      "loss": 0.02818494,
+      "step": 11006
+    },
+    {
+      "epoch": 22.014,
+      "grad_norm": 1.5576436519622803,
+      "learning_rate": 2e-05,
+      "loss": 0.05102572,
+      "step": 11007
+    },
+    {
+      "epoch": 22.016,
+      "grad_norm": 0.9978459477424622,
+      "learning_rate": 2e-05,
+      "loss": 0.03394839,
+      "step": 11008
+    },
+    {
+      "epoch": 22.018,
+      "grad_norm": 1.1211631298065186,
+      "learning_rate": 2e-05,
+      "loss": 0.04935118,
+      "step": 11009
+    },
+    {
+      "epoch": 22.02,
+      "grad_norm": 2.0036747455596924,
+      "learning_rate": 2e-05,
+      "loss": 0.04285323,
+      "step": 11010
+    },
+    {
+      "epoch": 22.022,
+      "grad_norm": 1.1582741737365723,
+      "learning_rate": 2e-05,
+      "loss": 0.04079919,
+      "step": 11011
+    },
+    {
+      "epoch": 22.024,
+      "grad_norm": 1.5606409311294556,
+      "learning_rate": 2e-05,
+      "loss": 0.03285683,
+      "step": 11012
+    },
+    {
+      "epoch": 22.026,
+      "grad_norm": 2.8261466026306152,
+      "learning_rate": 2e-05,
+      "loss": 0.04419506,
+      "step": 11013
+    },
+    {
+      "epoch": 22.028,
+      "grad_norm": 1.250126838684082,
+      "learning_rate": 2e-05,
+      "loss": 0.03550693,
+      "step": 11014
+    },
+    {
+      "epoch": 22.03,
+      "grad_norm": 2.362985372543335,
+      "learning_rate": 2e-05,
+      "loss": 0.03660453,
+      "step": 11015
+    },
+    {
+      "epoch": 22.032,
+      "grad_norm": 1.1419157981872559,
+      "learning_rate": 2e-05,
+      "loss": 0.03367894,
+      "step": 11016
+    },
+    {
+      "epoch": 22.034,
+      "grad_norm": 1.2348449230194092,
+      "learning_rate": 2e-05,
+      "loss": 0.03856073,
+      "step": 11017
+    },
+    {
+      "epoch": 22.036,
+      "grad_norm": 0.8628196120262146,
+      "learning_rate": 2e-05,
+      "loss": 0.02549962,
+      "step": 11018
+    },
+    {
+      "epoch": 22.038,
+      "grad_norm": 1.2292640209197998,
+      "learning_rate": 2e-05,
+      "loss": 0.03300656,
+      "step": 11019
+    },
+    {
+      "epoch": 22.04,
+      "grad_norm": 1.549729824066162,
+      "learning_rate": 2e-05,
+      "loss": 0.03245661,
+      "step": 11020
+    },
+    {
+      "epoch": 22.042,
+      "grad_norm": 1.387235164642334,
+      "learning_rate": 2e-05,
+      "loss": 0.03017138,
+      "step": 11021
+    },
+    {
+      "epoch": 22.044,
+      "grad_norm": 1.231819987297058,
+      "learning_rate": 2e-05,
+      "loss": 0.03833802,
+      "step": 11022
+    },
+    {
+      "epoch": 22.046,
+      "grad_norm": 0.9583532214164734,
+      "learning_rate": 2e-05,
+      "loss": 0.0348784,
+      "step": 11023
+    },
+    {
+      "epoch": 22.048,
+      "grad_norm": 2.0372843742370605,
+      "learning_rate": 2e-05,
+      "loss": 0.0497199,
+      "step": 11024
+    },
+    {
+      "epoch": 22.05,
+      "grad_norm": 1.5443356037139893,
+      "learning_rate": 2e-05,
+      "loss": 0.04319365,
+      "step": 11025
+    },
+    {
+      "epoch": 22.052,
+      "grad_norm": 0.9617211818695068,
+      "learning_rate": 2e-05,
+      "loss": 0.02863034,
+      "step": 11026
+    },
+    {
+      "epoch": 22.054,
+      "grad_norm": 1.008378028869629,
+      "learning_rate": 2e-05,
+      "loss": 0.03430276,
+      "step": 11027
+    },
+    {
+      "epoch": 22.056,
+      "grad_norm": 0.8539184927940369,
+      "learning_rate": 2e-05,
+      "loss": 0.02441457,
+      "step": 11028
+    },
+    {
+      "epoch": 22.058,
+      "grad_norm": 0.9425522685050964,
+      "learning_rate": 2e-05,
+      "loss": 0.03636635,
+      "step": 11029
+    },
+    {
+      "epoch": 22.06,
+      "grad_norm": 1.6154694557189941,
+      "learning_rate": 2e-05,
+      "loss": 0.03873938,
+      "step": 11030
+    },
+    {
+      "epoch": 22.062,
+      "grad_norm": 0.9321594834327698,
+      "learning_rate": 2e-05,
+      "loss": 0.02615494,
+      "step": 11031
+    },
+    {
+      "epoch": 22.064,
+      "grad_norm": 1.0639960765838623,
+      "learning_rate": 2e-05,
+      "loss": 0.03128612,
+      "step": 11032
+    },
+    {
+      "epoch": 22.066,
+      "grad_norm": 2.2421159744262695,
+      "learning_rate": 2e-05,
+      "loss": 0.04077518,
+      "step": 11033
+    },
+    {
+      "epoch": 22.068,
+      "grad_norm": 1.059031367301941,
+      "learning_rate": 2e-05,
+      "loss": 0.03726078,
+      "step": 11034
+    },
+    {
+      "epoch": 22.07,
+      "grad_norm": 1.1438366174697876,
+      "learning_rate": 2e-05,
+      "loss": 0.03756997,
+      "step": 11035
+    },
+    {
+      "epoch": 22.072,
+      "grad_norm": 1.1501072645187378,
+      "learning_rate": 2e-05,
+      "loss": 0.02969356,
+      "step": 11036
+    },
+    {
+      "epoch": 22.074,
+      "grad_norm": 1.339308738708496,
+      "learning_rate": 2e-05,
+      "loss": 0.02925887,
+      "step": 11037
+    },
+    {
+      "epoch": 22.076,
+      "grad_norm": 1.7447677850723267,
+      "learning_rate": 2e-05,
+      "loss": 0.04287693,
+      "step": 11038
+    },
+    {
+      "epoch": 22.078,
+      "grad_norm": 1.0596781969070435,
+      "learning_rate": 2e-05,
+      "loss": 0.02210912,
+      "step": 11039
+    },
+    {
+      "epoch": 22.08,
+      "grad_norm": 1.0472910404205322,
+      "learning_rate": 2e-05,
+      "loss": 0.03132582,
+      "step": 11040
+    },
+    {
+      "epoch": 22.082,
+      "grad_norm": 1.045034408569336,
+      "learning_rate": 2e-05,
+      "loss": 0.03018015,
+      "step": 11041
+    },
+    {
+      "epoch": 22.084,
+      "grad_norm": 0.9737959504127502,
+      "learning_rate": 2e-05,
+      "loss": 0.0249765,
+      "step": 11042
+    },
+    {
+      "epoch": 22.086,
+      "grad_norm": 1.3689839839935303,
+      "learning_rate": 2e-05,
+      "loss": 0.03519817,
+      "step": 11043
+    },
+    {
+      "epoch": 22.088,
+      "grad_norm": 1.2619421482086182,
+      "learning_rate": 2e-05,
+      "loss": 0.03652804,
+      "step": 11044
+    },
+    {
+      "epoch": 22.09,
+      "grad_norm": 1.0860421657562256,
+      "learning_rate": 2e-05,
+      "loss": 0.02551399,
+      "step": 11045
+    },
+    {
+      "epoch": 22.092,
+      "grad_norm": 0.9912066459655762,
+      "learning_rate": 2e-05,
+      "loss": 0.03179378,
+      "step": 11046
+    },
+    {
+      "epoch": 22.094,
+      "grad_norm": 1.0879795551300049,
+      "learning_rate": 2e-05,
+      "loss": 0.03158774,
+      "step": 11047
+    },
+    {
+      "epoch": 22.096,
+      "grad_norm": 0.8615609407424927,
+      "learning_rate": 2e-05,
+      "loss": 0.02629388,
+      "step": 11048
+    },
+    {
+      "epoch": 22.098,
+      "grad_norm": 1.7215999364852905,
+      "learning_rate": 2e-05,
+      "loss": 0.03771915,
+      "step": 11049
+    },
+    {
+      "epoch": 22.1,
+      "grad_norm": 0.9128824472427368,
+      "learning_rate": 2e-05,
+      "loss": 0.02178533,
+      "step": 11050
+    },
+    {
+      "epoch": 22.102,
+      "grad_norm": 1.2859452962875366,
+      "learning_rate": 2e-05,
+      "loss": 0.04280837,
+      "step": 11051
+    },
+    {
+      "epoch": 22.104,
+      "grad_norm": 1.0833191871643066,
+      "learning_rate": 2e-05,
+      "loss": 0.03400262,
+      "step": 11052
+    },
+    {
+      "epoch": 22.106,
+      "grad_norm": 0.8177392482757568,
+      "learning_rate": 2e-05,
+      "loss": 0.02097586,
+      "step": 11053
+    },
+    {
+      "epoch": 22.108,
+      "grad_norm": 3.141331195831299,
+      "learning_rate": 2e-05,
+      "loss": 0.02519385,
+      "step": 11054
+    },
+    {
+      "epoch": 22.11,
+      "grad_norm": 1.0663577318191528,
+      "learning_rate": 2e-05,
+      "loss": 0.03236621,
+      "step": 11055
+    },
+    {
+      "epoch": 22.112,
+      "grad_norm": 3.876330614089966,
+      "learning_rate": 2e-05,
+      "loss": 0.05157751,
+      "step": 11056
+    },
+    {
+      "epoch": 22.114,
+      "grad_norm": 1.0990570783615112,
+      "learning_rate": 2e-05,
+      "loss": 0.03127418,
+      "step": 11057
+    },
+    {
+      "epoch": 22.116,
+      "grad_norm": 2.1724092960357666,
+      "learning_rate": 2e-05,
+      "loss": 0.05966609,
+      "step": 11058
+    },
+    {
+      "epoch": 22.118,
+      "grad_norm": 2.2459487915039062,
+      "learning_rate": 2e-05,
+      "loss": 0.04710309,
+      "step": 11059
+    },
+    {
+      "epoch": 22.12,
+      "grad_norm": 1.9107338190078735,
+      "learning_rate": 2e-05,
+      "loss": 0.04102355,
+      "step": 11060
+    },
+    {
+      "epoch": 22.122,
+      "grad_norm": 1.542749285697937,
+      "learning_rate": 2e-05,
+      "loss": 0.04050046,
+      "step": 11061
+    },
+    {
+      "epoch": 22.124,
+      "grad_norm": 1.8505321741104126,
+      "learning_rate": 2e-05,
+      "loss": 0.04142891,
+      "step": 11062
+    },
+    {
+      "epoch": 22.126,
+      "grad_norm": 1.4331146478652954,
+      "learning_rate": 2e-05,
+      "loss": 0.03895901,
+      "step": 11063
+    },
+    {
+      "epoch": 22.128,
+      "grad_norm": 1.0269427299499512,
+      "learning_rate": 2e-05,
+      "loss": 0.03234866,
+      "step": 11064
+    },
+    {
+      "epoch": 22.13,
+      "grad_norm": 1.1322760581970215,
+      "learning_rate": 2e-05,
+      "loss": 0.03056391,
+      "step": 11065
+    },
+    {
+      "epoch": 22.132,
+      "grad_norm": 0.8751665949821472,
+      "learning_rate": 2e-05,
+      "loss": 0.01894474,
+      "step": 11066
+    },
+    {
+      "epoch": 22.134,
+      "grad_norm": 1.1677522659301758,
+      "learning_rate": 2e-05,
+      "loss": 0.03264951,
+      "step": 11067
+    },
+    {
+      "epoch": 22.136,
+      "grad_norm": 1.8770594596862793,
+      "learning_rate": 2e-05,
+      "loss": 0.04727862,
+      "step": 11068
+    },
+    {
+      "epoch": 22.138,
+      "grad_norm": 0.9931758642196655,
+      "learning_rate": 2e-05,
+      "loss": 0.02625104,
+      "step": 11069
+    },
+    {
+      "epoch": 22.14,
+      "grad_norm": 1.8161505460739136,
+      "learning_rate": 2e-05,
+      "loss": 0.0335126,
+      "step": 11070
+    },
+    {
+      "epoch": 22.142,
+      "grad_norm": 1.2563273906707764,
+      "learning_rate": 2e-05,
+      "loss": 0.03935399,
+      "step": 11071
+    },
+    {
+      "epoch": 22.144,
+      "grad_norm": 1.6449320316314697,
+      "learning_rate": 2e-05,
+      "loss": 0.03845852,
+      "step": 11072
+    },
+    {
+      "epoch": 22.146,
+      "grad_norm": 1.3035930395126343,
+      "learning_rate": 2e-05,
+      "loss": 0.03262253,
+      "step": 11073
+    },
+    {
+      "epoch": 22.148,
+      "grad_norm": 1.5313411951065063,
+      "learning_rate": 2e-05,
+      "loss": 0.03509212,
+      "step": 11074
+    },
+    {
+      "epoch": 22.15,
+      "grad_norm": 2.219724178314209,
+      "learning_rate": 2e-05,
+      "loss": 0.05217472,
+      "step": 11075
+    },
+    {
+      "epoch": 22.152,
+      "grad_norm": 1.2089931964874268,
+      "learning_rate": 2e-05,
+      "loss": 0.03936118,
+      "step": 11076
+    },
+    {
+      "epoch": 22.154,
+      "grad_norm": 1.1982935667037964,
+      "learning_rate": 2e-05,
+      "loss": 0.03750388,
+      "step": 11077
+    },
+    {
+      "epoch": 22.156,
+      "grad_norm": 1.3776034116744995,
+      "learning_rate": 2e-05,
+      "loss": 0.02474811,
+      "step": 11078
+    },
+    {
+      "epoch": 22.158,
+      "grad_norm": 1.5565482378005981,
+      "learning_rate": 2e-05,
+      "loss": 0.04205162,
+      "step": 11079
+    },
+    {
+      "epoch": 22.16,
+      "grad_norm": 1.5331981182098389,
+      "learning_rate": 2e-05,
+      "loss": 0.05201468,
+      "step": 11080
+    },
+    {
+      "epoch": 22.162,
+      "grad_norm": 1.260038137435913,
+      "learning_rate": 2e-05,
+      "loss": 0.03021825,
+      "step": 11081
+    },
+    {
+      "epoch": 22.164,
+      "grad_norm": 1.128806471824646,
+      "learning_rate": 2e-05,
+      "loss": 0.03233521,
+      "step": 11082
+    },
+    {
+      "epoch": 22.166,
+      "grad_norm": 1.0538668632507324,
+      "learning_rate": 2e-05,
+      "loss": 0.02831938,
+      "step": 11083
+    },
+    {
+      "epoch": 22.168,
+      "grad_norm": 1.506290078163147,
+      "learning_rate": 2e-05,
+      "loss": 0.03654964,
+      "step": 11084
+    },
+    {
+      "epoch": 22.17,
+      "grad_norm": 1.4972366094589233,
+      "learning_rate": 2e-05,
+      "loss": 0.04231321,
+      "step": 11085
+    },
+    {
+      "epoch": 22.172,
+      "grad_norm": 1.2762951850891113,
+      "learning_rate": 2e-05,
+      "loss": 0.04460282,
+      "step": 11086
+    },
+    {
+      "epoch": 22.174,
+      "grad_norm": 1.0688859224319458,
+      "learning_rate": 2e-05,
+      "loss": 0.03068903,
+      "step": 11087
+    },
+    {
+      "epoch": 22.176,
+      "grad_norm": 1.194064974784851,
+      "learning_rate": 2e-05,
+      "loss": 0.03391276,
+      "step": 11088
+    },
+    {
+      "epoch": 22.178,
+      "grad_norm": 0.7239233255386353,
+      "learning_rate": 2e-05,
+      "loss": 0.01693652,
+      "step": 11089
+    },
+    {
+      "epoch": 22.18,
+      "grad_norm": 1.0344641208648682,
+      "learning_rate": 2e-05,
+      "loss": 0.03364387,
+      "step": 11090
+    },
+    {
+      "epoch": 22.182,
+      "grad_norm": 1.1779170036315918,
+      "learning_rate": 2e-05,
+      "loss": 0.03767635,
+      "step": 11091
+    },
+    {
+      "epoch": 22.184,
+      "grad_norm": 1.9103333950042725,
+      "learning_rate": 2e-05,
+      "loss": 0.04368252,
+      "step": 11092
+    },
+    {
+      "epoch": 22.186,
+      "grad_norm": 1.6527727842330933,
+      "learning_rate": 2e-05,
+      "loss": 0.03112076,
+      "step": 11093
+    },
+    {
+      "epoch": 22.188,
+      "grad_norm": 0.8650185465812683,
+      "learning_rate": 2e-05,
+      "loss": 0.0230502,
+      "step": 11094
+    },
+    {
+      "epoch": 22.19,
+      "grad_norm": 2.3402581214904785,
+      "learning_rate": 2e-05,
+      "loss": 0.05151432,
+      "step": 11095
+    },
+    {
+      "epoch": 22.192,
+      "grad_norm": 1.087644100189209,
+      "learning_rate": 2e-05,
+      "loss": 0.02510383,
+      "step": 11096
+    },
+    {
+      "epoch": 22.194,
+      "grad_norm": 1.062873125076294,
+      "learning_rate": 2e-05,
+      "loss": 0.02815459,
+      "step": 11097
+    },
+    {
+      "epoch": 22.196,
+      "grad_norm": 1.5720903873443604,
+      "learning_rate": 2e-05,
+      "loss": 0.04108073,
+      "step": 11098
+    },
+    {
+      "epoch": 22.198,
+      "grad_norm": 1.1922218799591064,
+      "learning_rate": 2e-05,
+      "loss": 0.03966674,
+      "step": 11099
+    },
+    {
+      "epoch": 22.2,
+      "grad_norm": 1.7009609937667847,
+      "learning_rate": 2e-05,
+      "loss": 0.03648136,
+      "step": 11100
+    },
+    {
+      "epoch": 22.202,
+      "grad_norm": 1.3058127164840698,
+      "learning_rate": 2e-05,
+      "loss": 0.030958,
+      "step": 11101
+    },
+    {
+      "epoch": 22.204,
+      "grad_norm": 1.2523670196533203,
+      "learning_rate": 2e-05,
+      "loss": 0.03727802,
+      "step": 11102
+    },
+    {
+      "epoch": 22.206,
+      "grad_norm": 2.1313955783843994,
+      "learning_rate": 2e-05,
+      "loss": 0.03327603,
+      "step": 11103
+    },
+    {
+      "epoch": 22.208,
+      "grad_norm": 1.0386605262756348,
+      "learning_rate": 2e-05,
+      "loss": 0.03811013,
+      "step": 11104
+    },
+    {
+      "epoch": 22.21,
+      "grad_norm": 0.9282265305519104,
+      "learning_rate": 2e-05,
+      "loss": 0.02790239,
+      "step": 11105
+    },
+    {
+      "epoch": 22.212,
+      "grad_norm": 1.176169991493225,
+      "learning_rate": 2e-05,
+      "loss": 0.03345915,
+      "step": 11106
+    },
+    {
+      "epoch": 22.214,
+      "grad_norm": 1.745169758796692,
+      "learning_rate": 2e-05,
+      "loss": 0.04287583,
+      "step": 11107
+    },
+    {
+      "epoch": 22.216,
+      "grad_norm": 1.2296968698501587,
+      "learning_rate": 2e-05,
+      "loss": 0.04217129,
+      "step": 11108
+    },
+    {
+      "epoch": 22.218,
+      "grad_norm": 1.1114269495010376,
+      "learning_rate": 2e-05,
+      "loss": 0.04333063,
+      "step": 11109
+    },
+    {
+      "epoch": 22.22,
+      "grad_norm": 1.1733880043029785,
+      "learning_rate": 2e-05,
+      "loss": 0.03284467,
+      "step": 11110
+    },
+    {
+      "epoch": 22.222,
+      "grad_norm": 1.0831984281539917,
+      "learning_rate": 2e-05,
+      "loss": 0.03438616,
+      "step": 11111
+    },
+    {
+      "epoch": 22.224,
+      "grad_norm": 1.352116346359253,
+      "learning_rate": 2e-05,
+      "loss": 0.04775696,
+      "step": 11112
+    },
+    {
+      "epoch": 22.226,
+      "grad_norm": 1.0606354475021362,
+      "learning_rate": 2e-05,
+      "loss": 0.02460128,
+      "step": 11113
+    },
+    {
+      "epoch": 22.228,
+      "grad_norm": 1.1735827922821045,
+      "learning_rate": 2e-05,
+      "loss": 0.03722972,
+      "step": 11114
+    },
+    {
+      "epoch": 22.23,
+      "grad_norm": 0.9459177851676941,
+      "learning_rate": 2e-05,
+      "loss": 0.03487729,
+      "step": 11115
+    },
+    {
+      "epoch": 22.232,
+      "grad_norm": 1.0781097412109375,
+      "learning_rate": 2e-05,
+      "loss": 0.02943964,
+      "step": 11116
+    },
+    {
+      "epoch": 22.234,
+      "grad_norm": 1.2076375484466553,
+      "learning_rate": 2e-05,
+      "loss": 0.03915054,
+      "step": 11117
+    },
+    {
+      "epoch": 22.236,
+      "grad_norm": 0.8902519941329956,
+      "learning_rate": 2e-05,
+      "loss": 0.02953203,
+      "step": 11118
+    },
+    {
+      "epoch": 22.238,
+      "grad_norm": 1.423532247543335,
+      "learning_rate": 2e-05,
+      "loss": 0.03786097,
+      "step": 11119
+    },
+    {
+      "epoch": 22.24,
+      "grad_norm": 1.068610668182373,
+      "learning_rate": 2e-05,
+      "loss": 0.0284631,
+      "step": 11120
+    },
+    {
+      "epoch": 22.242,
+      "grad_norm": 1.3024731874465942,
+      "learning_rate": 2e-05,
+      "loss": 0.0295128,
+      "step": 11121
+    },
+    {
+      "epoch": 22.244,
+      "grad_norm": 0.8875964283943176,
+      "learning_rate": 2e-05,
+      "loss": 0.02831073,
+      "step": 11122
+    },
+    {
+      "epoch": 22.246,
+      "grad_norm": 1.04351806640625,
+      "learning_rate": 2e-05,
+      "loss": 0.03519507,
+      "step": 11123
+    },
+    {
+      "epoch": 22.248,
+      "grad_norm": 0.969316303730011,
+      "learning_rate": 2e-05,
+      "loss": 0.02787546,
+      "step": 11124
+    },
+    {
+      "epoch": 22.25,
+      "grad_norm": 1.1152201890945435,
+      "learning_rate": 2e-05,
+      "loss": 0.03151672,
+      "step": 11125
+    },
+    {
+      "epoch": 22.252,
+      "grad_norm": 1.1251158714294434,
+      "learning_rate": 2e-05,
+      "loss": 0.03841231,
+      "step": 11126
+    },
+    {
+      "epoch": 22.254,
+      "grad_norm": 1.6095986366271973,
+      "learning_rate": 2e-05,
+      "loss": 0.03633893,
+      "step": 11127
+    },
+    {
+      "epoch": 22.256,
+      "grad_norm": 1.3410676717758179,
+      "learning_rate": 2e-05,
+      "loss": 0.0304535,
+      "step": 11128
+    },
+    {
+      "epoch": 22.258,
+      "grad_norm": 1.5073579549789429,
+      "learning_rate": 2e-05,
+      "loss": 0.04205672,
+      "step": 11129
+    },
+    {
+      "epoch": 22.26,
+      "grad_norm": 1.2273343801498413,
+      "learning_rate": 2e-05,
+      "loss": 0.02800684,
+      "step": 11130
+    },
+    {
+      "epoch": 22.262,
+      "grad_norm": 1.2454593181610107,
+      "learning_rate": 2e-05,
+      "loss": 0.03668667,
+      "step": 11131
+    },
+    {
+      "epoch": 22.264,
+      "grad_norm": 0.9008411169052124,
+      "learning_rate": 2e-05,
+      "loss": 0.02018595,
+      "step": 11132
+    },
+    {
+      "epoch": 22.266,
+      "grad_norm": 1.4407562017440796,
+      "learning_rate": 2e-05,
+      "loss": 0.02953599,
+      "step": 11133
+    },
+    {
+      "epoch": 22.268,
+      "grad_norm": 1.4390043020248413,
+      "learning_rate": 2e-05,
+      "loss": 0.02695351,
+      "step": 11134
+    },
+    {
+      "epoch": 22.27,
+      "grad_norm": 1.291393756866455,
+      "learning_rate": 2e-05,
+      "loss": 0.03049627,
+      "step": 11135
+    },
+    {
+      "epoch": 22.272,
+      "grad_norm": 1.229018211364746,
+      "learning_rate": 2e-05,
+      "loss": 0.03298253,
+      "step": 11136
+    },
+    {
+      "epoch": 22.274,
+      "grad_norm": 2.3607177734375,
+      "learning_rate": 2e-05,
+      "loss": 0.03460927,
+      "step": 11137
+    },
+    {
+      "epoch": 22.276,
+      "grad_norm": 4.348043918609619,
+      "learning_rate": 2e-05,
+      "loss": 0.04996082,
+      "step": 11138
+    },
+    {
+      "epoch": 22.278,
+      "grad_norm": 1.7652407884597778,
+      "learning_rate": 2e-05,
+      "loss": 0.02721683,
+      "step": 11139
+    },
+    {
+      "epoch": 22.28,
+      "grad_norm": 1.593738317489624,
+      "learning_rate": 2e-05,
+      "loss": 0.03514788,
+      "step": 11140
+    },
+    {
+      "epoch": 22.282,
+      "grad_norm": 1.2631458044052124,
+      "learning_rate": 2e-05,
+      "loss": 0.03659892,
+      "step": 11141
+    },
+    {
+      "epoch": 22.284,
+      "grad_norm": 1.642995834350586,
+      "learning_rate": 2e-05,
+      "loss": 0.03794821,
+      "step": 11142
+    },
+    {
+      "epoch": 22.286,
+      "grad_norm": 1.9356567859649658,
+      "learning_rate": 2e-05,
+      "loss": 0.03757293,
+      "step": 11143
+    },
+    {
+      "epoch": 22.288,
+      "grad_norm": 0.9092358946800232,
+      "learning_rate": 2e-05,
+      "loss": 0.02223909,
+      "step": 11144
+    },
+    {
+      "epoch": 22.29,
+      "grad_norm": 3.2695510387420654,
+      "learning_rate": 2e-05,
+      "loss": 0.03481711,
+      "step": 11145
+    },
+    {
+      "epoch": 22.292,
+      "grad_norm": 0.9674291610717773,
+      "learning_rate": 2e-05,
+      "loss": 0.02906576,
+      "step": 11146
+    },
+    {
+      "epoch": 22.294,
+      "grad_norm": 6.927521705627441,
+      "learning_rate": 2e-05,
+      "loss": 0.05944148,
+      "step": 11147
+    },
+    {
+      "epoch": 22.296,
+      "grad_norm": 1.8510019779205322,
+      "learning_rate": 2e-05,
+      "loss": 0.04159633,
+      "step": 11148
+    },
+    {
+      "epoch": 22.298,
+      "grad_norm": 1.190011739730835,
+      "learning_rate": 2e-05,
+      "loss": 0.03650186,
+      "step": 11149
+    },
+    {
+      "epoch": 22.3,
+      "grad_norm": 0.9493358731269836,
+      "learning_rate": 2e-05,
+      "loss": 0.03242558,
+      "step": 11150
+    },
+    {
+      "epoch": 22.302,
+      "grad_norm": 3.193427324295044,
+      "learning_rate": 2e-05,
+      "loss": 0.04184601,
+      "step": 11151
+    },
+    {
+      "epoch": 22.304,
+      "grad_norm": 1.372183918952942,
+      "learning_rate": 2e-05,
+      "loss": 0.03488796,
+      "step": 11152
+    },
+    {
+      "epoch": 22.306,
+      "grad_norm": 2.2507684230804443,
+      "learning_rate": 2e-05,
+      "loss": 0.04868758,
+      "step": 11153
+    },
+    {
+      "epoch": 22.308,
+      "grad_norm": 1.2131235599517822,
+      "learning_rate": 2e-05,
+      "loss": 0.02608864,
+      "step": 11154
+    },
+    {
+      "epoch": 22.31,
+      "grad_norm": 1.179132103919983,
+      "learning_rate": 2e-05,
+      "loss": 0.03008761,
+      "step": 11155
+    },
+    {
+      "epoch": 22.312,
+      "grad_norm": 1.2756638526916504,
+      "learning_rate": 2e-05,
+      "loss": 0.03774351,
+      "step": 11156
+    },
+    {
+      "epoch": 22.314,
+      "grad_norm": 1.5968365669250488,
+      "learning_rate": 2e-05,
+      "loss": 0.0491349,
+      "step": 11157
+    },
+    {
+      "epoch": 22.316,
+      "grad_norm": 2.220818519592285,
+      "learning_rate": 2e-05,
+      "loss": 0.02767688,
+      "step": 11158
+    },
+    {
+      "epoch": 22.318,
+      "grad_norm": 0.9472852349281311,
+      "learning_rate": 2e-05,
+      "loss": 0.02426755,
+      "step": 11159
+    },
+    {
+      "epoch": 22.32,
+      "grad_norm": 2.503251552581787,
+      "learning_rate": 2e-05,
+      "loss": 0.0384932,
+      "step": 11160
+    },
+    {
+      "epoch": 22.322,
+      "grad_norm": 3.294299602508545,
+      "learning_rate": 2e-05,
+      "loss": 0.06644698,
+      "step": 11161
+    },
+    {
+      "epoch": 22.324,
+      "grad_norm": 1.2162548303604126,
+      "learning_rate": 2e-05,
+      "loss": 0.03608144,
+      "step": 11162
+    },
+    {
+      "epoch": 22.326,
+      "grad_norm": 1.3543113470077515,
+      "learning_rate": 2e-05,
+      "loss": 0.03112835,
+      "step": 11163
+    },
+    {
+      "epoch": 22.328,
+      "grad_norm": 1.1519101858139038,
+      "learning_rate": 2e-05,
+      "loss": 0.02835603,
+      "step": 11164
+    },
+    {
+      "epoch": 22.33,
+      "grad_norm": 0.905073344707489,
+      "learning_rate": 2e-05,
+      "loss": 0.03140076,
+      "step": 11165
+    },
+    {
+      "epoch": 22.332,
+      "grad_norm": 1.0447617769241333,
+      "learning_rate": 2e-05,
+      "loss": 0.02909181,
+      "step": 11166
+    },
+    {
+      "epoch": 22.334,
+      "grad_norm": 1.8674046993255615,
+      "learning_rate": 2e-05,
+      "loss": 0.0350634,
+      "step": 11167
+    },
+    {
+      "epoch": 22.336,
+      "grad_norm": 1.658823013305664,
+      "learning_rate": 2e-05,
+      "loss": 0.04215661,
+      "step": 11168
+    },
+    {
+      "epoch": 22.338,
+      "grad_norm": 0.9590745568275452,
+      "learning_rate": 2e-05,
+      "loss": 0.03000915,
+      "step": 11169
+    },
+    {
+      "epoch": 22.34,
+      "grad_norm": 1.0610401630401611,
+      "learning_rate": 2e-05,
+      "loss": 0.03618003,
+      "step": 11170
+    },
+    {
+      "epoch": 22.342,
+      "grad_norm": 1.1848183870315552,
+      "learning_rate": 2e-05,
+      "loss": 0.03476301,
+      "step": 11171
+    },
+    {
+      "epoch": 22.344,
+      "grad_norm": 1.2776687145233154,
+      "learning_rate": 2e-05,
+      "loss": 0.03656887,
+      "step": 11172
+    },
+    {
+      "epoch": 22.346,
+      "grad_norm": 1.9115201234817505,
+      "learning_rate": 2e-05,
+      "loss": 0.04781888,
+      "step": 11173
+    },
+    {
+      "epoch": 22.348,
+      "grad_norm": 2.2884371280670166,
+      "learning_rate": 2e-05,
+      "loss": 0.03745548,
+      "step": 11174
+    },
+    {
+      "epoch": 22.35,
+      "grad_norm": 0.9647842049598694,
+      "learning_rate": 2e-05,
+      "loss": 0.02903546,
+      "step": 11175
+    },
+    {
+      "epoch": 22.352,
+      "grad_norm": 1.4671611785888672,
+      "learning_rate": 2e-05,
+      "loss": 0.05149737,
+      "step": 11176
+    },
+    {
+      "epoch": 22.354,
+      "grad_norm": 1.0586042404174805,
+      "learning_rate": 2e-05,
+      "loss": 0.0410578,
+      "step": 11177
+    },
+    {
+      "epoch": 22.356,
+      "grad_norm": 1.2867225408554077,
+      "learning_rate": 2e-05,
+      "loss": 0.04615509,
+      "step": 11178
+    },
+    {
+      "epoch": 22.358,
+      "grad_norm": 1.0811021327972412,
+      "learning_rate": 2e-05,
+      "loss": 0.03351441,
+      "step": 11179
+    },
+    {
+      "epoch": 22.36,
+      "grad_norm": 2.024754762649536,
+      "learning_rate": 2e-05,
+      "loss": 0.04169374,
+      "step": 11180
+    },
+    {
+      "epoch": 22.362,
+      "grad_norm": 1.5441582202911377,
+      "learning_rate": 2e-05,
+      "loss": 0.03294868,
+      "step": 11181
+    },
+    {
+      "epoch": 22.364,
+      "grad_norm": 1.9213157892227173,
+      "learning_rate": 2e-05,
+      "loss": 0.04192325,
+      "step": 11182
+    },
+    {
+      "epoch": 22.366,
+      "grad_norm": 1.0890470743179321,
+      "learning_rate": 2e-05,
+      "loss": 0.03932363,
+      "step": 11183
+    },
+    {
+      "epoch": 22.368,
+      "grad_norm": 1.6594054698944092,
+      "learning_rate": 2e-05,
+      "loss": 0.04021504,
+      "step": 11184
+    },
+    {
+      "epoch": 22.37,
+      "grad_norm": 1.811708927154541,
+      "learning_rate": 2e-05,
+      "loss": 0.05453177,
+      "step": 11185
+    },
+    {
+      "epoch": 22.372,
+      "grad_norm": 1.0302735567092896,
+      "learning_rate": 2e-05,
+      "loss": 0.03242699,
+      "step": 11186
+    },
+    {
+      "epoch": 22.374,
+      "grad_norm": 0.9932673573493958,
+      "learning_rate": 2e-05,
+      "loss": 0.02922393,
+      "step": 11187
+    },
+    {
+      "epoch": 22.376,
+      "grad_norm": 1.15402352809906,
+      "learning_rate": 2e-05,
+      "loss": 0.03236395,
+      "step": 11188
+    },
+    {
+      "epoch": 22.378,
+      "grad_norm": 0.9961234927177429,
+      "learning_rate": 2e-05,
+      "loss": 0.03148089,
+      "step": 11189
+    },
+    {
+      "epoch": 22.38,
+      "grad_norm": 1.1807105541229248,
+      "learning_rate": 2e-05,
+      "loss": 0.03198289,
+      "step": 11190
+    },
+    {
+      "epoch": 22.382,
+      "grad_norm": 1.0972753763198853,
+      "learning_rate": 2e-05,
+      "loss": 0.03932333,
+      "step": 11191
+    },
+    {
+      "epoch": 22.384,
+      "grad_norm": 1.3407927751541138,
+      "learning_rate": 2e-05,
+      "loss": 0.02748748,
+      "step": 11192
+    },
+    {
+      "epoch": 22.386,
+      "grad_norm": 1.1336009502410889,
+      "learning_rate": 2e-05,
+      "loss": 0.03101401,
+      "step": 11193
+    },
+    {
+      "epoch": 22.388,
+      "grad_norm": 0.8580100536346436,
+      "learning_rate": 2e-05,
+      "loss": 0.01859001,
+      "step": 11194
+    },
+    {
+      "epoch": 22.39,
+      "grad_norm": 1.9567729234695435,
+      "learning_rate": 2e-05,
+      "loss": 0.03903724,
+      "step": 11195
+    },
+    {
+      "epoch": 22.392,
+      "grad_norm": 1.1536086797714233,
+      "learning_rate": 2e-05,
+      "loss": 0.03743483,
+      "step": 11196
+    },
+    {
+      "epoch": 22.394,
+      "grad_norm": 1.7297052145004272,
+      "learning_rate": 2e-05,
+      "loss": 0.04060826,
+      "step": 11197
+    },
+    {
+      "epoch": 22.396,
+      "grad_norm": 1.1482161283493042,
+      "learning_rate": 2e-05,
+      "loss": 0.04031848,
+      "step": 11198
+    },
+    {
+      "epoch": 22.398,
+      "grad_norm": 1.4597668647766113,
+      "learning_rate": 2e-05,
+      "loss": 0.04166424,
+      "step": 11199
+    },
+    {
+      "epoch": 22.4,
+      "grad_norm": 1.0787426233291626,
+      "learning_rate": 2e-05,
+      "loss": 0.0363807,
+      "step": 11200
+    },
+    {
+      "epoch": 22.402,
+      "grad_norm": 2.6459505558013916,
+      "learning_rate": 2e-05,
+      "loss": 0.03180039,
+      "step": 11201
+    },
+    {
+      "epoch": 22.404,
+      "grad_norm": 1.28476881980896,
+      "learning_rate": 2e-05,
+      "loss": 0.03239727,
+      "step": 11202
+    },
+    {
+      "epoch": 22.406,
+      "grad_norm": 1.4840121269226074,
+      "learning_rate": 2e-05,
+      "loss": 0.02138729,
+      "step": 11203
+    },
+    {
+      "epoch": 22.408,
+      "grad_norm": 1.6134271621704102,
+      "learning_rate": 2e-05,
+      "loss": 0.05612726,
+      "step": 11204
+    },
+    {
+      "epoch": 22.41,
+      "grad_norm": 1.0505448579788208,
+      "learning_rate": 2e-05,
+      "loss": 0.02743955,
+      "step": 11205
+    },
+    {
+      "epoch": 22.412,
+      "grad_norm": 1.2720935344696045,
+      "learning_rate": 2e-05,
+      "loss": 0.04470626,
+      "step": 11206
+    },
+    {
+      "epoch": 22.414,
+      "grad_norm": 1.216604232788086,
+      "learning_rate": 2e-05,
+      "loss": 0.03580245,
+      "step": 11207
+    },
+    {
+      "epoch": 22.416,
+      "grad_norm": 0.9434407949447632,
+      "learning_rate": 2e-05,
+      "loss": 0.02070105,
+      "step": 11208
+    },
+    {
+      "epoch": 22.418,
+      "grad_norm": 1.0345122814178467,
+      "learning_rate": 2e-05,
+      "loss": 0.0290225,
+      "step": 11209
+    },
+    {
+      "epoch": 22.42,
+      "grad_norm": 1.173581600189209,
+      "learning_rate": 2e-05,
+      "loss": 0.03839279,
+      "step": 11210
+    },
+    {
+      "epoch": 22.422,
+      "grad_norm": 1.084951400756836,
+      "learning_rate": 2e-05,
+      "loss": 0.03221672,
+      "step": 11211
+    },
+    {
+      "epoch": 22.424,
+      "grad_norm": 1.0095467567443848,
+      "learning_rate": 2e-05,
+      "loss": 0.02846614,
+      "step": 11212
+    },
+    {
+      "epoch": 22.426,
+      "grad_norm": 1.254839539527893,
+      "learning_rate": 2e-05,
+      "loss": 0.03883304,
+      "step": 11213
+    },
+    {
+      "epoch": 22.428,
+      "grad_norm": 2.123760461807251,
+      "learning_rate": 2e-05,
+      "loss": 0.03277697,
+      "step": 11214
+    },
+    {
+      "epoch": 22.43,
+      "grad_norm": 1.1430293321609497,
+      "learning_rate": 2e-05,
+      "loss": 0.02167347,
+      "step": 11215
+    },
+    {
+      "epoch": 22.432,
+      "grad_norm": 1.242515206336975,
+      "learning_rate": 2e-05,
+      "loss": 0.0301038,
+      "step": 11216
+    },
+    {
+      "epoch": 22.434,
+      "grad_norm": 1.651579737663269,
+      "learning_rate": 2e-05,
+      "loss": 0.03766736,
+      "step": 11217
+    },
+    {
+      "epoch": 22.436,
+      "grad_norm": 0.8291965126991272,
+      "learning_rate": 2e-05,
+      "loss": 0.02428487,
+      "step": 11218
+    },
+    {
+      "epoch": 22.438,
+      "grad_norm": 0.8632979393005371,
+      "learning_rate": 2e-05,
+      "loss": 0.0235811,
+      "step": 11219
+    },
+    {
+      "epoch": 22.44,
+      "grad_norm": 1.366252064704895,
+      "learning_rate": 2e-05,
+      "loss": 0.04416512,
+      "step": 11220
+    },
+    {
+      "epoch": 22.442,
+      "grad_norm": 0.796024739742279,
+      "learning_rate": 2e-05,
+      "loss": 0.0193195,
+      "step": 11221
+    },
+    {
+      "epoch": 22.444,
+      "grad_norm": 1.679572582244873,
+      "learning_rate": 2e-05,
+      "loss": 0.04569577,
+      "step": 11222
+    },
+    {
+      "epoch": 22.446,
+      "grad_norm": 1.6671518087387085,
+      "learning_rate": 2e-05,
+      "loss": 0.03679072,
+      "step": 11223
+    },
+    {
+      "epoch": 22.448,
+      "grad_norm": 1.1129558086395264,
+      "learning_rate": 2e-05,
+      "loss": 0.03054724,
+      "step": 11224
+    },
+    {
+      "epoch": 22.45,
+      "grad_norm": 5.476849555969238,
+      "learning_rate": 2e-05,
+      "loss": 0.03754855,
+      "step": 11225
+    },
+    {
+      "epoch": 22.452,
+      "grad_norm": 1.0009487867355347,
+      "learning_rate": 2e-05,
+      "loss": 0.03189157,
+      "step": 11226
+    },
+    {
+      "epoch": 22.454,
+      "grad_norm": 1.420681118965149,
+      "learning_rate": 2e-05,
+      "loss": 0.0333651,
+      "step": 11227
+    },
+    {
+      "epoch": 22.456,
+      "grad_norm": 1.3200653791427612,
+      "learning_rate": 2e-05,
+      "loss": 0.0283935,
+      "step": 11228
+    },
+    {
+      "epoch": 22.458,
+      "grad_norm": 2.4781174659729004,
+      "learning_rate": 2e-05,
+      "loss": 0.0337758,
+      "step": 11229
+    },
+    {
+      "epoch": 22.46,
+      "grad_norm": 1.2006012201309204,
+      "learning_rate": 2e-05,
+      "loss": 0.04407885,
+      "step": 11230
+    },
+    {
+      "epoch": 22.462,
+      "grad_norm": 0.8424386978149414,
+      "learning_rate": 2e-05,
+      "loss": 0.02036972,
+      "step": 11231
+    },
+    {
+      "epoch": 22.464,
+      "grad_norm": 1.0086331367492676,
+      "learning_rate": 2e-05,
+      "loss": 0.02813908,
+      "step": 11232
+    },
+    {
+      "epoch": 22.466,
+      "grad_norm": 1.6784151792526245,
+      "learning_rate": 2e-05,
+      "loss": 0.05713883,
+      "step": 11233
+    },
+    {
+      "epoch": 22.468,
+      "grad_norm": 1.6650753021240234,
+      "learning_rate": 2e-05,
+      "loss": 0.0437688,
+      "step": 11234
+    },
+    {
+      "epoch": 22.47,
+      "grad_norm": 1.033752679824829,
+      "learning_rate": 2e-05,
+      "loss": 0.02289294,
+      "step": 11235
+    },
+    {
+      "epoch": 22.472,
+      "grad_norm": 1.9619020223617554,
+      "learning_rate": 2e-05,
+      "loss": 0.03555152,
+      "step": 11236
+    },
+    {
+      "epoch": 22.474,
+      "grad_norm": 0.9467611908912659,
+      "learning_rate": 2e-05,
+      "loss": 0.02932581,
+      "step": 11237
+    },
+    {
+      "epoch": 22.476,
+      "grad_norm": 1.2341922521591187,
+      "learning_rate": 2e-05,
+      "loss": 0.0385951,
+      "step": 11238
+    },
+    {
+      "epoch": 22.478,
+      "grad_norm": 1.1940884590148926,
+      "learning_rate": 2e-05,
+      "loss": 0.03371501,
+      "step": 11239
+    },
+    {
+      "epoch": 22.48,
+      "grad_norm": 1.9683741331100464,
+      "learning_rate": 2e-05,
+      "loss": 0.02918104,
+      "step": 11240
+    },
+    {
+      "epoch": 22.482,
+      "grad_norm": 1.0450471639633179,
+      "learning_rate": 2e-05,
+      "loss": 0.02672218,
+      "step": 11241
+    },
+    {
+      "epoch": 22.484,
+      "grad_norm": 2.0168466567993164,
+      "learning_rate": 2e-05,
+      "loss": 0.03515015,
+      "step": 11242
+    },
+    {
+      "epoch": 22.486,
+      "grad_norm": 1.5114184617996216,
+      "learning_rate": 2e-05,
+      "loss": 0.04158841,
+      "step": 11243
+    },
+    {
+      "epoch": 22.488,
+      "grad_norm": 1.5208250284194946,
+      "learning_rate": 2e-05,
+      "loss": 0.05399328,
+      "step": 11244
+    },
+    {
+      "epoch": 22.49,
+      "grad_norm": 1.0174354314804077,
+      "learning_rate": 2e-05,
+      "loss": 0.02212554,
+      "step": 11245
+    },
+    {
+      "epoch": 22.492,
+      "grad_norm": 1.1009485721588135,
+      "learning_rate": 2e-05,
+      "loss": 0.03259212,
+      "step": 11246
+    },
+    {
+      "epoch": 22.494,
+      "grad_norm": 1.2722481489181519,
+      "learning_rate": 2e-05,
+      "loss": 0.03970824,
+      "step": 11247
+    },
+    {
+      "epoch": 22.496,
+      "grad_norm": 1.4732874631881714,
+      "learning_rate": 2e-05,
+      "loss": 0.03239143,
+      "step": 11248
+    },
+    {
+      "epoch": 22.498,
+      "grad_norm": 1.8055660724639893,
+      "learning_rate": 2e-05,
+      "loss": 0.04277326,
+      "step": 11249
+    },
+    {
+      "epoch": 22.5,
+      "grad_norm": 1.5957167148590088,
+      "learning_rate": 2e-05,
+      "loss": 0.04040106,
+      "step": 11250
+    },
+    {
+      "epoch": 22.502,
+      "grad_norm": 1.0191160440444946,
+      "learning_rate": 2e-05,
+      "loss": 0.03819025,
+      "step": 11251
+    },
+    {
+      "epoch": 22.504,
+      "grad_norm": 0.8751825094223022,
+      "learning_rate": 2e-05,
+      "loss": 0.02078329,
+      "step": 11252
+    },
+    {
+      "epoch": 22.506,
+      "grad_norm": 0.9703699350357056,
+      "learning_rate": 2e-05,
+      "loss": 0.02371472,
+      "step": 11253
+    },
+    {
+      "epoch": 22.508,
+      "grad_norm": 1.4609352350234985,
+      "learning_rate": 2e-05,
+      "loss": 0.04331794,
+      "step": 11254
+    },
+    {
+      "epoch": 22.51,
+      "grad_norm": 0.7537633180618286,
+      "learning_rate": 2e-05,
+      "loss": 0.02106013,
+      "step": 11255
+    },
+    {
+      "epoch": 22.512,
+      "grad_norm": 1.1532191038131714,
+      "learning_rate": 2e-05,
+      "loss": 0.02696463,
+      "step": 11256
+    },
+    {
+      "epoch": 22.514,
+      "grad_norm": 1.2470663785934448,
+      "learning_rate": 2e-05,
+      "loss": 0.03676067,
+      "step": 11257
+    },
+    {
+      "epoch": 22.516,
+      "grad_norm": 1.034314751625061,
+      "learning_rate": 2e-05,
+      "loss": 0.03010433,
+      "step": 11258
+    },
+    {
+      "epoch": 22.518,
+      "grad_norm": 1.0208652019500732,
+      "learning_rate": 2e-05,
+      "loss": 0.0345837,
+      "step": 11259
+    },
+    {
+      "epoch": 22.52,
+      "grad_norm": 1.0119693279266357,
+      "learning_rate": 2e-05,
+      "loss": 0.02655143,
+      "step": 11260
+    },
+    {
+      "epoch": 22.522,
+      "grad_norm": 1.8333549499511719,
+      "learning_rate": 2e-05,
+      "loss": 0.04552222,
+      "step": 11261
+    },
+    {
+      "epoch": 22.524,
+      "grad_norm": 1.5406705141067505,
+      "learning_rate": 2e-05,
+      "loss": 0.04645106,
+      "step": 11262
+    },
+    {
+      "epoch": 22.526,
+      "grad_norm": 0.7743955850601196,
+      "learning_rate": 2e-05,
+      "loss": 0.02030542,
+      "step": 11263
+    },
+    {
+      "epoch": 22.528,
+      "grad_norm": 1.0403064489364624,
+      "learning_rate": 2e-05,
+      "loss": 0.02845855,
+      "step": 11264
+    },
+    {
+      "epoch": 22.53,
+      "grad_norm": 1.3076633214950562,
+      "learning_rate": 2e-05,
+      "loss": 0.04241259,
+      "step": 11265
+    },
+    {
+      "epoch": 22.532,
+      "grad_norm": 1.454092025756836,
+      "learning_rate": 2e-05,
+      "loss": 0.04193448,
+      "step": 11266
+    },
+    {
+      "epoch": 22.534,
+      "grad_norm": 1.0106709003448486,
+      "learning_rate": 2e-05,
+      "loss": 0.03059579,
+      "step": 11267
+    },
+    {
+      "epoch": 22.536,
+      "grad_norm": 1.4397797584533691,
+      "learning_rate": 2e-05,
+      "loss": 0.03595772,
+      "step": 11268
+    },
+    {
+      "epoch": 22.538,
+      "grad_norm": 1.0913630723953247,
+      "learning_rate": 2e-05,
+      "loss": 0.02504985,
+      "step": 11269
+    },
+    {
+      "epoch": 22.54,
+      "grad_norm": 1.1001238822937012,
+      "learning_rate": 2e-05,
+      "loss": 0.03245225,
+      "step": 11270
+    },
+    {
+      "epoch": 22.542,
+      "grad_norm": 1.6088941097259521,
+      "learning_rate": 2e-05,
+      "loss": 0.05301979,
+      "step": 11271
+    },
+    {
+      "epoch": 22.544,
+      "grad_norm": 1.3868272304534912,
+      "learning_rate": 2e-05,
+      "loss": 0.04549325,
+      "step": 11272
+    },
+    {
+      "epoch": 22.546,
+      "grad_norm": 1.3786357641220093,
+      "learning_rate": 2e-05,
+      "loss": 0.03595918,
+      "step": 11273
+    },
+    {
+      "epoch": 22.548000000000002,
+      "grad_norm": 0.9569694995880127,
+      "learning_rate": 2e-05,
+      "loss": 0.03288563,
+      "step": 11274
+    },
+    {
+      "epoch": 22.55,
+      "grad_norm": 1.6313834190368652,
+      "learning_rate": 2e-05,
+      "loss": 0.0272957,
+      "step": 11275
+    },
+    {
+      "epoch": 22.552,
+      "grad_norm": 1.3030887842178345,
+      "learning_rate": 2e-05,
+      "loss": 0.03881227,
+      "step": 11276
+    },
+    {
+      "epoch": 22.554,
+      "grad_norm": 1.4454890489578247,
+      "learning_rate": 2e-05,
+      "loss": 0.03911527,
+      "step": 11277
+    },
+    {
+      "epoch": 22.556,
+      "grad_norm": 1.0935059785842896,
+      "learning_rate": 2e-05,
+      "loss": 0.02929432,
+      "step": 11278
+    },
+    {
+      "epoch": 22.558,
+      "grad_norm": 1.0467413663864136,
+      "learning_rate": 2e-05,
+      "loss": 0.03879336,
+      "step": 11279
+    },
+    {
+      "epoch": 22.56,
+      "grad_norm": 1.3743126392364502,
+      "learning_rate": 2e-05,
+      "loss": 0.03689353,
+      "step": 11280
+    },
+    {
+      "epoch": 22.562,
+      "grad_norm": 0.9728202819824219,
+      "learning_rate": 2e-05,
+      "loss": 0.02886998,
+      "step": 11281
+    },
+    {
+      "epoch": 22.564,
+      "grad_norm": 1.2006583213806152,
+      "learning_rate": 2e-05,
+      "loss": 0.03440745,
+      "step": 11282
+    },
+    {
+      "epoch": 22.566,
+      "grad_norm": 1.1317431926727295,
+      "learning_rate": 2e-05,
+      "loss": 0.03354093,
+      "step": 11283
+    },
+    {
+      "epoch": 22.568,
+      "grad_norm": 1.0374213457107544,
+      "learning_rate": 2e-05,
+      "loss": 0.0318547,
+      "step": 11284
+    },
+    {
+      "epoch": 22.57,
+      "grad_norm": 1.3933595418930054,
+      "learning_rate": 2e-05,
+      "loss": 0.03949068,
+      "step": 11285
+    },
+    {
+      "epoch": 22.572,
+      "grad_norm": 1.9650115966796875,
+      "learning_rate": 2e-05,
+      "loss": 0.0620589,
+      "step": 11286
+    },
+    {
+      "epoch": 22.574,
+      "grad_norm": 1.1582067012786865,
+      "learning_rate": 2e-05,
+      "loss": 0.03287086,
+      "step": 11287
+    },
+    {
+      "epoch": 22.576,
+      "grad_norm": 1.4262588024139404,
+      "learning_rate": 2e-05,
+      "loss": 0.0474546,
+      "step": 11288
+    },
+    {
+      "epoch": 22.578,
+      "grad_norm": 1.8528952598571777,
+      "learning_rate": 2e-05,
+      "loss": 0.03344478,
+      "step": 11289
+    },
+    {
+      "epoch": 22.58,
+      "grad_norm": 2.4040780067443848,
+      "learning_rate": 2e-05,
+      "loss": 0.04099388,
+      "step": 11290
+    },
+    {
+      "epoch": 22.582,
+      "grad_norm": 1.1463083028793335,
+      "learning_rate": 2e-05,
+      "loss": 0.02880567,
+      "step": 11291
+    },
+    {
+      "epoch": 22.584,
+      "grad_norm": 1.1943498849868774,
+      "learning_rate": 2e-05,
+      "loss": 0.03802372,
+      "step": 11292
+    },
+    {
+      "epoch": 22.586,
+      "grad_norm": 0.8871561288833618,
+      "learning_rate": 2e-05,
+      "loss": 0.02874197,
+      "step": 11293
+    },
+    {
+      "epoch": 22.588,
+      "grad_norm": 1.1465365886688232,
+      "learning_rate": 2e-05,
+      "loss": 0.0340536,
+      "step": 11294
+    },
+    {
+      "epoch": 22.59,
+      "grad_norm": 0.9088796973228455,
+      "learning_rate": 2e-05,
+      "loss": 0.0175512,
+      "step": 11295
+    },
+    {
+      "epoch": 22.592,
+      "grad_norm": 1.8269764184951782,
+      "learning_rate": 2e-05,
+      "loss": 0.04000066,
+      "step": 11296
+    },
+    {
+      "epoch": 22.594,
+      "grad_norm": 0.9784151315689087,
+      "learning_rate": 2e-05,
+      "loss": 0.02507006,
+      "step": 11297
+    },
+    {
+      "epoch": 22.596,
+      "grad_norm": 1.2063928842544556,
+      "learning_rate": 2e-05,
+      "loss": 0.04113289,
+      "step": 11298
+    },
+    {
+      "epoch": 22.598,
+      "grad_norm": 1.7086107730865479,
+      "learning_rate": 2e-05,
+      "loss": 0.03416719,
+      "step": 11299
+    },
+    {
+      "epoch": 22.6,
+      "grad_norm": 0.8789715766906738,
+      "learning_rate": 2e-05,
+      "loss": 0.02553638,
+      "step": 11300
+    },
+    {
+      "epoch": 22.602,
+      "grad_norm": 1.5890235900878906,
+      "learning_rate": 2e-05,
+      "loss": 0.03483776,
+      "step": 11301
+    },
+    {
+      "epoch": 22.604,
+      "grad_norm": 1.4625310897827148,
+      "learning_rate": 2e-05,
+      "loss": 0.04801542,
+      "step": 11302
+    },
+    {
+      "epoch": 22.606,
+      "grad_norm": 1.0777349472045898,
+      "learning_rate": 2e-05,
+      "loss": 0.03591266,
+      "step": 11303
+    },
+    {
+      "epoch": 22.608,
+      "grad_norm": 0.9330995082855225,
+      "learning_rate": 2e-05,
+      "loss": 0.02881388,
+      "step": 11304
+    },
+    {
+      "epoch": 22.61,
+      "grad_norm": 1.3750163316726685,
+      "learning_rate": 2e-05,
+      "loss": 0.03394581,
+      "step": 11305
+    },
+    {
+      "epoch": 22.612,
+      "grad_norm": 1.35263192653656,
+      "learning_rate": 2e-05,
+      "loss": 0.03068662,
+      "step": 11306
+    },
+    {
+      "epoch": 22.614,
+      "grad_norm": 1.291716456413269,
+      "learning_rate": 2e-05,
+      "loss": 0.0435909,
+      "step": 11307
+    },
+    {
+      "epoch": 22.616,
+      "grad_norm": 1.3781776428222656,
+      "learning_rate": 2e-05,
+      "loss": 0.03991321,
+      "step": 11308
+    },
+    {
+      "epoch": 22.618,
+      "grad_norm": 1.09850013256073,
+      "learning_rate": 2e-05,
+      "loss": 0.04068404,
+      "step": 11309
+    },
+    {
+      "epoch": 22.62,
+      "grad_norm": 1.2705893516540527,
+      "learning_rate": 2e-05,
+      "loss": 0.04384391,
+      "step": 11310
+    },
+    {
+      "epoch": 22.622,
+      "grad_norm": 1.6616660356521606,
+      "learning_rate": 2e-05,
+      "loss": 0.02786833,
+      "step": 11311
+    },
+    {
+      "epoch": 22.624,
+      "grad_norm": 1.111756443977356,
+      "learning_rate": 2e-05,
+      "loss": 0.02686176,
+      "step": 11312
+    },
+    {
+      "epoch": 22.626,
+      "grad_norm": 1.2425123453140259,
+      "learning_rate": 2e-05,
+      "loss": 0.04210493,
+      "step": 11313
+    },
+    {
+      "epoch": 22.628,
+      "grad_norm": 1.0276856422424316,
+      "learning_rate": 2e-05,
+      "loss": 0.03203695,
+      "step": 11314
+    },
+    {
+      "epoch": 22.63,
+      "grad_norm": 1.0379737615585327,
+      "learning_rate": 2e-05,
+      "loss": 0.03296281,
+      "step": 11315
+    },
+    {
+      "epoch": 22.632,
+      "grad_norm": 0.9909905195236206,
+      "learning_rate": 2e-05,
+      "loss": 0.02666735,
+      "step": 11316
+    },
+    {
+      "epoch": 22.634,
+      "grad_norm": 1.0876964330673218,
+      "learning_rate": 2e-05,
+      "loss": 0.03161968,
+      "step": 11317
+    },
+    {
+      "epoch": 22.636,
+      "grad_norm": 1.085287094116211,
+      "learning_rate": 2e-05,
+      "loss": 0.02732705,
+      "step": 11318
+    },
+    {
+      "epoch": 22.638,
+      "grad_norm": 2.1646599769592285,
+      "learning_rate": 2e-05,
+      "loss": 0.02609478,
+      "step": 11319
+    },
+    {
+      "epoch": 22.64,
+      "grad_norm": 0.967041552066803,
+      "learning_rate": 2e-05,
+      "loss": 0.02320741,
+      "step": 11320
+    },
+    {
+      "epoch": 22.642,
+      "grad_norm": 1.5888314247131348,
+      "learning_rate": 2e-05,
+      "loss": 0.05063655,
+      "step": 11321
+    },
+    {
+      "epoch": 22.644,
+      "grad_norm": 0.9226511120796204,
+      "learning_rate": 2e-05,
+      "loss": 0.03058358,
+      "step": 11322
+    },
+    {
+      "epoch": 22.646,
+      "grad_norm": 1.701810598373413,
+      "learning_rate": 2e-05,
+      "loss": 0.02915188,
+      "step": 11323
+    },
+    {
+      "epoch": 22.648,
+      "grad_norm": 1.497314453125,
+      "learning_rate": 2e-05,
+      "loss": 0.04008635,
+      "step": 11324
+    },
+    {
+      "epoch": 22.65,
+      "grad_norm": 1.3982007503509521,
+      "learning_rate": 2e-05,
+      "loss": 0.026863,
+      "step": 11325
+    },
+    {
+      "epoch": 22.652,
+      "grad_norm": 1.4211523532867432,
+      "learning_rate": 2e-05,
+      "loss": 0.04486757,
+      "step": 11326
+    },
+    {
+      "epoch": 22.654,
+      "grad_norm": 1.1829328536987305,
+      "learning_rate": 2e-05,
+      "loss": 0.02879838,
+      "step": 11327
+    },
+    {
+      "epoch": 22.656,
+      "grad_norm": 1.7771282196044922,
+      "learning_rate": 2e-05,
+      "loss": 0.03814802,
+      "step": 11328
+    },
+    {
+      "epoch": 22.658,
+      "grad_norm": 2.1755659580230713,
+      "learning_rate": 2e-05,
+      "loss": 0.03225455,
+      "step": 11329
+    },
+    {
+      "epoch": 22.66,
+      "grad_norm": 1.3810690641403198,
+      "learning_rate": 2e-05,
+      "loss": 0.03953212,
+      "step": 11330
+    },
+    {
+      "epoch": 22.662,
+      "grad_norm": 1.7033722400665283,
+      "learning_rate": 2e-05,
+      "loss": 0.04444302,
+      "step": 11331
+    },
+    {
+      "epoch": 22.664,
+      "grad_norm": 2.3877339363098145,
+      "learning_rate": 2e-05,
+      "loss": 0.04218495,
+      "step": 11332
+    },
+    {
+      "epoch": 22.666,
+      "grad_norm": 1.8948878049850464,
+      "learning_rate": 2e-05,
+      "loss": 0.05076663,
+      "step": 11333
+    },
+    {
+      "epoch": 22.668,
+      "grad_norm": 1.181063175201416,
+      "learning_rate": 2e-05,
+      "loss": 0.02133227,
+      "step": 11334
+    },
+    {
+      "epoch": 22.67,
+      "grad_norm": 2.0228304862976074,
+      "learning_rate": 2e-05,
+      "loss": 0.04120112,
+      "step": 11335
+    },
+    {
+      "epoch": 22.672,
+      "grad_norm": 1.0375754833221436,
+      "learning_rate": 2e-05,
+      "loss": 0.03073654,
+      "step": 11336
+    },
+    {
+      "epoch": 22.674,
+      "grad_norm": 2.258004665374756,
+      "learning_rate": 2e-05,
+      "loss": 0.05255279,
+      "step": 11337
+    },
+    {
+      "epoch": 22.676,
+      "grad_norm": 1.752378225326538,
+      "learning_rate": 2e-05,
+      "loss": 0.05190774,
+      "step": 11338
+    },
+    {
+      "epoch": 22.678,
+      "grad_norm": 2.363682746887207,
+      "learning_rate": 2e-05,
+      "loss": 0.04572538,
+      "step": 11339
+    },
+    {
+      "epoch": 22.68,
+      "grad_norm": 1.088084101676941,
+      "learning_rate": 2e-05,
+      "loss": 0.03764492,
+      "step": 11340
+    },
+    {
+      "epoch": 22.682,
+      "grad_norm": 0.8578435182571411,
+      "learning_rate": 2e-05,
+      "loss": 0.02391421,
+      "step": 11341
+    },
+    {
+      "epoch": 22.684,
+      "grad_norm": 1.1827582120895386,
+      "learning_rate": 2e-05,
+      "loss": 0.03264995,
+      "step": 11342
+    },
+    {
+      "epoch": 22.686,
+      "grad_norm": 1.3935420513153076,
+      "learning_rate": 2e-05,
+      "loss": 0.03195801,
+      "step": 11343
+    },
+    {
+      "epoch": 22.688,
+      "grad_norm": 0.9729677438735962,
+      "learning_rate": 2e-05,
+      "loss": 0.02463444,
+      "step": 11344
+    },
+    {
+      "epoch": 22.69,
+      "grad_norm": 3.6322379112243652,
+      "learning_rate": 2e-05,
+      "loss": 0.04378824,
+      "step": 11345
+    },
+    {
+      "epoch": 22.692,
+      "grad_norm": 1.002524971961975,
+      "learning_rate": 2e-05,
+      "loss": 0.02371977,
+      "step": 11346
+    },
+    {
+      "epoch": 22.694,
+      "grad_norm": 1.1274267435073853,
+      "learning_rate": 2e-05,
+      "loss": 0.02771427,
+      "step": 11347
+    },
+    {
+      "epoch": 22.696,
+      "grad_norm": 0.9226782321929932,
+      "learning_rate": 2e-05,
+      "loss": 0.0263548,
+      "step": 11348
+    },
+    {
+      "epoch": 22.698,
+      "grad_norm": 1.6948686838150024,
+      "learning_rate": 2e-05,
+      "loss": 0.03164578,
+      "step": 11349
+    },
+    {
+      "epoch": 22.7,
+      "grad_norm": 1.1010253429412842,
+      "learning_rate": 2e-05,
+      "loss": 0.03652514,
+      "step": 11350
+    },
+    {
+      "epoch": 22.701999999999998,
+      "grad_norm": 1.2103954553604126,
+      "learning_rate": 2e-05,
+      "loss": 0.03659312,
+      "step": 11351
+    },
+    {
+      "epoch": 22.704,
+      "grad_norm": 1.1126821041107178,
+      "learning_rate": 2e-05,
+      "loss": 0.02752934,
+      "step": 11352
+    },
+    {
+      "epoch": 22.706,
+      "grad_norm": 0.9469572305679321,
+      "learning_rate": 2e-05,
+      "loss": 0.02585907,
+      "step": 11353
+    },
+    {
+      "epoch": 22.708,
+      "grad_norm": 1.8642154932022095,
+      "learning_rate": 2e-05,
+      "loss": 0.04325049,
+      "step": 11354
+    },
+    {
+      "epoch": 22.71,
+      "grad_norm": 1.1645115613937378,
+      "learning_rate": 2e-05,
+      "loss": 0.04005027,
+      "step": 11355
+    },
+    {
+      "epoch": 22.712,
+      "grad_norm": 1.520301342010498,
+      "learning_rate": 2e-05,
+      "loss": 0.0399612,
+      "step": 11356
+    },
+    {
+      "epoch": 22.714,
+      "grad_norm": 1.235013723373413,
+      "learning_rate": 2e-05,
+      "loss": 0.03438204,
+      "step": 11357
+    },
+    {
+      "epoch": 22.716,
+      "grad_norm": 0.9834000468254089,
+      "learning_rate": 2e-05,
+      "loss": 0.02958825,
+      "step": 11358
+    },
+    {
+      "epoch": 22.718,
+      "grad_norm": 0.9529434442520142,
+      "learning_rate": 2e-05,
+      "loss": 0.02942239,
+      "step": 11359
+    },
+    {
+      "epoch": 22.72,
+      "grad_norm": 1.0147749185562134,
+      "learning_rate": 2e-05,
+      "loss": 0.02894795,
+      "step": 11360
+    },
+    {
+      "epoch": 22.722,
+      "grad_norm": 2.125749111175537,
+      "learning_rate": 2e-05,
+      "loss": 0.02771318,
+      "step": 11361
+    },
+    {
+      "epoch": 22.724,
+      "grad_norm": 1.1008570194244385,
+      "learning_rate": 2e-05,
+      "loss": 0.0344195,
+      "step": 11362
+    },
+    {
+      "epoch": 22.726,
+      "grad_norm": 2.6095540523529053,
+      "learning_rate": 2e-05,
+      "loss": 0.04947619,
+      "step": 11363
+    },
+    {
+      "epoch": 22.728,
+      "grad_norm": 1.228231430053711,
+      "learning_rate": 2e-05,
+      "loss": 0.03951849,
+      "step": 11364
+    },
+    {
+      "epoch": 22.73,
+      "grad_norm": 3.1176652908325195,
+      "learning_rate": 2e-05,
+      "loss": 0.03356297,
+      "step": 11365
+    },
+    {
+      "epoch": 22.732,
+      "grad_norm": 1.1963483095169067,
+      "learning_rate": 2e-05,
+      "loss": 0.02985599,
+      "step": 11366
+    },
+    {
+      "epoch": 22.734,
+      "grad_norm": 1.533469319343567,
+      "learning_rate": 2e-05,
+      "loss": 0.03753369,
+      "step": 11367
+    },
+    {
+      "epoch": 22.736,
+      "grad_norm": 0.9125386476516724,
+      "learning_rate": 2e-05,
+      "loss": 0.03140819,
+      "step": 11368
+    },
+    {
+      "epoch": 22.738,
+      "grad_norm": 1.361525058746338,
+      "learning_rate": 2e-05,
+      "loss": 0.03976889,
+      "step": 11369
+    },
+    {
+      "epoch": 22.74,
+      "grad_norm": 1.0508217811584473,
+      "learning_rate": 2e-05,
+      "loss": 0.03463693,
+      "step": 11370
+    },
+    {
+      "epoch": 22.742,
+      "grad_norm": 1.6200356483459473,
+      "learning_rate": 2e-05,
+      "loss": 0.04815769,
+      "step": 11371
+    },
+    {
+      "epoch": 22.744,
+      "grad_norm": 1.0508980751037598,
+      "learning_rate": 2e-05,
+      "loss": 0.04034566,
+      "step": 11372
+    },
+    {
+      "epoch": 22.746,
+      "grad_norm": 1.024295687675476,
+      "learning_rate": 2e-05,
+      "loss": 0.03280937,
+      "step": 11373
+    },
+    {
+      "epoch": 22.748,
+      "grad_norm": 1.413459062576294,
+      "learning_rate": 2e-05,
+      "loss": 0.04194808,
+      "step": 11374
+    },
+    {
+      "epoch": 22.75,
+      "grad_norm": 1.0012433528900146,
+      "learning_rate": 2e-05,
+      "loss": 0.02298179,
+      "step": 11375
+    },
+    {
+      "epoch": 22.752,
+      "grad_norm": 2.7343661785125732,
+      "learning_rate": 2e-05,
+      "loss": 0.04880745,
+      "step": 11376
+    },
+    {
+      "epoch": 22.754,
+      "grad_norm": 1.2125567197799683,
+      "learning_rate": 2e-05,
+      "loss": 0.0418051,
+      "step": 11377
+    },
+    {
+      "epoch": 22.756,
+      "grad_norm": 1.162410855293274,
+      "learning_rate": 2e-05,
+      "loss": 0.02900402,
+      "step": 11378
+    },
+    {
+      "epoch": 22.758,
+      "grad_norm": 0.9885851740837097,
+      "learning_rate": 2e-05,
+      "loss": 0.02954095,
+      "step": 11379
+    },
+    {
+      "epoch": 22.76,
+      "grad_norm": 1.2678016424179077,
+      "learning_rate": 2e-05,
+      "loss": 0.04300134,
+      "step": 11380
+    },
+    {
+      "epoch": 22.762,
+      "grad_norm": 1.4912469387054443,
+      "learning_rate": 2e-05,
+      "loss": 0.03353298,
+      "step": 11381
+    },
+    {
+      "epoch": 22.764,
+      "grad_norm": 1.0253992080688477,
+      "learning_rate": 2e-05,
+      "loss": 0.02656168,
+      "step": 11382
+    },
+    {
+      "epoch": 22.766,
+      "grad_norm": 0.8745793104171753,
+      "learning_rate": 2e-05,
+      "loss": 0.02277788,
+      "step": 11383
+    },
+    {
+      "epoch": 22.768,
+      "grad_norm": 1.3656665086746216,
+      "learning_rate": 2e-05,
+      "loss": 0.03005461,
+      "step": 11384
+    },
+    {
+      "epoch": 22.77,
+      "grad_norm": 1.9831993579864502,
+      "learning_rate": 2e-05,
+      "loss": 0.04527285,
+      "step": 11385
+    },
+    {
+      "epoch": 22.772,
+      "grad_norm": 1.8033963441848755,
+      "learning_rate": 2e-05,
+      "loss": 0.05182678,
+      "step": 11386
+    },
+    {
+      "epoch": 22.774,
+      "grad_norm": 1.5675684213638306,
+      "learning_rate": 2e-05,
+      "loss": 0.03885451,
+      "step": 11387
+    },
+    {
+      "epoch": 22.776,
+      "grad_norm": 2.476839065551758,
+      "learning_rate": 2e-05,
+      "loss": 0.05386474,
+      "step": 11388
+    },
+    {
+      "epoch": 22.778,
+      "grad_norm": 0.9441741108894348,
+      "learning_rate": 2e-05,
+      "loss": 0.02855249,
+      "step": 11389
+    },
+    {
+      "epoch": 22.78,
+      "grad_norm": 1.4872889518737793,
+      "learning_rate": 2e-05,
+      "loss": 0.04495686,
+      "step": 11390
+    },
+    {
+      "epoch": 22.782,
+      "grad_norm": 2.2526416778564453,
+      "learning_rate": 2e-05,
+      "loss": 0.04384479,
+      "step": 11391
+    },
+    {
+      "epoch": 22.784,
+      "grad_norm": 1.2508896589279175,
+      "learning_rate": 2e-05,
+      "loss": 0.03676173,
+      "step": 11392
+    },
+    {
+      "epoch": 22.786,
+      "grad_norm": 1.5125991106033325,
+      "learning_rate": 2e-05,
+      "loss": 0.04031794,
+      "step": 11393
+    },
+    {
+      "epoch": 22.788,
+      "grad_norm": 0.8906210660934448,
+      "learning_rate": 2e-05,
+      "loss": 0.02660401,
+      "step": 11394
+    },
+    {
+      "epoch": 22.79,
+      "grad_norm": 3.060696601867676,
+      "learning_rate": 2e-05,
+      "loss": 0.04931217,
+      "step": 11395
+    },
+    {
+      "epoch": 22.792,
+      "grad_norm": 1.2167766094207764,
+      "learning_rate": 2e-05,
+      "loss": 0.0397851,
+      "step": 11396
+    },
+    {
+      "epoch": 22.794,
+      "grad_norm": 1.559085726737976,
+      "learning_rate": 2e-05,
+      "loss": 0.04445662,
+      "step": 11397
+    },
+    {
+      "epoch": 22.796,
+      "grad_norm": 2.269927978515625,
+      "learning_rate": 2e-05,
+      "loss": 0.05920038,
+      "step": 11398
+    },
+    {
+      "epoch": 22.798000000000002,
+      "grad_norm": 1.1354275941848755,
+      "learning_rate": 2e-05,
+      "loss": 0.03149124,
+      "step": 11399
+    },
+    {
+      "epoch": 22.8,
+      "grad_norm": 1.1973793506622314,
+      "learning_rate": 2e-05,
+      "loss": 0.03659835,
+      "step": 11400
+    },
+    {
+      "epoch": 22.802,
+      "grad_norm": 1.178956151008606,
+      "learning_rate": 2e-05,
+      "loss": 0.03625498,
+      "step": 11401
+    },
+    {
+      "epoch": 22.804,
+      "grad_norm": 0.8932549357414246,
+      "learning_rate": 2e-05,
+      "loss": 0.02956768,
+      "step": 11402
+    },
+    {
+      "epoch": 22.806,
+      "grad_norm": 1.0559618473052979,
+      "learning_rate": 2e-05,
+      "loss": 0.03430596,
+      "step": 11403
+    },
+    {
+      "epoch": 22.808,
+      "grad_norm": 2.337083578109741,
+      "learning_rate": 2e-05,
+      "loss": 0.04307479,
+      "step": 11404
+    },
+    {
+      "epoch": 22.81,
+      "grad_norm": 1.271933913230896,
+      "learning_rate": 2e-05,
+      "loss": 0.03103825,
+      "step": 11405
+    },
+    {
+      "epoch": 22.812,
+      "grad_norm": 1.2939378023147583,
+      "learning_rate": 2e-05,
+      "loss": 0.04132862,
+      "step": 11406
+    },
+    {
+      "epoch": 22.814,
+      "grad_norm": 1.2300240993499756,
+      "learning_rate": 2e-05,
+      "loss": 0.03786105,
+      "step": 11407
+    },
+    {
+      "epoch": 22.816,
+      "grad_norm": 1.1687867641448975,
+      "learning_rate": 2e-05,
+      "loss": 0.04594323,
+      "step": 11408
+    },
+    {
+      "epoch": 22.818,
+      "grad_norm": 1.2083779573440552,
+      "learning_rate": 2e-05,
+      "loss": 0.04165336,
+      "step": 11409
+    },
+    {
+      "epoch": 22.82,
+      "grad_norm": 2.3479809761047363,
+      "learning_rate": 2e-05,
+      "loss": 0.0412714,
+      "step": 11410
+    },
+    {
+      "epoch": 22.822,
+      "grad_norm": 1.5195822715759277,
+      "learning_rate": 2e-05,
+      "loss": 0.02944362,
+      "step": 11411
+    },
+    {
+      "epoch": 22.824,
+      "grad_norm": 1.2599931955337524,
+      "learning_rate": 2e-05,
+      "loss": 0.039657,
+      "step": 11412
+    },
+    {
+      "epoch": 22.826,
+      "grad_norm": 1.5169256925582886,
+      "learning_rate": 2e-05,
+      "loss": 0.0313296,
+      "step": 11413
+    },
+    {
+      "epoch": 22.828,
+      "grad_norm": 1.3896204233169556,
+      "learning_rate": 2e-05,
+      "loss": 0.041568,
+      "step": 11414
+    },
+    {
+      "epoch": 22.83,
+      "grad_norm": 1.3151553869247437,
+      "learning_rate": 2e-05,
+      "loss": 0.03976612,
+      "step": 11415
+    },
+    {
+      "epoch": 22.832,
+      "grad_norm": 1.2248992919921875,
+      "learning_rate": 2e-05,
+      "loss": 0.032742,
+      "step": 11416
+    },
+    {
+      "epoch": 22.834,
+      "grad_norm": 1.6058671474456787,
+      "learning_rate": 2e-05,
+      "loss": 0.04475835,
+      "step": 11417
+    },
+    {
+      "epoch": 22.836,
+      "grad_norm": 1.131249189376831,
+      "learning_rate": 2e-05,
+      "loss": 0.03879447,
+      "step": 11418
+    },
+    {
+      "epoch": 22.838,
+      "grad_norm": 1.2209913730621338,
+      "learning_rate": 2e-05,
+      "loss": 0.03018184,
+      "step": 11419
+    },
+    {
+      "epoch": 22.84,
+      "grad_norm": 1.2565336227416992,
+      "learning_rate": 2e-05,
+      "loss": 0.03886155,
+      "step": 11420
+    },
+    {
+      "epoch": 22.842,
+      "grad_norm": 1.2150452136993408,
+      "learning_rate": 2e-05,
+      "loss": 0.03672185,
+      "step": 11421
+    },
+    {
+      "epoch": 22.844,
+      "grad_norm": 1.4771347045898438,
+      "learning_rate": 2e-05,
+      "loss": 0.02761836,
+      "step": 11422
+    },
+    {
+      "epoch": 22.846,
+      "grad_norm": 2.192431926727295,
+      "learning_rate": 2e-05,
+      "loss": 0.0543457,
+      "step": 11423
+    },
+    {
+      "epoch": 22.848,
+      "grad_norm": 1.1383287906646729,
+      "learning_rate": 2e-05,
+      "loss": 0.03061396,
+      "step": 11424
+    },
+    {
+      "epoch": 22.85,
+      "grad_norm": 2.1792497634887695,
+      "learning_rate": 2e-05,
+      "loss": 0.03490089,
+      "step": 11425
+    },
+    {
+      "epoch": 22.852,
+      "grad_norm": 1.2106961011886597,
+      "learning_rate": 2e-05,
+      "loss": 0.03852124,
+      "step": 11426
+    },
+    {
+      "epoch": 22.854,
+      "grad_norm": 4.948497772216797,
+      "learning_rate": 2e-05,
+      "loss": 0.02499816,
+      "step": 11427
+    },
+    {
+      "epoch": 22.856,
+      "grad_norm": 1.3616306781768799,
+      "learning_rate": 2e-05,
+      "loss": 0.04259917,
+      "step": 11428
+    },
+    {
+      "epoch": 22.858,
+      "grad_norm": 1.405735969543457,
+      "learning_rate": 2e-05,
+      "loss": 0.03377235,
+      "step": 11429
+    },
+    {
+      "epoch": 22.86,
+      "grad_norm": 1.426137924194336,
+      "learning_rate": 2e-05,
+      "loss": 0.04095915,
+      "step": 11430
+    },
+    {
+      "epoch": 22.862,
+      "grad_norm": 1.1575813293457031,
+      "learning_rate": 2e-05,
+      "loss": 0.03657633,
+      "step": 11431
+    },
+    {
+      "epoch": 22.864,
+      "grad_norm": 1.048122525215149,
+      "learning_rate": 2e-05,
+      "loss": 0.03505576,
+      "step": 11432
+    },
+    {
+      "epoch": 22.866,
+      "grad_norm": 1.589388132095337,
+      "learning_rate": 2e-05,
+      "loss": 0.02945594,
+      "step": 11433
+    },
+    {
+      "epoch": 22.868,
+      "grad_norm": 1.7383077144622803,
+      "learning_rate": 2e-05,
+      "loss": 0.03888205,
+      "step": 11434
+    },
+    {
+      "epoch": 22.87,
+      "grad_norm": 1.2866278886795044,
+      "learning_rate": 2e-05,
+      "loss": 0.02404101,
+      "step": 11435
+    },
+    {
+      "epoch": 22.872,
+      "grad_norm": 1.8006558418273926,
+      "learning_rate": 2e-05,
+      "loss": 0.02283414,
+      "step": 11436
+    },
+    {
+      "epoch": 22.874,
+      "grad_norm": 0.7845028638839722,
+      "learning_rate": 2e-05,
+      "loss": 0.01730304,
+      "step": 11437
+    },
+    {
+      "epoch": 22.876,
+      "grad_norm": 1.3924261331558228,
+      "learning_rate": 2e-05,
+      "loss": 0.03165732,
+      "step": 11438
+    },
+    {
+      "epoch": 22.878,
+      "grad_norm": 1.0454896688461304,
+      "learning_rate": 2e-05,
+      "loss": 0.03067488,
+      "step": 11439
+    },
+    {
+      "epoch": 22.88,
+      "grad_norm": 1.8691692352294922,
+      "learning_rate": 2e-05,
+      "loss": 0.03286079,
+      "step": 11440
+    },
+    {
+      "epoch": 22.882,
+      "grad_norm": 1.0586732625961304,
+      "learning_rate": 2e-05,
+      "loss": 0.02909458,
+      "step": 11441
+    },
+    {
+      "epoch": 22.884,
+      "grad_norm": 1.525641679763794,
+      "learning_rate": 2e-05,
+      "loss": 0.03261862,
+      "step": 11442
+    },
+    {
+      "epoch": 22.886,
+      "grad_norm": 1.164318323135376,
+      "learning_rate": 2e-05,
+      "loss": 0.02912915,
+      "step": 11443
+    },
+    {
+      "epoch": 22.888,
+      "grad_norm": 1.5031243562698364,
+      "learning_rate": 2e-05,
+      "loss": 0.03876949,
+      "step": 11444
+    },
+    {
+      "epoch": 22.89,
+      "grad_norm": 1.0397024154663086,
+      "learning_rate": 2e-05,
+      "loss": 0.02776971,
+      "step": 11445
+    },
+    {
+      "epoch": 22.892,
+      "grad_norm": 1.2293657064437866,
+      "learning_rate": 2e-05,
+      "loss": 0.04357108,
+      "step": 11446
+    },
+    {
+      "epoch": 22.894,
+      "grad_norm": 1.8432687520980835,
+      "learning_rate": 2e-05,
+      "loss": 0.02928544,
+      "step": 11447
+    },
+    {
+      "epoch": 22.896,
+      "grad_norm": 1.107912540435791,
+      "learning_rate": 2e-05,
+      "loss": 0.02876203,
+      "step": 11448
+    },
+    {
+      "epoch": 22.898,
+      "grad_norm": 1.5207486152648926,
+      "learning_rate": 2e-05,
+      "loss": 0.04288258,
+      "step": 11449
+    },
+    {
+      "epoch": 22.9,
+      "grad_norm": 1.0726830959320068,
+      "learning_rate": 2e-05,
+      "loss": 0.03220513,
+      "step": 11450
+    },
+    {
+      "epoch": 22.902,
+      "grad_norm": 1.3904377222061157,
+      "learning_rate": 2e-05,
+      "loss": 0.03931183,
+      "step": 11451
+    },
+    {
+      "epoch": 22.904,
+      "grad_norm": 1.601686954498291,
+      "learning_rate": 2e-05,
+      "loss": 0.04329685,
+      "step": 11452
+    },
+    {
+      "epoch": 22.906,
+      "grad_norm": 1.7275676727294922,
+      "learning_rate": 2e-05,
+      "loss": 0.03662112,
+      "step": 11453
+    },
+    {
+      "epoch": 22.908,
+      "grad_norm": 0.9452254176139832,
+      "learning_rate": 2e-05,
+      "loss": 0.03591129,
+      "step": 11454
+    },
+    {
+      "epoch": 22.91,
+      "grad_norm": 0.9925097823143005,
+      "learning_rate": 2e-05,
+      "loss": 0.02755562,
+      "step": 11455
+    },
+    {
+      "epoch": 22.912,
+      "grad_norm": 1.9532277584075928,
+      "learning_rate": 2e-05,
+      "loss": 0.03555189,
+      "step": 11456
+    },
+    {
+      "epoch": 22.914,
+      "grad_norm": 0.8371837139129639,
+      "learning_rate": 2e-05,
+      "loss": 0.02261299,
+      "step": 11457
+    },
+    {
+      "epoch": 22.916,
+      "grad_norm": 2.2544491291046143,
+      "learning_rate": 2e-05,
+      "loss": 0.05361641,
+      "step": 11458
+    },
+    {
+      "epoch": 22.918,
+      "grad_norm": 2.455197334289551,
+      "learning_rate": 2e-05,
+      "loss": 0.04926509,
+      "step": 11459
+    },
+    {
+      "epoch": 22.92,
+      "grad_norm": 1.0468469858169556,
+      "learning_rate": 2e-05,
+      "loss": 0.03623208,
+      "step": 11460
+    },
+    {
+      "epoch": 22.922,
+      "grad_norm": 1.6605795621871948,
+      "learning_rate": 2e-05,
+      "loss": 0.03582647,
+      "step": 11461
+    },
+    {
+      "epoch": 22.924,
+      "grad_norm": 1.6512527465820312,
+      "learning_rate": 2e-05,
+      "loss": 0.03864128,
+      "step": 11462
+    },
+    {
+      "epoch": 22.926,
+      "grad_norm": 1.015683889389038,
+      "learning_rate": 2e-05,
+      "loss": 0.02952475,
+      "step": 11463
+    },
+    {
+      "epoch": 22.928,
+      "grad_norm": 2.490291118621826,
+      "learning_rate": 2e-05,
+      "loss": 0.04171167,
+      "step": 11464
+    },
+    {
+      "epoch": 22.93,
+      "grad_norm": 1.092077612876892,
+      "learning_rate": 2e-05,
+      "loss": 0.03160276,
+      "step": 11465
+    },
+    {
+      "epoch": 22.932,
+      "grad_norm": 1.0306111574172974,
+      "learning_rate": 2e-05,
+      "loss": 0.02688064,
+      "step": 11466
+    },
+    {
+      "epoch": 22.934,
+      "grad_norm": 1.20967435836792,
+      "learning_rate": 2e-05,
+      "loss": 0.03792361,
+      "step": 11467
+    },
+    {
+      "epoch": 22.936,
+      "grad_norm": 1.6543128490447998,
+      "learning_rate": 2e-05,
+      "loss": 0.0479852,
+      "step": 11468
+    },
+    {
+      "epoch": 22.938,
+      "grad_norm": 1.193781852722168,
+      "learning_rate": 2e-05,
+      "loss": 0.03420633,
+      "step": 11469
+    },
+    {
+      "epoch": 22.94,
+      "grad_norm": 1.764358639717102,
+      "learning_rate": 2e-05,
+      "loss": 0.03625751,
+      "step": 11470
+    },
+    {
+      "epoch": 22.942,
+      "grad_norm": 1.6757545471191406,
+      "learning_rate": 2e-05,
+      "loss": 0.03107898,
+      "step": 11471
+    },
+    {
+      "epoch": 22.944,
+      "grad_norm": 0.9526821970939636,
+      "learning_rate": 2e-05,
+      "loss": 0.03141288,
+      "step": 11472
+    },
+    {
+      "epoch": 22.946,
+      "grad_norm": 1.6071827411651611,
+      "learning_rate": 2e-05,
+      "loss": 0.03736177,
+      "step": 11473
+    },
+    {
+      "epoch": 22.948,
+      "grad_norm": 1.1540484428405762,
+      "learning_rate": 2e-05,
+      "loss": 0.03194441,
+      "step": 11474
+    },
+    {
+      "epoch": 22.95,
+      "grad_norm": 1.1837899684906006,
+      "learning_rate": 2e-05,
+      "loss": 0.0288153,
+      "step": 11475
+    },
+    {
+      "epoch": 22.951999999999998,
+      "grad_norm": 1.1613291501998901,
+      "learning_rate": 2e-05,
+      "loss": 0.03203413,
+      "step": 11476
+    },
+    {
+      "epoch": 22.954,
+      "grad_norm": 2.0555005073547363,
+      "learning_rate": 2e-05,
+      "loss": 0.04438991,
+      "step": 11477
+    },
+    {
+      "epoch": 22.956,
+      "grad_norm": 1.4944390058517456,
+      "learning_rate": 2e-05,
+      "loss": 0.05310993,
+      "step": 11478
+    },
+    {
+      "epoch": 22.958,
+      "grad_norm": 1.4034525156021118,
+      "learning_rate": 2e-05,
+      "loss": 0.0385785,
+      "step": 11479
+    },
+    {
+      "epoch": 22.96,
+      "grad_norm": 2.2362515926361084,
+      "learning_rate": 2e-05,
+      "loss": 0.03343879,
+      "step": 11480
+    },
+    {
+      "epoch": 22.962,
+      "grad_norm": 1.9142746925354004,
+      "learning_rate": 2e-05,
+      "loss": 0.04086325,
+      "step": 11481
+    },
+    {
+      "epoch": 22.964,
+      "grad_norm": 3.3769659996032715,
+      "learning_rate": 2e-05,
+      "loss": 0.03347282,
+      "step": 11482
+    },
+    {
+      "epoch": 22.966,
+      "grad_norm": 1.3966118097305298,
+      "learning_rate": 2e-05,
+      "loss": 0.02449038,
+      "step": 11483
+    },
+    {
+      "epoch": 22.968,
+      "grad_norm": 1.6454405784606934,
+      "learning_rate": 2e-05,
+      "loss": 0.03337567,
+      "step": 11484
+    },
+    {
+      "epoch": 22.97,
+      "grad_norm": 1.2182114124298096,
+      "learning_rate": 2e-05,
+      "loss": 0.02583343,
+      "step": 11485
+    },
+    {
+      "epoch": 22.972,
+      "grad_norm": 2.4437437057495117,
+      "learning_rate": 2e-05,
+      "loss": 0.04662333,
+      "step": 11486
+    },
+    {
+      "epoch": 22.974,
+      "grad_norm": 1.3717036247253418,
+      "learning_rate": 2e-05,
+      "loss": 0.04354615,
+      "step": 11487
+    },
+    {
+      "epoch": 22.976,
+      "grad_norm": 1.0817804336547852,
+      "learning_rate": 2e-05,
+      "loss": 0.03474058,
+      "step": 11488
+    },
+    {
+      "epoch": 22.978,
+      "grad_norm": 1.8078384399414062,
+      "learning_rate": 2e-05,
+      "loss": 0.04659226,
+      "step": 11489
+    },
+    {
+      "epoch": 22.98,
+      "grad_norm": 1.3917157649993896,
+      "learning_rate": 2e-05,
+      "loss": 0.04078533,
+      "step": 11490
+    },
+    {
+      "epoch": 22.982,
+      "grad_norm": 0.8220876455307007,
+      "learning_rate": 2e-05,
+      "loss": 0.02258881,
+      "step": 11491
+    },
+    {
+      "epoch": 22.984,
+      "grad_norm": 1.2913342714309692,
+      "learning_rate": 2e-05,
+      "loss": 0.04604527,
+      "step": 11492
+    },
+    {
+      "epoch": 22.986,
+      "grad_norm": 1.271330714225769,
+      "learning_rate": 2e-05,
+      "loss": 0.03667264,
+      "step": 11493
+    },
+    {
+      "epoch": 22.988,
+      "grad_norm": 1.3295248746871948,
+      "learning_rate": 2e-05,
+      "loss": 0.03829803,
+      "step": 11494
+    },
+    {
+      "epoch": 22.99,
+      "grad_norm": 1.3785157203674316,
+      "learning_rate": 2e-05,
+      "loss": 0.04370276,
+      "step": 11495
+    },
+    {
+      "epoch": 22.992,
+      "grad_norm": 1.7691086530685425,
+      "learning_rate": 2e-05,
+      "loss": 0.04192688,
+      "step": 11496
+    },
+    {
+      "epoch": 22.994,
+      "grad_norm": 0.933940052986145,
+      "learning_rate": 2e-05,
+      "loss": 0.01630483,
+      "step": 11497
+    },
+    {
+      "epoch": 22.996,
+      "grad_norm": 1.4123674631118774,
+      "learning_rate": 2e-05,
+      "loss": 0.03609384,
+      "step": 11498
+    },
+    {
+      "epoch": 22.998,
+      "grad_norm": 1.3186585903167725,
+      "learning_rate": 2e-05,
+      "loss": 0.04399614,
+      "step": 11499
+    },
+    {
+      "epoch": 23.0,
+      "grad_norm": 0.9407261610031128,
+      "learning_rate": 2e-05,
+      "loss": 0.02819065,
+      "step": 11500
+    },
+    {
+      "epoch": 23.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9780439121756487,
+        "Equal_1": 0.998,
+        "Equal_2": 0.9780439121756487,
+        "Equal_3": 0.9760479041916168,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 1.0,
+        "Parallel_1": 0.9899799599198397,
+        "Parallel_2": 0.9919839679358717,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.992,
+        "Perpendicular_3": 0.8386773547094188,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9956666666666667,
+        "PointLiesOnCircle_3": 0.9896666666666667,
+        "PointLiesOnLine_1": 0.9959919839679359,
+        "PointLiesOnLine_2": 0.9939879759519038,
+        "PointLiesOnLine_3": 0.9800399201596807
+      },
+      "eval_runtime": 319.8554,
+      "eval_samples_per_second": 32.827,
+      "eval_steps_per_second": 0.657,
+      "step": 11500
+    },
+    {
+      "epoch": 23.002,
+      "grad_norm": 1.2540278434753418,
+      "learning_rate": 2e-05,
+      "loss": 0.05130012,
+      "step": 11501
+    },
+    {
+      "epoch": 23.004,
+      "grad_norm": 1.3764632940292358,
+      "learning_rate": 2e-05,
+      "loss": 0.04294942,
+      "step": 11502
+    },
+    {
+      "epoch": 23.006,
+      "grad_norm": 1.6750069856643677,
+      "learning_rate": 2e-05,
+      "loss": 0.05229911,
+      "step": 11503
+    },
+    {
+      "epoch": 23.008,
+      "grad_norm": 1.1598528623580933,
+      "learning_rate": 2e-05,
+      "loss": 0.04085714,
+      "step": 11504
+    },
+    {
+      "epoch": 23.01,
+      "grad_norm": 1.1853110790252686,
+      "learning_rate": 2e-05,
+      "loss": 0.03996443,
+      "step": 11505
+    },
+    {
+      "epoch": 23.012,
+      "grad_norm": 1.4552375078201294,
+      "learning_rate": 2e-05,
+      "loss": 0.05250798,
+      "step": 11506
+    },
+    {
+      "epoch": 23.014,
+      "grad_norm": 1.4703574180603027,
+      "learning_rate": 2e-05,
+      "loss": 0.05003911,
+      "step": 11507
+    },
+    {
+      "epoch": 23.016,
+      "grad_norm": 1.011425256729126,
+      "learning_rate": 2e-05,
+      "loss": 0.03787163,
+      "step": 11508
+    },
+    {
+      "epoch": 23.018,
+      "grad_norm": 0.9965170621871948,
+      "learning_rate": 2e-05,
+      "loss": 0.04001056,
+      "step": 11509
+    },
+    {
+      "epoch": 23.02,
+      "grad_norm": 1.442737102508545,
+      "learning_rate": 2e-05,
+      "loss": 0.03976019,
+      "step": 11510
+    },
+    {
+      "epoch": 23.022,
+      "grad_norm": 1.067285418510437,
+      "learning_rate": 2e-05,
+      "loss": 0.03311403,
+      "step": 11511
+    },
+    {
+      "epoch": 23.024,
+      "grad_norm": 1.3310978412628174,
+      "learning_rate": 2e-05,
+      "loss": 0.05192403,
+      "step": 11512
+    },
+    {
+      "epoch": 23.026,
+      "grad_norm": 2.071756601333618,
+      "learning_rate": 2e-05,
+      "loss": 0.04093051,
+      "step": 11513
+    },
+    {
+      "epoch": 23.028,
+      "grad_norm": 1.8485080003738403,
+      "learning_rate": 2e-05,
+      "loss": 0.04342106,
+      "step": 11514
+    },
+    {
+      "epoch": 23.03,
+      "grad_norm": 1.2285382747650146,
+      "learning_rate": 2e-05,
+      "loss": 0.03801531,
+      "step": 11515
+    },
+    {
+      "epoch": 23.032,
+      "grad_norm": 1.2549402713775635,
+      "learning_rate": 2e-05,
+      "loss": 0.04710529,
+      "step": 11516
+    },
+    {
+      "epoch": 23.034,
+      "grad_norm": 1.2575291395187378,
+      "learning_rate": 2e-05,
+      "loss": 0.04867338,
+      "step": 11517
+    },
+    {
+      "epoch": 23.036,
+      "grad_norm": 1.4146618843078613,
+      "learning_rate": 2e-05,
+      "loss": 0.04894332,
+      "step": 11518
+    },
+    {
+      "epoch": 23.038,
+      "grad_norm": 1.2756978273391724,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 11519
+    },
+    {
+      "epoch": 23.04,
+      "grad_norm": 1.3770201206207275,
+      "learning_rate": 2e-05,
+      "loss": 0.05650711,
+      "step": 11520
+    },
+    {
+      "epoch": 23.042,
+      "grad_norm": 1.5191999673843384,
+      "learning_rate": 2e-05,
+      "loss": 0.05803034,
+      "step": 11521
+    },
+    {
+      "epoch": 23.044,
+      "grad_norm": 1.6032277345657349,
+      "learning_rate": 2e-05,
+      "loss": 0.04611324,
+      "step": 11522
+    },
+    {
+      "epoch": 23.046,
+      "grad_norm": 1.0860769748687744,
+      "learning_rate": 2e-05,
+      "loss": 0.03498707,
+      "step": 11523
+    },
+    {
+      "epoch": 23.048,
+      "grad_norm": 1.0449998378753662,
+      "learning_rate": 2e-05,
+      "loss": 0.03648846,
+      "step": 11524
+    },
+    {
+      "epoch": 23.05,
+      "grad_norm": 2.3625895977020264,
+      "learning_rate": 2e-05,
+      "loss": 0.03183797,
+      "step": 11525
+    },
+    {
+      "epoch": 23.052,
+      "grad_norm": 1.3209874629974365,
+      "learning_rate": 2e-05,
+      "loss": 0.0523079,
+      "step": 11526
+    },
+    {
+      "epoch": 23.054,
+      "grad_norm": 0.8777402639389038,
+      "learning_rate": 2e-05,
+      "loss": 0.02545576,
+      "step": 11527
+    },
+    {
+      "epoch": 23.056,
+      "grad_norm": 1.1120661497116089,
+      "learning_rate": 2e-05,
+      "loss": 0.0336577,
+      "step": 11528
+    },
+    {
+      "epoch": 23.058,
+      "grad_norm": 1.3430697917938232,
+      "learning_rate": 2e-05,
+      "loss": 0.03209985,
+      "step": 11529
+    },
+    {
+      "epoch": 23.06,
+      "grad_norm": 1.346922755241394,
+      "learning_rate": 2e-05,
+      "loss": 0.02798453,
+      "step": 11530
+    },
+    {
+      "epoch": 23.062,
+      "grad_norm": 1.8189756870269775,
+      "learning_rate": 2e-05,
+      "loss": 0.05564386,
+      "step": 11531
+    },
+    {
+      "epoch": 23.064,
+      "grad_norm": 3.2581892013549805,
+      "learning_rate": 2e-05,
+      "loss": 0.04301608,
+      "step": 11532
+    },
+    {
+      "epoch": 23.066,
+      "grad_norm": 1.3685656785964966,
+      "learning_rate": 2e-05,
+      "loss": 0.03543826,
+      "step": 11533
+    },
+    {
+      "epoch": 23.068,
+      "grad_norm": 1.6216603517532349,
+      "learning_rate": 2e-05,
+      "loss": 0.04750851,
+      "step": 11534
+    },
+    {
+      "epoch": 23.07,
+      "grad_norm": 2.4956471920013428,
+      "learning_rate": 2e-05,
+      "loss": 0.05645571,
+      "step": 11535
+    },
+    {
+      "epoch": 23.072,
+      "grad_norm": 3.800372362136841,
+      "learning_rate": 2e-05,
+      "loss": 0.03642789,
+      "step": 11536
+    },
+    {
+      "epoch": 23.074,
+      "grad_norm": 1.135061502456665,
+      "learning_rate": 2e-05,
+      "loss": 0.03842633,
+      "step": 11537
+    },
+    {
+      "epoch": 23.076,
+      "grad_norm": 1.3326822519302368,
+      "learning_rate": 2e-05,
+      "loss": 0.05297691,
+      "step": 11538
+    },
+    {
+      "epoch": 23.078,
+      "grad_norm": 1.7269304990768433,
+      "learning_rate": 2e-05,
+      "loss": 0.04963746,
+      "step": 11539
+    },
+    {
+      "epoch": 23.08,
+      "grad_norm": 1.300743818283081,
+      "learning_rate": 2e-05,
+      "loss": 0.04434649,
+      "step": 11540
+    },
+    {
+      "epoch": 23.082,
+      "grad_norm": 2.736264705657959,
+      "learning_rate": 2e-05,
+      "loss": 0.05283449,
+      "step": 11541
+    },
+    {
+      "epoch": 23.084,
+      "grad_norm": 1.9172672033309937,
+      "learning_rate": 2e-05,
+      "loss": 0.05265498,
+      "step": 11542
+    },
+    {
+      "epoch": 23.086,
+      "grad_norm": 1.1808491945266724,
+      "learning_rate": 2e-05,
+      "loss": 0.0372861,
+      "step": 11543
+    },
+    {
+      "epoch": 23.088,
+      "grad_norm": 1.2032355070114136,
+      "learning_rate": 2e-05,
+      "loss": 0.04825965,
+      "step": 11544
+    },
+    {
+      "epoch": 23.09,
+      "grad_norm": 1.292203426361084,
+      "learning_rate": 2e-05,
+      "loss": 0.04288067,
+      "step": 11545
+    },
+    {
+      "epoch": 23.092,
+      "grad_norm": 1.23992919921875,
+      "learning_rate": 2e-05,
+      "loss": 0.03847829,
+      "step": 11546
+    },
+    {
+      "epoch": 23.094,
+      "grad_norm": 1.6928577423095703,
+      "learning_rate": 2e-05,
+      "loss": 0.05047772,
+      "step": 11547
+    },
+    {
+      "epoch": 23.096,
+      "grad_norm": 1.2044388055801392,
+      "learning_rate": 2e-05,
+      "loss": 0.04760785,
+      "step": 11548
+    },
+    {
+      "epoch": 23.098,
+      "grad_norm": 4.208912372589111,
+      "learning_rate": 2e-05,
+      "loss": 0.04748537,
+      "step": 11549
+    },
+    {
+      "epoch": 23.1,
+      "grad_norm": 1.1794403791427612,
+      "learning_rate": 2e-05,
+      "loss": 0.03520359,
+      "step": 11550
+    },
+    {
+      "epoch": 23.102,
+      "grad_norm": 2.432137966156006,
+      "learning_rate": 2e-05,
+      "loss": 0.05768427,
+      "step": 11551
+    },
+    {
+      "epoch": 23.104,
+      "grad_norm": 1.1416656970977783,
+      "learning_rate": 2e-05,
+      "loss": 0.04104606,
+      "step": 11552
+    },
+    {
+      "epoch": 23.106,
+      "grad_norm": 0.929355263710022,
+      "learning_rate": 2e-05,
+      "loss": 0.0206061,
+      "step": 11553
+    },
+    {
+      "epoch": 23.108,
+      "grad_norm": 1.0629310607910156,
+      "learning_rate": 2e-05,
+      "loss": 0.02847815,
+      "step": 11554
+    },
+    {
+      "epoch": 23.11,
+      "grad_norm": 2.062800168991089,
+      "learning_rate": 2e-05,
+      "loss": 0.043769,
+      "step": 11555
+    },
+    {
+      "epoch": 23.112,
+      "grad_norm": 1.8286502361297607,
+      "learning_rate": 2e-05,
+      "loss": 0.05828995,
+      "step": 11556
+    },
+    {
+      "epoch": 23.114,
+      "grad_norm": 1.8750455379486084,
+      "learning_rate": 2e-05,
+      "loss": 0.04257144,
+      "step": 11557
+    },
+    {
+      "epoch": 23.116,
+      "grad_norm": 1.3888318538665771,
+      "learning_rate": 2e-05,
+      "loss": 0.04705369,
+      "step": 11558
+    },
+    {
+      "epoch": 23.118,
+      "grad_norm": 1.7153780460357666,
+      "learning_rate": 2e-05,
+      "loss": 0.0691988,
+      "step": 11559
+    },
+    {
+      "epoch": 23.12,
+      "grad_norm": 1.4313924312591553,
+      "learning_rate": 2e-05,
+      "loss": 0.04791881,
+      "step": 11560
+    },
+    {
+      "epoch": 23.122,
+      "grad_norm": 1.0603506565093994,
+      "learning_rate": 2e-05,
+      "loss": 0.0366033,
+      "step": 11561
+    },
+    {
+      "epoch": 23.124,
+      "grad_norm": 1.405933141708374,
+      "learning_rate": 2e-05,
+      "loss": 0.03930571,
+      "step": 11562
+    },
+    {
+      "epoch": 23.126,
+      "grad_norm": 1.920100212097168,
+      "learning_rate": 2e-05,
+      "loss": 0.04094848,
+      "step": 11563
+    },
+    {
+      "epoch": 23.128,
+      "grad_norm": 5.326123237609863,
+      "learning_rate": 2e-05,
+      "loss": 0.06398241,
+      "step": 11564
+    },
+    {
+      "epoch": 23.13,
+      "grad_norm": 1.9219155311584473,
+      "learning_rate": 2e-05,
+      "loss": 0.02961898,
+      "step": 11565
+    },
+    {
+      "epoch": 23.132,
+      "grad_norm": 2.914210081100464,
+      "learning_rate": 2e-05,
+      "loss": 0.04696817,
+      "step": 11566
+    },
+    {
+      "epoch": 23.134,
+      "grad_norm": 0.8635774254798889,
+      "learning_rate": 2e-05,
+      "loss": 0.02733534,
+      "step": 11567
+    },
+    {
+      "epoch": 23.136,
+      "grad_norm": 1.496962308883667,
+      "learning_rate": 2e-05,
+      "loss": 0.03688932,
+      "step": 11568
+    },
+    {
+      "epoch": 23.138,
+      "grad_norm": 2.043313980102539,
+      "learning_rate": 2e-05,
+      "loss": 0.03681432,
+      "step": 11569
+    },
+    {
+      "epoch": 23.14,
+      "grad_norm": 1.5973297357559204,
+      "learning_rate": 2e-05,
+      "loss": 0.04360689,
+      "step": 11570
+    },
+    {
+      "epoch": 23.142,
+      "grad_norm": 1.3972605466842651,
+      "learning_rate": 2e-05,
+      "loss": 0.04391249,
+      "step": 11571
+    },
+    {
+      "epoch": 23.144,
+      "grad_norm": 1.3387508392333984,
+      "learning_rate": 2e-05,
+      "loss": 0.04608027,
+      "step": 11572
+    },
+    {
+      "epoch": 23.146,
+      "grad_norm": 2.011808395385742,
+      "learning_rate": 2e-05,
+      "loss": 0.05625965,
+      "step": 11573
+    },
+    {
+      "epoch": 23.148,
+      "grad_norm": 1.5013598203659058,
+      "learning_rate": 2e-05,
+      "loss": 0.05085777,
+      "step": 11574
+    },
+    {
+      "epoch": 23.15,
+      "grad_norm": 2.186159133911133,
+      "learning_rate": 2e-05,
+      "loss": 0.04711618,
+      "step": 11575
+    },
+    {
+      "epoch": 23.152,
+      "grad_norm": 2.2853493690490723,
+      "learning_rate": 2e-05,
+      "loss": 0.0342935,
+      "step": 11576
+    },
+    {
+      "epoch": 23.154,
+      "grad_norm": 1.3616294860839844,
+      "learning_rate": 2e-05,
+      "loss": 0.04885361,
+      "step": 11577
+    },
+    {
+      "epoch": 23.156,
+      "grad_norm": 1.2394838333129883,
+      "learning_rate": 2e-05,
+      "loss": 0.03285236,
+      "step": 11578
+    },
+    {
+      "epoch": 23.158,
+      "grad_norm": 1.339077353477478,
+      "learning_rate": 2e-05,
+      "loss": 0.02606569,
+      "step": 11579
+    },
+    {
+      "epoch": 23.16,
+      "grad_norm": 0.8776892423629761,
+      "learning_rate": 2e-05,
+      "loss": 0.02158548,
+      "step": 11580
+    },
+    {
+      "epoch": 23.162,
+      "grad_norm": 3.6260955333709717,
+      "learning_rate": 2e-05,
+      "loss": 0.05039684,
+      "step": 11581
+    },
+    {
+      "epoch": 23.164,
+      "grad_norm": 1.242028832435608,
+      "learning_rate": 2e-05,
+      "loss": 0.04099006,
+      "step": 11582
+    },
+    {
+      "epoch": 23.166,
+      "grad_norm": 1.870340347290039,
+      "learning_rate": 2e-05,
+      "loss": 0.05608212,
+      "step": 11583
+    },
+    {
+      "epoch": 23.168,
+      "grad_norm": 1.346118688583374,
+      "learning_rate": 2e-05,
+      "loss": 0.03667539,
+      "step": 11584
+    },
+    {
+      "epoch": 23.17,
+      "grad_norm": 1.3202615976333618,
+      "learning_rate": 2e-05,
+      "loss": 0.06091684,
+      "step": 11585
+    },
+    {
+      "epoch": 23.172,
+      "grad_norm": 1.252225399017334,
+      "learning_rate": 2e-05,
+      "loss": 0.0603791,
+      "step": 11586
+    },
+    {
+      "epoch": 23.174,
+      "grad_norm": 1.584975242614746,
+      "learning_rate": 2e-05,
+      "loss": 0.04847322,
+      "step": 11587
+    },
+    {
+      "epoch": 23.176,
+      "grad_norm": 1.3901718854904175,
+      "learning_rate": 2e-05,
+      "loss": 0.03669205,
+      "step": 11588
+    },
+    {
+      "epoch": 23.178,
+      "grad_norm": 1.0474369525909424,
+      "learning_rate": 2e-05,
+      "loss": 0.04350015,
+      "step": 11589
+    },
+    {
+      "epoch": 23.18,
+      "grad_norm": 1.0090280771255493,
+      "learning_rate": 2e-05,
+      "loss": 0.03717218,
+      "step": 11590
+    },
+    {
+      "epoch": 23.182,
+      "grad_norm": 1.004239797592163,
+      "learning_rate": 2e-05,
+      "loss": 0.02906239,
+      "step": 11591
+    },
+    {
+      "epoch": 23.184,
+      "grad_norm": 1.3345508575439453,
+      "learning_rate": 2e-05,
+      "loss": 0.03678144,
+      "step": 11592
+    },
+    {
+      "epoch": 23.186,
+      "grad_norm": 1.837895393371582,
+      "learning_rate": 2e-05,
+      "loss": 0.03756856,
+      "step": 11593
+    },
+    {
+      "epoch": 23.188,
+      "grad_norm": 1.3044272661209106,
+      "learning_rate": 2e-05,
+      "loss": 0.04445122,
+      "step": 11594
+    },
+    {
+      "epoch": 23.19,
+      "grad_norm": 1.1923726797103882,
+      "learning_rate": 2e-05,
+      "loss": 0.03589686,
+      "step": 11595
+    },
+    {
+      "epoch": 23.192,
+      "grad_norm": 1.0984914302825928,
+      "learning_rate": 2e-05,
+      "loss": 0.04005678,
+      "step": 11596
+    },
+    {
+      "epoch": 23.194,
+      "grad_norm": 1.199487566947937,
+      "learning_rate": 2e-05,
+      "loss": 0.03045943,
+      "step": 11597
+    },
+    {
+      "epoch": 23.196,
+      "grad_norm": 1.032716989517212,
+      "learning_rate": 2e-05,
+      "loss": 0.03927022,
+      "step": 11598
+    },
+    {
+      "epoch": 23.198,
+      "grad_norm": 2.102476119995117,
+      "learning_rate": 2e-05,
+      "loss": 0.06727884,
+      "step": 11599
+    },
+    {
+      "epoch": 23.2,
+      "grad_norm": 1.172778844833374,
+      "learning_rate": 2e-05,
+      "loss": 0.05053917,
+      "step": 11600
+    },
+    {
+      "epoch": 23.202,
+      "grad_norm": 1.736275315284729,
+      "learning_rate": 2e-05,
+      "loss": 0.03909831,
+      "step": 11601
+    },
+    {
+      "epoch": 23.204,
+      "grad_norm": 1.0708051919937134,
+      "learning_rate": 2e-05,
+      "loss": 0.03059747,
+      "step": 11602
+    },
+    {
+      "epoch": 23.206,
+      "grad_norm": 1.1244726181030273,
+      "learning_rate": 2e-05,
+      "loss": 0.04023635,
+      "step": 11603
+    },
+    {
+      "epoch": 23.208,
+      "grad_norm": 0.9452853202819824,
+      "learning_rate": 2e-05,
+      "loss": 0.03030682,
+      "step": 11604
+    },
+    {
+      "epoch": 23.21,
+      "grad_norm": 2.1545941829681396,
+      "learning_rate": 2e-05,
+      "loss": 0.04214806,
+      "step": 11605
+    },
+    {
+      "epoch": 23.212,
+      "grad_norm": 1.2645689249038696,
+      "learning_rate": 2e-05,
+      "loss": 0.03720896,
+      "step": 11606
+    },
+    {
+      "epoch": 23.214,
+      "grad_norm": 1.2697169780731201,
+      "learning_rate": 2e-05,
+      "loss": 0.04333124,
+      "step": 11607
+    },
+    {
+      "epoch": 23.216,
+      "grad_norm": 1.4149627685546875,
+      "learning_rate": 2e-05,
+      "loss": 0.05068656,
+      "step": 11608
+    },
+    {
+      "epoch": 23.218,
+      "grad_norm": 2.079878568649292,
+      "learning_rate": 2e-05,
+      "loss": 0.06022862,
+      "step": 11609
+    },
+    {
+      "epoch": 23.22,
+      "grad_norm": 1.221092700958252,
+      "learning_rate": 2e-05,
+      "loss": 0.02790497,
+      "step": 11610
+    },
+    {
+      "epoch": 23.222,
+      "grad_norm": 0.9675205945968628,
+      "learning_rate": 2e-05,
+      "loss": 0.03455825,
+      "step": 11611
+    },
+    {
+      "epoch": 23.224,
+      "grad_norm": 0.8989473581314087,
+      "learning_rate": 2e-05,
+      "loss": 0.02922,
+      "step": 11612
+    },
+    {
+      "epoch": 23.226,
+      "grad_norm": 1.088036060333252,
+      "learning_rate": 2e-05,
+      "loss": 0.05001824,
+      "step": 11613
+    },
+    {
+      "epoch": 23.228,
+      "grad_norm": 1.4129053354263306,
+      "learning_rate": 2e-05,
+      "loss": 0.04396573,
+      "step": 11614
+    },
+    {
+      "epoch": 23.23,
+      "grad_norm": 1.193305253982544,
+      "learning_rate": 2e-05,
+      "loss": 0.03746897,
+      "step": 11615
+    },
+    {
+      "epoch": 23.232,
+      "grad_norm": 1.3615189790725708,
+      "learning_rate": 2e-05,
+      "loss": 0.03328097,
+      "step": 11616
+    },
+    {
+      "epoch": 23.234,
+      "grad_norm": 3.6548335552215576,
+      "learning_rate": 2e-05,
+      "loss": 0.0415346,
+      "step": 11617
+    },
+    {
+      "epoch": 23.236,
+      "grad_norm": 1.218424677848816,
+      "learning_rate": 2e-05,
+      "loss": 0.02998128,
+      "step": 11618
+    },
+    {
+      "epoch": 23.238,
+      "grad_norm": 1.3885092735290527,
+      "learning_rate": 2e-05,
+      "loss": 0.04785194,
+      "step": 11619
+    },
+    {
+      "epoch": 23.24,
+      "grad_norm": 2.74619722366333,
+      "learning_rate": 2e-05,
+      "loss": 0.04847107,
+      "step": 11620
+    },
+    {
+      "epoch": 23.242,
+      "grad_norm": 1.482759952545166,
+      "learning_rate": 2e-05,
+      "loss": 0.04352796,
+      "step": 11621
+    },
+    {
+      "epoch": 23.244,
+      "grad_norm": 1.3058183193206787,
+      "learning_rate": 2e-05,
+      "loss": 0.0443982,
+      "step": 11622
+    },
+    {
+      "epoch": 23.246,
+      "grad_norm": 1.1419992446899414,
+      "learning_rate": 2e-05,
+      "loss": 0.03190926,
+      "step": 11623
+    },
+    {
+      "epoch": 23.248,
+      "grad_norm": 3.0687108039855957,
+      "learning_rate": 2e-05,
+      "loss": 0.04174676,
+      "step": 11624
+    },
+    {
+      "epoch": 23.25,
+      "grad_norm": 1.1151496171951294,
+      "learning_rate": 2e-05,
+      "loss": 0.0575142,
+      "step": 11625
+    },
+    {
+      "epoch": 23.252,
+      "grad_norm": 1.7354116439819336,
+      "learning_rate": 2e-05,
+      "loss": 0.03637656,
+      "step": 11626
+    },
+    {
+      "epoch": 23.254,
+      "grad_norm": 1.7158818244934082,
+      "learning_rate": 2e-05,
+      "loss": 0.04089367,
+      "step": 11627
+    },
+    {
+      "epoch": 23.256,
+      "grad_norm": 1.027208685874939,
+      "learning_rate": 2e-05,
+      "loss": 0.04252267,
+      "step": 11628
+    },
+    {
+      "epoch": 23.258,
+      "grad_norm": 1.3836479187011719,
+      "learning_rate": 2e-05,
+      "loss": 0.04314929,
+      "step": 11629
+    },
+    {
+      "epoch": 23.26,
+      "grad_norm": 1.6362491846084595,
+      "learning_rate": 2e-05,
+      "loss": 0.04100268,
+      "step": 11630
+    },
+    {
+      "epoch": 23.262,
+      "grad_norm": 1.1033443212509155,
+      "learning_rate": 2e-05,
+      "loss": 0.04861282,
+      "step": 11631
+    },
+    {
+      "epoch": 23.264,
+      "grad_norm": 1.354112148284912,
+      "learning_rate": 2e-05,
+      "loss": 0.03558677,
+      "step": 11632
+    },
+    {
+      "epoch": 23.266,
+      "grad_norm": 2.211174726486206,
+      "learning_rate": 2e-05,
+      "loss": 0.05863594,
+      "step": 11633
+    },
+    {
+      "epoch": 23.268,
+      "grad_norm": 1.1978131532669067,
+      "learning_rate": 2e-05,
+      "loss": 0.02527563,
+      "step": 11634
+    },
+    {
+      "epoch": 23.27,
+      "grad_norm": 1.079159140586853,
+      "learning_rate": 2e-05,
+      "loss": 0.04221547,
+      "step": 11635
+    },
+    {
+      "epoch": 23.272,
+      "grad_norm": 2.7079126834869385,
+      "learning_rate": 2e-05,
+      "loss": 0.04663559,
+      "step": 11636
+    },
+    {
+      "epoch": 23.274,
+      "grad_norm": 1.3492555618286133,
+      "learning_rate": 2e-05,
+      "loss": 0.05419156,
+      "step": 11637
+    },
+    {
+      "epoch": 23.276,
+      "grad_norm": 2.4996025562286377,
+      "learning_rate": 2e-05,
+      "loss": 0.03772485,
+      "step": 11638
+    },
+    {
+      "epoch": 23.278,
+      "grad_norm": 2.083409070968628,
+      "learning_rate": 2e-05,
+      "loss": 0.06437519,
+      "step": 11639
+    },
+    {
+      "epoch": 23.28,
+      "grad_norm": 1.45590341091156,
+      "learning_rate": 2e-05,
+      "loss": 0.03550784,
+      "step": 11640
+    },
+    {
+      "epoch": 23.282,
+      "grad_norm": 1.1977746486663818,
+      "learning_rate": 2e-05,
+      "loss": 0.03832368,
+      "step": 11641
+    },
+    {
+      "epoch": 23.284,
+      "grad_norm": 1.926700234413147,
+      "learning_rate": 2e-05,
+      "loss": 0.06392578,
+      "step": 11642
+    },
+    {
+      "epoch": 23.286,
+      "grad_norm": 2.1961700916290283,
+      "learning_rate": 2e-05,
+      "loss": 0.03601138,
+      "step": 11643
+    },
+    {
+      "epoch": 23.288,
+      "grad_norm": 1.1713889837265015,
+      "learning_rate": 2e-05,
+      "loss": 0.03601219,
+      "step": 11644
+    },
+    {
+      "epoch": 23.29,
+      "grad_norm": 1.6894010305404663,
+      "learning_rate": 2e-05,
+      "loss": 0.0445708,
+      "step": 11645
+    },
+    {
+      "epoch": 23.292,
+      "grad_norm": 1.3147658109664917,
+      "learning_rate": 2e-05,
+      "loss": 0.0355294,
+      "step": 11646
+    },
+    {
+      "epoch": 23.294,
+      "grad_norm": 1.2042319774627686,
+      "learning_rate": 2e-05,
+      "loss": 0.04006883,
+      "step": 11647
+    },
+    {
+      "epoch": 23.296,
+      "grad_norm": 1.1680289506912231,
+      "learning_rate": 2e-05,
+      "loss": 0.04503386,
+      "step": 11648
+    },
+    {
+      "epoch": 23.298,
+      "grad_norm": 1.135467529296875,
+      "learning_rate": 2e-05,
+      "loss": 0.0402963,
+      "step": 11649
+    },
+    {
+      "epoch": 23.3,
+      "grad_norm": 1.505689024925232,
+      "learning_rate": 2e-05,
+      "loss": 0.03736635,
+      "step": 11650
+    },
+    {
+      "epoch": 23.302,
+      "grad_norm": 2.4752161502838135,
+      "learning_rate": 2e-05,
+      "loss": 0.04134458,
+      "step": 11651
+    },
+    {
+      "epoch": 23.304,
+      "grad_norm": 1.4017146825790405,
+      "learning_rate": 2e-05,
+      "loss": 0.04087723,
+      "step": 11652
+    },
+    {
+      "epoch": 23.306,
+      "grad_norm": 1.2549480199813843,
+      "learning_rate": 2e-05,
+      "loss": 0.04560962,
+      "step": 11653
+    },
+    {
+      "epoch": 23.308,
+      "grad_norm": 1.496455192565918,
+      "learning_rate": 2e-05,
+      "loss": 0.04883281,
+      "step": 11654
+    },
+    {
+      "epoch": 23.31,
+      "grad_norm": 1.4462777376174927,
+      "learning_rate": 2e-05,
+      "loss": 0.04287733,
+      "step": 11655
+    },
+    {
+      "epoch": 23.312,
+      "grad_norm": 1.8287996053695679,
+      "learning_rate": 2e-05,
+      "loss": 0.03942668,
+      "step": 11656
+    },
+    {
+      "epoch": 23.314,
+      "grad_norm": 1.9385639429092407,
+      "learning_rate": 2e-05,
+      "loss": 0.06809842,
+      "step": 11657
+    },
+    {
+      "epoch": 23.316,
+      "grad_norm": 0.842710018157959,
+      "learning_rate": 2e-05,
+      "loss": 0.02134952,
+      "step": 11658
+    },
+    {
+      "epoch": 23.318,
+      "grad_norm": 2.0979163646698,
+      "learning_rate": 2e-05,
+      "loss": 0.05558499,
+      "step": 11659
+    },
+    {
+      "epoch": 23.32,
+      "grad_norm": 1.2247816324234009,
+      "learning_rate": 2e-05,
+      "loss": 0.04178263,
+      "step": 11660
+    },
+    {
+      "epoch": 23.322,
+      "grad_norm": 1.5208022594451904,
+      "learning_rate": 2e-05,
+      "loss": 0.04280531,
+      "step": 11661
+    },
+    {
+      "epoch": 23.324,
+      "grad_norm": 2.318500280380249,
+      "learning_rate": 2e-05,
+      "loss": 0.06325538,
+      "step": 11662
+    },
+    {
+      "epoch": 23.326,
+      "grad_norm": 1.6465061902999878,
+      "learning_rate": 2e-05,
+      "loss": 0.05929951,
+      "step": 11663
+    },
+    {
+      "epoch": 23.328,
+      "grad_norm": 1.3455055952072144,
+      "learning_rate": 2e-05,
+      "loss": 0.03735639,
+      "step": 11664
+    },
+    {
+      "epoch": 23.33,
+      "grad_norm": 1.0528393983840942,
+      "learning_rate": 2e-05,
+      "loss": 0.03686676,
+      "step": 11665
+    },
+    {
+      "epoch": 23.332,
+      "grad_norm": 2.8287322521209717,
+      "learning_rate": 2e-05,
+      "loss": 0.04164189,
+      "step": 11666
+    },
+    {
+      "epoch": 23.334,
+      "grad_norm": 1.6893528699874878,
+      "learning_rate": 2e-05,
+      "loss": 0.04093581,
+      "step": 11667
+    },
+    {
+      "epoch": 23.336,
+      "grad_norm": 1.6616655588150024,
+      "learning_rate": 2e-05,
+      "loss": 0.03945962,
+      "step": 11668
+    },
+    {
+      "epoch": 23.338,
+      "grad_norm": 1.084054946899414,
+      "learning_rate": 2e-05,
+      "loss": 0.04473053,
+      "step": 11669
+    },
+    {
+      "epoch": 23.34,
+      "grad_norm": 1.4993929862976074,
+      "learning_rate": 2e-05,
+      "loss": 0.03447219,
+      "step": 11670
+    },
+    {
+      "epoch": 23.342,
+      "grad_norm": 1.050879955291748,
+      "learning_rate": 2e-05,
+      "loss": 0.03411791,
+      "step": 11671
+    },
+    {
+      "epoch": 23.344,
+      "grad_norm": 1.2374294996261597,
+      "learning_rate": 2e-05,
+      "loss": 0.0351023,
+      "step": 11672
+    },
+    {
+      "epoch": 23.346,
+      "grad_norm": 1.3457894325256348,
+      "learning_rate": 2e-05,
+      "loss": 0.04368747,
+      "step": 11673
+    },
+    {
+      "epoch": 23.348,
+      "grad_norm": 0.8786593675613403,
+      "learning_rate": 2e-05,
+      "loss": 0.03530385,
+      "step": 11674
+    },
+    {
+      "epoch": 23.35,
+      "grad_norm": 1.326228380203247,
+      "learning_rate": 2e-05,
+      "loss": 0.04968689,
+      "step": 11675
+    },
+    {
+      "epoch": 23.352,
+      "grad_norm": 0.9884204864501953,
+      "learning_rate": 2e-05,
+      "loss": 0.03281524,
+      "step": 11676
+    },
+    {
+      "epoch": 23.354,
+      "grad_norm": 1.3991303443908691,
+      "learning_rate": 2e-05,
+      "loss": 0.04180757,
+      "step": 11677
+    },
+    {
+      "epoch": 23.356,
+      "grad_norm": 6.225818157196045,
+      "learning_rate": 2e-05,
+      "loss": 0.03679702,
+      "step": 11678
+    },
+    {
+      "epoch": 23.358,
+      "grad_norm": 0.8489519953727722,
+      "learning_rate": 2e-05,
+      "loss": 0.02504526,
+      "step": 11679
+    },
+    {
+      "epoch": 23.36,
+      "grad_norm": 1.34859299659729,
+      "learning_rate": 2e-05,
+      "loss": 0.05301417,
+      "step": 11680
+    },
+    {
+      "epoch": 23.362,
+      "grad_norm": 2.322361469268799,
+      "learning_rate": 2e-05,
+      "loss": 0.0461299,
+      "step": 11681
+    },
+    {
+      "epoch": 23.364,
+      "grad_norm": 1.9422720670700073,
+      "learning_rate": 2e-05,
+      "loss": 0.04207643,
+      "step": 11682
+    },
+    {
+      "epoch": 23.366,
+      "grad_norm": 2.020577907562256,
+      "learning_rate": 2e-05,
+      "loss": 0.04851538,
+      "step": 11683
+    },
+    {
+      "epoch": 23.368,
+      "grad_norm": 1.196955919265747,
+      "learning_rate": 2e-05,
+      "loss": 0.04109028,
+      "step": 11684
+    },
+    {
+      "epoch": 23.37,
+      "grad_norm": 2.016350507736206,
+      "learning_rate": 2e-05,
+      "loss": 0.04552692,
+      "step": 11685
+    },
+    {
+      "epoch": 23.372,
+      "grad_norm": 1.13778555393219,
+      "learning_rate": 2e-05,
+      "loss": 0.02917241,
+      "step": 11686
+    },
+    {
+      "epoch": 23.374,
+      "grad_norm": 1.7037264108657837,
+      "learning_rate": 2e-05,
+      "loss": 0.05261172,
+      "step": 11687
+    },
+    {
+      "epoch": 23.376,
+      "grad_norm": 6.611384868621826,
+      "learning_rate": 2e-05,
+      "loss": 0.03373696,
+      "step": 11688
+    },
+    {
+      "epoch": 23.378,
+      "grad_norm": 1.5197391510009766,
+      "learning_rate": 2e-05,
+      "loss": 0.05965453,
+      "step": 11689
+    },
+    {
+      "epoch": 23.38,
+      "grad_norm": 1.1589999198913574,
+      "learning_rate": 2e-05,
+      "loss": 0.02925105,
+      "step": 11690
+    },
+    {
+      "epoch": 23.382,
+      "grad_norm": 1.4183608293533325,
+      "learning_rate": 2e-05,
+      "loss": 0.03728583,
+      "step": 11691
+    },
+    {
+      "epoch": 23.384,
+      "grad_norm": 1.9492428302764893,
+      "learning_rate": 2e-05,
+      "loss": 0.04366478,
+      "step": 11692
+    },
+    {
+      "epoch": 23.386,
+      "grad_norm": 1.5098069906234741,
+      "learning_rate": 2e-05,
+      "loss": 0.04676771,
+      "step": 11693
+    },
+    {
+      "epoch": 23.388,
+      "grad_norm": 2.645620107650757,
+      "learning_rate": 2e-05,
+      "loss": 0.03401638,
+      "step": 11694
+    },
+    {
+      "epoch": 23.39,
+      "grad_norm": 1.2716730833053589,
+      "learning_rate": 2e-05,
+      "loss": 0.03060681,
+      "step": 11695
+    },
+    {
+      "epoch": 23.392,
+      "grad_norm": 1.4287192821502686,
+      "learning_rate": 2e-05,
+      "loss": 0.04606835,
+      "step": 11696
+    },
+    {
+      "epoch": 23.394,
+      "grad_norm": 2.0073654651641846,
+      "learning_rate": 2e-05,
+      "loss": 0.03770929,
+      "step": 11697
+    },
+    {
+      "epoch": 23.396,
+      "grad_norm": 1.0809135437011719,
+      "learning_rate": 2e-05,
+      "loss": 0.04094427,
+      "step": 11698
+    },
+    {
+      "epoch": 23.398,
+      "grad_norm": 1.341443419456482,
+      "learning_rate": 2e-05,
+      "loss": 0.03574315,
+      "step": 11699
+    },
+    {
+      "epoch": 23.4,
+      "grad_norm": 1.4604313373565674,
+      "learning_rate": 2e-05,
+      "loss": 0.04919671,
+      "step": 11700
+    },
+    {
+      "epoch": 23.402,
+      "grad_norm": 1.32045578956604,
+      "learning_rate": 2e-05,
+      "loss": 0.04904388,
+      "step": 11701
+    },
+    {
+      "epoch": 23.404,
+      "grad_norm": 1.5795214176177979,
+      "learning_rate": 2e-05,
+      "loss": 0.04413631,
+      "step": 11702
+    },
+    {
+      "epoch": 23.406,
+      "grad_norm": 1.566278338432312,
+      "learning_rate": 2e-05,
+      "loss": 0.04938511,
+      "step": 11703
+    },
+    {
+      "epoch": 23.408,
+      "grad_norm": 1.6116101741790771,
+      "learning_rate": 2e-05,
+      "loss": 0.03976731,
+      "step": 11704
+    },
+    {
+      "epoch": 23.41,
+      "grad_norm": 1.1294550895690918,
+      "learning_rate": 2e-05,
+      "loss": 0.04623659,
+      "step": 11705
+    },
+    {
+      "epoch": 23.412,
+      "grad_norm": 2.9337821006774902,
+      "learning_rate": 2e-05,
+      "loss": 0.05919053,
+      "step": 11706
+    },
+    {
+      "epoch": 23.414,
+      "grad_norm": 1.2668352127075195,
+      "learning_rate": 2e-05,
+      "loss": 0.04299584,
+      "step": 11707
+    },
+    {
+      "epoch": 23.416,
+      "grad_norm": 1.6017807722091675,
+      "learning_rate": 2e-05,
+      "loss": 0.04490201,
+      "step": 11708
+    },
+    {
+      "epoch": 23.418,
+      "grad_norm": 1.9470270872116089,
+      "learning_rate": 2e-05,
+      "loss": 0.04873323,
+      "step": 11709
+    },
+    {
+      "epoch": 23.42,
+      "grad_norm": 0.8678619265556335,
+      "learning_rate": 2e-05,
+      "loss": 0.01984765,
+      "step": 11710
+    },
+    {
+      "epoch": 23.422,
+      "grad_norm": 1.015408992767334,
+      "learning_rate": 2e-05,
+      "loss": 0.02890258,
+      "step": 11711
+    },
+    {
+      "epoch": 23.424,
+      "grad_norm": 1.9212663173675537,
+      "learning_rate": 2e-05,
+      "loss": 0.05686199,
+      "step": 11712
+    },
+    {
+      "epoch": 23.426,
+      "grad_norm": 1.5647587776184082,
+      "learning_rate": 2e-05,
+      "loss": 0.04229132,
+      "step": 11713
+    },
+    {
+      "epoch": 23.428,
+      "grad_norm": 1.157758116722107,
+      "learning_rate": 2e-05,
+      "loss": 0.05381518,
+      "step": 11714
+    },
+    {
+      "epoch": 23.43,
+      "grad_norm": 2.0962116718292236,
+      "learning_rate": 2e-05,
+      "loss": 0.04939672,
+      "step": 11715
+    },
+    {
+      "epoch": 23.432,
+      "grad_norm": 1.778320074081421,
+      "learning_rate": 2e-05,
+      "loss": 0.03275887,
+      "step": 11716
+    },
+    {
+      "epoch": 23.434,
+      "grad_norm": 1.3979294300079346,
+      "learning_rate": 2e-05,
+      "loss": 0.04589324,
+      "step": 11717
+    },
+    {
+      "epoch": 23.436,
+      "grad_norm": 1.1280796527862549,
+      "learning_rate": 2e-05,
+      "loss": 0.04539976,
+      "step": 11718
+    },
+    {
+      "epoch": 23.438,
+      "grad_norm": 2.4845614433288574,
+      "learning_rate": 2e-05,
+      "loss": 0.04491439,
+      "step": 11719
+    },
+    {
+      "epoch": 23.44,
+      "grad_norm": 1.0096144676208496,
+      "learning_rate": 2e-05,
+      "loss": 0.03461198,
+      "step": 11720
+    },
+    {
+      "epoch": 23.442,
+      "grad_norm": 2.4556872844696045,
+      "learning_rate": 2e-05,
+      "loss": 0.05000499,
+      "step": 11721
+    },
+    {
+      "epoch": 23.444,
+      "grad_norm": 5.186877727508545,
+      "learning_rate": 2e-05,
+      "loss": 0.05488808,
+      "step": 11722
+    },
+    {
+      "epoch": 23.446,
+      "grad_norm": 1.524732232093811,
+      "learning_rate": 2e-05,
+      "loss": 0.03662727,
+      "step": 11723
+    },
+    {
+      "epoch": 23.448,
+      "grad_norm": 2.764235734939575,
+      "learning_rate": 2e-05,
+      "loss": 0.04267842,
+      "step": 11724
+    },
+    {
+      "epoch": 23.45,
+      "grad_norm": 3.6539905071258545,
+      "learning_rate": 2e-05,
+      "loss": 0.05041397,
+      "step": 11725
+    },
+    {
+      "epoch": 23.452,
+      "grad_norm": 2.682654619216919,
+      "learning_rate": 2e-05,
+      "loss": 0.0332579,
+      "step": 11726
+    },
+    {
+      "epoch": 23.454,
+      "grad_norm": 1.5499696731567383,
+      "learning_rate": 2e-05,
+      "loss": 0.03875547,
+      "step": 11727
+    },
+    {
+      "epoch": 23.456,
+      "grad_norm": 2.318291664123535,
+      "learning_rate": 2e-05,
+      "loss": 0.02854845,
+      "step": 11728
+    },
+    {
+      "epoch": 23.458,
+      "grad_norm": 1.7998384237289429,
+      "learning_rate": 2e-05,
+      "loss": 0.05344633,
+      "step": 11729
+    },
+    {
+      "epoch": 23.46,
+      "grad_norm": 0.9075918197631836,
+      "learning_rate": 2e-05,
+      "loss": 0.02585726,
+      "step": 11730
+    },
+    {
+      "epoch": 23.462,
+      "grad_norm": 1.5107016563415527,
+      "learning_rate": 2e-05,
+      "loss": 0.05805733,
+      "step": 11731
+    },
+    {
+      "epoch": 23.464,
+      "grad_norm": 2.0827245712280273,
+      "learning_rate": 2e-05,
+      "loss": 0.04989189,
+      "step": 11732
+    },
+    {
+      "epoch": 23.466,
+      "grad_norm": 2.3105974197387695,
+      "learning_rate": 2e-05,
+      "loss": 0.06807564,
+      "step": 11733
+    },
+    {
+      "epoch": 23.468,
+      "grad_norm": 1.2695666551589966,
+      "learning_rate": 2e-05,
+      "loss": 0.03619879,
+      "step": 11734
+    },
+    {
+      "epoch": 23.47,
+      "grad_norm": 1.2384021282196045,
+      "learning_rate": 2e-05,
+      "loss": 0.02629324,
+      "step": 11735
+    },
+    {
+      "epoch": 23.472,
+      "grad_norm": 1.6881636381149292,
+      "learning_rate": 2e-05,
+      "loss": 0.04189314,
+      "step": 11736
+    },
+    {
+      "epoch": 23.474,
+      "grad_norm": 2.447591543197632,
+      "learning_rate": 2e-05,
+      "loss": 0.04387872,
+      "step": 11737
+    },
+    {
+      "epoch": 23.476,
+      "grad_norm": 0.8882629871368408,
+      "learning_rate": 2e-05,
+      "loss": 0.02441452,
+      "step": 11738
+    },
+    {
+      "epoch": 23.478,
+      "grad_norm": 1.4331190586090088,
+      "learning_rate": 2e-05,
+      "loss": 0.03524774,
+      "step": 11739
+    },
+    {
+      "epoch": 23.48,
+      "grad_norm": 2.20993709564209,
+      "learning_rate": 2e-05,
+      "loss": 0.05033822,
+      "step": 11740
+    },
+    {
+      "epoch": 23.482,
+      "grad_norm": 1.4944759607315063,
+      "learning_rate": 2e-05,
+      "loss": 0.05018061,
+      "step": 11741
+    },
+    {
+      "epoch": 23.484,
+      "grad_norm": 2.6251118183135986,
+      "learning_rate": 2e-05,
+      "loss": 0.04994038,
+      "step": 11742
+    },
+    {
+      "epoch": 23.486,
+      "grad_norm": 1.345861792564392,
+      "learning_rate": 2e-05,
+      "loss": 0.044821,
+      "step": 11743
+    },
+    {
+      "epoch": 23.488,
+      "grad_norm": 1.239540934562683,
+      "learning_rate": 2e-05,
+      "loss": 0.04019895,
+      "step": 11744
+    },
+    {
+      "epoch": 23.49,
+      "grad_norm": 1.280549168586731,
+      "learning_rate": 2e-05,
+      "loss": 0.03652037,
+      "step": 11745
+    },
+    {
+      "epoch": 23.492,
+      "grad_norm": 1.523679256439209,
+      "learning_rate": 2e-05,
+      "loss": 0.04372611,
+      "step": 11746
+    },
+    {
+      "epoch": 23.494,
+      "grad_norm": 1.49849271774292,
+      "learning_rate": 2e-05,
+      "loss": 0.03744507,
+      "step": 11747
+    },
+    {
+      "epoch": 23.496,
+      "grad_norm": 1.3029149770736694,
+      "learning_rate": 2e-05,
+      "loss": 0.04071214,
+      "step": 11748
+    },
+    {
+      "epoch": 23.498,
+      "grad_norm": 2.212434768676758,
+      "learning_rate": 2e-05,
+      "loss": 0.06702287,
+      "step": 11749
+    },
+    {
+      "epoch": 23.5,
+      "grad_norm": 1.800525188446045,
+      "learning_rate": 2e-05,
+      "loss": 0.04466784,
+      "step": 11750
+    },
+    {
+      "epoch": 23.502,
+      "grad_norm": 1.8852155208587646,
+      "learning_rate": 2e-05,
+      "loss": 0.06043893,
+      "step": 11751
+    },
+    {
+      "epoch": 23.504,
+      "grad_norm": 1.5796104669570923,
+      "learning_rate": 2e-05,
+      "loss": 0.03296822,
+      "step": 11752
+    },
+    {
+      "epoch": 23.506,
+      "grad_norm": 1.2062172889709473,
+      "learning_rate": 2e-05,
+      "loss": 0.03532603,
+      "step": 11753
+    },
+    {
+      "epoch": 23.508,
+      "grad_norm": 2.296785831451416,
+      "learning_rate": 2e-05,
+      "loss": 0.06512903,
+      "step": 11754
+    },
+    {
+      "epoch": 23.51,
+      "grad_norm": 1.458593487739563,
+      "learning_rate": 2e-05,
+      "loss": 0.06009974,
+      "step": 11755
+    },
+    {
+      "epoch": 23.512,
+      "grad_norm": 1.1910109519958496,
+      "learning_rate": 2e-05,
+      "loss": 0.03358722,
+      "step": 11756
+    },
+    {
+      "epoch": 23.514,
+      "grad_norm": 1.1627408266067505,
+      "learning_rate": 2e-05,
+      "loss": 0.03376558,
+      "step": 11757
+    },
+    {
+      "epoch": 23.516,
+      "grad_norm": 2.2633957862854004,
+      "learning_rate": 2e-05,
+      "loss": 0.06547912,
+      "step": 11758
+    },
+    {
+      "epoch": 23.518,
+      "grad_norm": 0.9992890357971191,
+      "learning_rate": 2e-05,
+      "loss": 0.02558643,
+      "step": 11759
+    },
+    {
+      "epoch": 23.52,
+      "grad_norm": 2.1989448070526123,
+      "learning_rate": 2e-05,
+      "loss": 0.04395504,
+      "step": 11760
+    },
+    {
+      "epoch": 23.522,
+      "grad_norm": 2.1435000896453857,
+      "learning_rate": 2e-05,
+      "loss": 0.04401811,
+      "step": 11761
+    },
+    {
+      "epoch": 23.524,
+      "grad_norm": 1.109850287437439,
+      "learning_rate": 2e-05,
+      "loss": 0.02487316,
+      "step": 11762
+    },
+    {
+      "epoch": 23.526,
+      "grad_norm": 2.0983777046203613,
+      "learning_rate": 2e-05,
+      "loss": 0.04313958,
+      "step": 11763
+    },
+    {
+      "epoch": 23.528,
+      "grad_norm": 1.1474552154541016,
+      "learning_rate": 2e-05,
+      "loss": 0.04640581,
+      "step": 11764
+    },
+    {
+      "epoch": 23.53,
+      "grad_norm": 2.8319504261016846,
+      "learning_rate": 2e-05,
+      "loss": 0.05273167,
+      "step": 11765
+    },
+    {
+      "epoch": 23.532,
+      "grad_norm": 1.2033793926239014,
+      "learning_rate": 2e-05,
+      "loss": 0.0447002,
+      "step": 11766
+    },
+    {
+      "epoch": 23.534,
+      "grad_norm": 0.9693114757537842,
+      "learning_rate": 2e-05,
+      "loss": 0.02837565,
+      "step": 11767
+    },
+    {
+      "epoch": 23.536,
+      "grad_norm": 0.9325622916221619,
+      "learning_rate": 2e-05,
+      "loss": 0.03289568,
+      "step": 11768
+    },
+    {
+      "epoch": 23.538,
+      "grad_norm": 1.8937511444091797,
+      "learning_rate": 2e-05,
+      "loss": 0.05601943,
+      "step": 11769
+    },
+    {
+      "epoch": 23.54,
+      "grad_norm": 1.1438976526260376,
+      "learning_rate": 2e-05,
+      "loss": 0.04349836,
+      "step": 11770
+    },
+    {
+      "epoch": 23.542,
+      "grad_norm": 1.7353458404541016,
+      "learning_rate": 2e-05,
+      "loss": 0.06614293,
+      "step": 11771
+    },
+    {
+      "epoch": 23.544,
+      "grad_norm": 1.3022950887680054,
+      "learning_rate": 2e-05,
+      "loss": 0.03682633,
+      "step": 11772
+    },
+    {
+      "epoch": 23.546,
+      "grad_norm": 2.2560153007507324,
+      "learning_rate": 2e-05,
+      "loss": 0.03180346,
+      "step": 11773
+    },
+    {
+      "epoch": 23.548000000000002,
+      "grad_norm": 2.171947479248047,
+      "learning_rate": 2e-05,
+      "loss": 0.0739259,
+      "step": 11774
+    },
+    {
+      "epoch": 23.55,
+      "grad_norm": 1.3080508708953857,
+      "learning_rate": 2e-05,
+      "loss": 0.0323229,
+      "step": 11775
+    },
+    {
+      "epoch": 23.552,
+      "grad_norm": 1.3092063665390015,
+      "learning_rate": 2e-05,
+      "loss": 0.04459471,
+      "step": 11776
+    },
+    {
+      "epoch": 23.554,
+      "grad_norm": 1.0528229475021362,
+      "learning_rate": 2e-05,
+      "loss": 0.02914536,
+      "step": 11777
+    },
+    {
+      "epoch": 23.556,
+      "grad_norm": 1.155535340309143,
+      "learning_rate": 2e-05,
+      "loss": 0.03187998,
+      "step": 11778
+    },
+    {
+      "epoch": 23.558,
+      "grad_norm": 1.5524402856826782,
+      "learning_rate": 2e-05,
+      "loss": 0.04663787,
+      "step": 11779
+    },
+    {
+      "epoch": 23.56,
+      "grad_norm": 1.0907926559448242,
+      "learning_rate": 2e-05,
+      "loss": 0.04237938,
+      "step": 11780
+    },
+    {
+      "epoch": 23.562,
+      "grad_norm": 1.2452819347381592,
+      "learning_rate": 2e-05,
+      "loss": 0.04205526,
+      "step": 11781
+    },
+    {
+      "epoch": 23.564,
+      "grad_norm": 1.1769381761550903,
+      "learning_rate": 2e-05,
+      "loss": 0.03974335,
+      "step": 11782
+    },
+    {
+      "epoch": 23.566,
+      "grad_norm": 1.1212342977523804,
+      "learning_rate": 2e-05,
+      "loss": 0.04549305,
+      "step": 11783
+    },
+    {
+      "epoch": 23.568,
+      "grad_norm": 1.3117882013320923,
+      "learning_rate": 2e-05,
+      "loss": 0.03518543,
+      "step": 11784
+    },
+    {
+      "epoch": 23.57,
+      "grad_norm": 2.5933187007904053,
+      "learning_rate": 2e-05,
+      "loss": 0.05680026,
+      "step": 11785
+    },
+    {
+      "epoch": 23.572,
+      "grad_norm": 1.0752009153366089,
+      "learning_rate": 2e-05,
+      "loss": 0.03501371,
+      "step": 11786
+    },
+    {
+      "epoch": 23.574,
+      "grad_norm": 1.2819617986679077,
+      "learning_rate": 2e-05,
+      "loss": 0.04427963,
+      "step": 11787
+    },
+    {
+      "epoch": 23.576,
+      "grad_norm": 1.0761966705322266,
+      "learning_rate": 2e-05,
+      "loss": 0.03001852,
+      "step": 11788
+    },
+    {
+      "epoch": 23.578,
+      "grad_norm": 1.9416499137878418,
+      "learning_rate": 2e-05,
+      "loss": 0.05700012,
+      "step": 11789
+    },
+    {
+      "epoch": 23.58,
+      "grad_norm": 2.196974992752075,
+      "learning_rate": 2e-05,
+      "loss": 0.06811506,
+      "step": 11790
+    },
+    {
+      "epoch": 23.582,
+      "grad_norm": 1.458816409111023,
+      "learning_rate": 2e-05,
+      "loss": 0.03885439,
+      "step": 11791
+    },
+    {
+      "epoch": 23.584,
+      "grad_norm": 1.5310564041137695,
+      "learning_rate": 2e-05,
+      "loss": 0.04649789,
+      "step": 11792
+    },
+    {
+      "epoch": 23.586,
+      "grad_norm": 1.3450838327407837,
+      "learning_rate": 2e-05,
+      "loss": 0.05772156,
+      "step": 11793
+    },
+    {
+      "epoch": 23.588,
+      "grad_norm": 1.8807865381240845,
+      "learning_rate": 2e-05,
+      "loss": 0.04831642,
+      "step": 11794
+    },
+    {
+      "epoch": 23.59,
+      "grad_norm": 1.4770256280899048,
+      "learning_rate": 2e-05,
+      "loss": 0.05228359,
+      "step": 11795
+    },
+    {
+      "epoch": 23.592,
+      "grad_norm": 1.5434561967849731,
+      "learning_rate": 2e-05,
+      "loss": 0.04642072,
+      "step": 11796
+    },
+    {
+      "epoch": 23.594,
+      "grad_norm": 1.5032665729522705,
+      "learning_rate": 2e-05,
+      "loss": 0.04318558,
+      "step": 11797
+    },
+    {
+      "epoch": 23.596,
+      "grad_norm": 1.2868092060089111,
+      "learning_rate": 2e-05,
+      "loss": 0.05032202,
+      "step": 11798
+    },
+    {
+      "epoch": 23.598,
+      "grad_norm": 2.0231192111968994,
+      "learning_rate": 2e-05,
+      "loss": 0.0559545,
+      "step": 11799
+    },
+    {
+      "epoch": 23.6,
+      "grad_norm": 1.333901047706604,
+      "learning_rate": 2e-05,
+      "loss": 0.04533472,
+      "step": 11800
+    },
+    {
+      "epoch": 23.602,
+      "grad_norm": 1.2813880443572998,
+      "learning_rate": 2e-05,
+      "loss": 0.04703806,
+      "step": 11801
+    },
+    {
+      "epoch": 23.604,
+      "grad_norm": 2.0209999084472656,
+      "learning_rate": 2e-05,
+      "loss": 0.03075261,
+      "step": 11802
+    },
+    {
+      "epoch": 23.606,
+      "grad_norm": 0.8158363103866577,
+      "learning_rate": 2e-05,
+      "loss": 0.02488266,
+      "step": 11803
+    },
+    {
+      "epoch": 23.608,
+      "grad_norm": 1.2968827486038208,
+      "learning_rate": 2e-05,
+      "loss": 0.04825204,
+      "step": 11804
+    },
+    {
+      "epoch": 23.61,
+      "grad_norm": 1.1694176197052002,
+      "learning_rate": 2e-05,
+      "loss": 0.04675373,
+      "step": 11805
+    },
+    {
+      "epoch": 23.612,
+      "grad_norm": 2.4291164875030518,
+      "learning_rate": 2e-05,
+      "loss": 0.03782834,
+      "step": 11806
+    },
+    {
+      "epoch": 23.614,
+      "grad_norm": 3.2317073345184326,
+      "learning_rate": 2e-05,
+      "loss": 0.05336016,
+      "step": 11807
+    },
+    {
+      "epoch": 23.616,
+      "grad_norm": 2.4341442584991455,
+      "learning_rate": 2e-05,
+      "loss": 0.04534258,
+      "step": 11808
+    },
+    {
+      "epoch": 23.618,
+      "grad_norm": 1.3405523300170898,
+      "learning_rate": 2e-05,
+      "loss": 0.03542194,
+      "step": 11809
+    },
+    {
+      "epoch": 23.62,
+      "grad_norm": 0.952492356300354,
+      "learning_rate": 2e-05,
+      "loss": 0.02990831,
+      "step": 11810
+    },
+    {
+      "epoch": 23.622,
+      "grad_norm": 1.8398486375808716,
+      "learning_rate": 2e-05,
+      "loss": 0.04314672,
+      "step": 11811
+    },
+    {
+      "epoch": 23.624,
+      "grad_norm": 1.486546277999878,
+      "learning_rate": 2e-05,
+      "loss": 0.04819717,
+      "step": 11812
+    },
+    {
+      "epoch": 23.626,
+      "grad_norm": 1.3383139371871948,
+      "learning_rate": 2e-05,
+      "loss": 0.03825889,
+      "step": 11813
+    },
+    {
+      "epoch": 23.628,
+      "grad_norm": 1.57940673828125,
+      "learning_rate": 2e-05,
+      "loss": 0.03851825,
+      "step": 11814
+    },
+    {
+      "epoch": 23.63,
+      "grad_norm": 1.3672587871551514,
+      "learning_rate": 2e-05,
+      "loss": 0.04236236,
+      "step": 11815
+    },
+    {
+      "epoch": 23.632,
+      "grad_norm": 1.9322770833969116,
+      "learning_rate": 2e-05,
+      "loss": 0.04957227,
+      "step": 11816
+    },
+    {
+      "epoch": 23.634,
+      "grad_norm": 2.318655014038086,
+      "learning_rate": 2e-05,
+      "loss": 0.05477686,
+      "step": 11817
+    },
+    {
+      "epoch": 23.636,
+      "grad_norm": 1.3054866790771484,
+      "learning_rate": 2e-05,
+      "loss": 0.0492695,
+      "step": 11818
+    },
+    {
+      "epoch": 23.638,
+      "grad_norm": 1.7329834699630737,
+      "learning_rate": 2e-05,
+      "loss": 0.03611884,
+      "step": 11819
+    },
+    {
+      "epoch": 23.64,
+      "grad_norm": 1.0600239038467407,
+      "learning_rate": 2e-05,
+      "loss": 0.0398752,
+      "step": 11820
+    },
+    {
+      "epoch": 23.642,
+      "grad_norm": 1.5419100522994995,
+      "learning_rate": 2e-05,
+      "loss": 0.05457453,
+      "step": 11821
+    },
+    {
+      "epoch": 23.644,
+      "grad_norm": 1.1670844554901123,
+      "learning_rate": 2e-05,
+      "loss": 0.04318011,
+      "step": 11822
+    },
+    {
+      "epoch": 23.646,
+      "grad_norm": 1.8038170337677002,
+      "learning_rate": 2e-05,
+      "loss": 0.05024581,
+      "step": 11823
+    },
+    {
+      "epoch": 23.648,
+      "grad_norm": 1.4596983194351196,
+      "learning_rate": 2e-05,
+      "loss": 0.04731473,
+      "step": 11824
+    },
+    {
+      "epoch": 23.65,
+      "grad_norm": 2.0060338973999023,
+      "learning_rate": 2e-05,
+      "loss": 0.03213666,
+      "step": 11825
+    },
+    {
+      "epoch": 23.652,
+      "grad_norm": 2.789665699005127,
+      "learning_rate": 2e-05,
+      "loss": 0.05821307,
+      "step": 11826
+    },
+    {
+      "epoch": 23.654,
+      "grad_norm": 3.190103054046631,
+      "learning_rate": 2e-05,
+      "loss": 0.04068895,
+      "step": 11827
+    },
+    {
+      "epoch": 23.656,
+      "grad_norm": 1.2811437845230103,
+      "learning_rate": 2e-05,
+      "loss": 0.05212625,
+      "step": 11828
+    },
+    {
+      "epoch": 23.658,
+      "grad_norm": 1.1203440427780151,
+      "learning_rate": 2e-05,
+      "loss": 0.04808802,
+      "step": 11829
+    },
+    {
+      "epoch": 23.66,
+      "grad_norm": 2.3472464084625244,
+      "learning_rate": 2e-05,
+      "loss": 0.0419308,
+      "step": 11830
+    },
+    {
+      "epoch": 23.662,
+      "grad_norm": 1.1911481618881226,
+      "learning_rate": 2e-05,
+      "loss": 0.05066958,
+      "step": 11831
+    },
+    {
+      "epoch": 23.664,
+      "grad_norm": 1.5024276971817017,
+      "learning_rate": 2e-05,
+      "loss": 0.03530101,
+      "step": 11832
+    },
+    {
+      "epoch": 23.666,
+      "grad_norm": 2.032147169113159,
+      "learning_rate": 2e-05,
+      "loss": 0.04991383,
+      "step": 11833
+    },
+    {
+      "epoch": 23.668,
+      "grad_norm": 1.2729007005691528,
+      "learning_rate": 2e-05,
+      "loss": 0.04908589,
+      "step": 11834
+    },
+    {
+      "epoch": 23.67,
+      "grad_norm": 7.958946228027344,
+      "learning_rate": 2e-05,
+      "loss": 0.06328648,
+      "step": 11835
+    },
+    {
+      "epoch": 23.672,
+      "grad_norm": 1.1186507940292358,
+      "learning_rate": 2e-05,
+      "loss": 0.03300332,
+      "step": 11836
+    },
+    {
+      "epoch": 23.674,
+      "grad_norm": 1.4105454683303833,
+      "learning_rate": 2e-05,
+      "loss": 0.04340991,
+      "step": 11837
+    },
+    {
+      "epoch": 23.676,
+      "grad_norm": 1.504833698272705,
+      "learning_rate": 2e-05,
+      "loss": 0.05525521,
+      "step": 11838
+    },
+    {
+      "epoch": 23.678,
+      "grad_norm": 1.2396059036254883,
+      "learning_rate": 2e-05,
+      "loss": 0.04817509,
+      "step": 11839
+    },
+    {
+      "epoch": 23.68,
+      "grad_norm": 1.5307303667068481,
+      "learning_rate": 2e-05,
+      "loss": 0.02923251,
+      "step": 11840
+    },
+    {
+      "epoch": 23.682,
+      "grad_norm": 1.6105175018310547,
+      "learning_rate": 2e-05,
+      "loss": 0.03141434,
+      "step": 11841
+    },
+    {
+      "epoch": 23.684,
+      "grad_norm": 1.293770670890808,
+      "learning_rate": 2e-05,
+      "loss": 0.05276821,
+      "step": 11842
+    },
+    {
+      "epoch": 23.686,
+      "grad_norm": 1.8566535711288452,
+      "learning_rate": 2e-05,
+      "loss": 0.04680733,
+      "step": 11843
+    },
+    {
+      "epoch": 23.688,
+      "grad_norm": 1.2725869417190552,
+      "learning_rate": 2e-05,
+      "loss": 0.03304363,
+      "step": 11844
+    },
+    {
+      "epoch": 23.69,
+      "grad_norm": 1.192263126373291,
+      "learning_rate": 2e-05,
+      "loss": 0.03634157,
+      "step": 11845
+    },
+    {
+      "epoch": 23.692,
+      "grad_norm": 3.0075485706329346,
+      "learning_rate": 2e-05,
+      "loss": 0.04553998,
+      "step": 11846
+    },
+    {
+      "epoch": 23.694,
+      "grad_norm": 2.9852254390716553,
+      "learning_rate": 2e-05,
+      "loss": 0.06526848,
+      "step": 11847
+    },
+    {
+      "epoch": 23.696,
+      "grad_norm": 1.9898120164871216,
+      "learning_rate": 2e-05,
+      "loss": 0.03190294,
+      "step": 11848
+    },
+    {
+      "epoch": 23.698,
+      "grad_norm": 1.2109845876693726,
+      "learning_rate": 2e-05,
+      "loss": 0.04095478,
+      "step": 11849
+    },
+    {
+      "epoch": 23.7,
+      "grad_norm": 3.1448001861572266,
+      "learning_rate": 2e-05,
+      "loss": 0.0624724,
+      "step": 11850
+    },
+    {
+      "epoch": 23.701999999999998,
+      "grad_norm": 1.063796043395996,
+      "learning_rate": 2e-05,
+      "loss": 0.04226255,
+      "step": 11851
+    },
+    {
+      "epoch": 23.704,
+      "grad_norm": 1.0157510042190552,
+      "learning_rate": 2e-05,
+      "loss": 0.04364079,
+      "step": 11852
+    },
+    {
+      "epoch": 23.706,
+      "grad_norm": 0.9760161638259888,
+      "learning_rate": 2e-05,
+      "loss": 0.02745278,
+      "step": 11853
+    },
+    {
+      "epoch": 23.708,
+      "grad_norm": 1.4634634256362915,
+      "learning_rate": 2e-05,
+      "loss": 0.04308833,
+      "step": 11854
+    },
+    {
+      "epoch": 23.71,
+      "grad_norm": 1.4211335182189941,
+      "learning_rate": 2e-05,
+      "loss": 0.03839317,
+      "step": 11855
+    },
+    {
+      "epoch": 23.712,
+      "grad_norm": 1.0064549446105957,
+      "learning_rate": 2e-05,
+      "loss": 0.03243163,
+      "step": 11856
+    },
+    {
+      "epoch": 23.714,
+      "grad_norm": 1.0889908075332642,
+      "learning_rate": 2e-05,
+      "loss": 0.03015575,
+      "step": 11857
+    },
+    {
+      "epoch": 23.716,
+      "grad_norm": 1.499775767326355,
+      "learning_rate": 2e-05,
+      "loss": 0.04761718,
+      "step": 11858
+    },
+    {
+      "epoch": 23.718,
+      "grad_norm": 1.0862691402435303,
+      "learning_rate": 2e-05,
+      "loss": 0.03753242,
+      "step": 11859
+    },
+    {
+      "epoch": 23.72,
+      "grad_norm": 0.9295043349266052,
+      "learning_rate": 2e-05,
+      "loss": 0.02882018,
+      "step": 11860
+    },
+    {
+      "epoch": 23.722,
+      "grad_norm": 1.0936144590377808,
+      "learning_rate": 2e-05,
+      "loss": 0.03536011,
+      "step": 11861
+    },
+    {
+      "epoch": 23.724,
+      "grad_norm": 0.9443338513374329,
+      "learning_rate": 2e-05,
+      "loss": 0.03331787,
+      "step": 11862
+    },
+    {
+      "epoch": 23.726,
+      "grad_norm": 1.0032241344451904,
+      "learning_rate": 2e-05,
+      "loss": 0.02793972,
+      "step": 11863
+    },
+    {
+      "epoch": 23.728,
+      "grad_norm": 1.677073359489441,
+      "learning_rate": 2e-05,
+      "loss": 0.04015389,
+      "step": 11864
+    },
+    {
+      "epoch": 23.73,
+      "grad_norm": 2.5019454956054688,
+      "learning_rate": 2e-05,
+      "loss": 0.04113224,
+      "step": 11865
+    },
+    {
+      "epoch": 23.732,
+      "grad_norm": 1.2569458484649658,
+      "learning_rate": 2e-05,
+      "loss": 0.04976218,
+      "step": 11866
+    },
+    {
+      "epoch": 23.734,
+      "grad_norm": 1.3353568315505981,
+      "learning_rate": 2e-05,
+      "loss": 0.0452405,
+      "step": 11867
+    },
+    {
+      "epoch": 23.736,
+      "grad_norm": 1.0816160440444946,
+      "learning_rate": 2e-05,
+      "loss": 0.03934583,
+      "step": 11868
+    },
+    {
+      "epoch": 23.738,
+      "grad_norm": 2.60617733001709,
+      "learning_rate": 2e-05,
+      "loss": 0.04733625,
+      "step": 11869
+    },
+    {
+      "epoch": 23.74,
+      "grad_norm": 1.3264927864074707,
+      "learning_rate": 2e-05,
+      "loss": 0.04373094,
+      "step": 11870
+    },
+    {
+      "epoch": 23.742,
+      "grad_norm": 1.8265389204025269,
+      "learning_rate": 2e-05,
+      "loss": 0.04801701,
+      "step": 11871
+    },
+    {
+      "epoch": 23.744,
+      "grad_norm": 1.5918618440628052,
+      "learning_rate": 2e-05,
+      "loss": 0.03365224,
+      "step": 11872
+    },
+    {
+      "epoch": 23.746,
+      "grad_norm": 1.1717815399169922,
+      "learning_rate": 2e-05,
+      "loss": 0.04041979,
+      "step": 11873
+    },
+    {
+      "epoch": 23.748,
+      "grad_norm": 1.1238340139389038,
+      "learning_rate": 2e-05,
+      "loss": 0.03989327,
+      "step": 11874
+    },
+    {
+      "epoch": 23.75,
+      "grad_norm": 3.7857089042663574,
+      "learning_rate": 2e-05,
+      "loss": 0.04945645,
+      "step": 11875
+    },
+    {
+      "epoch": 23.752,
+      "grad_norm": 3.371209144592285,
+      "learning_rate": 2e-05,
+      "loss": 0.0476009,
+      "step": 11876
+    },
+    {
+      "epoch": 23.754,
+      "grad_norm": 1.3550797700881958,
+      "learning_rate": 2e-05,
+      "loss": 0.04116515,
+      "step": 11877
+    },
+    {
+      "epoch": 23.756,
+      "grad_norm": 1.1104875802993774,
+      "learning_rate": 2e-05,
+      "loss": 0.03957765,
+      "step": 11878
+    },
+    {
+      "epoch": 23.758,
+      "grad_norm": 1.2116479873657227,
+      "learning_rate": 2e-05,
+      "loss": 0.03302076,
+      "step": 11879
+    },
+    {
+      "epoch": 23.76,
+      "grad_norm": 1.565588355064392,
+      "learning_rate": 2e-05,
+      "loss": 0.05508092,
+      "step": 11880
+    },
+    {
+      "epoch": 23.762,
+      "grad_norm": 6.225358486175537,
+      "learning_rate": 2e-05,
+      "loss": 0.0478447,
+      "step": 11881
+    },
+    {
+      "epoch": 23.764,
+      "grad_norm": 1.688665747642517,
+      "learning_rate": 2e-05,
+      "loss": 0.0437132,
+      "step": 11882
+    },
+    {
+      "epoch": 23.766,
+      "grad_norm": 1.2847251892089844,
+      "learning_rate": 2e-05,
+      "loss": 0.04984112,
+      "step": 11883
+    },
+    {
+      "epoch": 23.768,
+      "grad_norm": 1.4837887287139893,
+      "learning_rate": 2e-05,
+      "loss": 0.03870174,
+      "step": 11884
+    },
+    {
+      "epoch": 23.77,
+      "grad_norm": 1.2898259162902832,
+      "learning_rate": 2e-05,
+      "loss": 0.0534673,
+      "step": 11885
+    },
+    {
+      "epoch": 23.772,
+      "grad_norm": 1.0326087474822998,
+      "learning_rate": 2e-05,
+      "loss": 0.0354575,
+      "step": 11886
+    },
+    {
+      "epoch": 23.774,
+      "grad_norm": 1.8046265840530396,
+      "learning_rate": 2e-05,
+      "loss": 0.0476476,
+      "step": 11887
+    },
+    {
+      "epoch": 23.776,
+      "grad_norm": 1.4603756666183472,
+      "learning_rate": 2e-05,
+      "loss": 0.04588185,
+      "step": 11888
+    },
+    {
+      "epoch": 23.778,
+      "grad_norm": 1.1678380966186523,
+      "learning_rate": 2e-05,
+      "loss": 0.04390156,
+      "step": 11889
+    },
+    {
+      "epoch": 23.78,
+      "grad_norm": 1.5542722940444946,
+      "learning_rate": 2e-05,
+      "loss": 0.04794715,
+      "step": 11890
+    },
+    {
+      "epoch": 23.782,
+      "grad_norm": 1.2043498754501343,
+      "learning_rate": 2e-05,
+      "loss": 0.03155842,
+      "step": 11891
+    },
+    {
+      "epoch": 23.784,
+      "grad_norm": 1.1357460021972656,
+      "learning_rate": 2e-05,
+      "loss": 0.04152321,
+      "step": 11892
+    },
+    {
+      "epoch": 23.786,
+      "grad_norm": 1.1794549226760864,
+      "learning_rate": 2e-05,
+      "loss": 0.04586139,
+      "step": 11893
+    },
+    {
+      "epoch": 23.788,
+      "grad_norm": 0.9895995855331421,
+      "learning_rate": 2e-05,
+      "loss": 0.0327672,
+      "step": 11894
+    },
+    {
+      "epoch": 23.79,
+      "grad_norm": 1.7362169027328491,
+      "learning_rate": 2e-05,
+      "loss": 0.05948992,
+      "step": 11895
+    },
+    {
+      "epoch": 23.792,
+      "grad_norm": 0.9421995282173157,
+      "learning_rate": 2e-05,
+      "loss": 0.02367941,
+      "step": 11896
+    },
+    {
+      "epoch": 23.794,
+      "grad_norm": 2.629747152328491,
+      "learning_rate": 2e-05,
+      "loss": 0.04058916,
+      "step": 11897
+    },
+    {
+      "epoch": 23.796,
+      "grad_norm": 1.0108635425567627,
+      "learning_rate": 2e-05,
+      "loss": 0.02985597,
+      "step": 11898
+    },
+    {
+      "epoch": 23.798000000000002,
+      "grad_norm": 3.040156126022339,
+      "learning_rate": 2e-05,
+      "loss": 0.0340285,
+      "step": 11899
+    },
+    {
+      "epoch": 23.8,
+      "grad_norm": 1.1744736433029175,
+      "learning_rate": 2e-05,
+      "loss": 0.04568225,
+      "step": 11900
+    },
+    {
+      "epoch": 23.802,
+      "grad_norm": 2.0291764736175537,
+      "learning_rate": 2e-05,
+      "loss": 0.05318277,
+      "step": 11901
+    },
+    {
+      "epoch": 23.804,
+      "grad_norm": 1.3046629428863525,
+      "learning_rate": 2e-05,
+      "loss": 0.05274621,
+      "step": 11902
+    },
+    {
+      "epoch": 23.806,
+      "grad_norm": 2.1963841915130615,
+      "learning_rate": 2e-05,
+      "loss": 0.04720778,
+      "step": 11903
+    },
+    {
+      "epoch": 23.808,
+      "grad_norm": 2.655916213989258,
+      "learning_rate": 2e-05,
+      "loss": 0.0420041,
+      "step": 11904
+    },
+    {
+      "epoch": 23.81,
+      "grad_norm": 1.176939845085144,
+      "learning_rate": 2e-05,
+      "loss": 0.04580169,
+      "step": 11905
+    },
+    {
+      "epoch": 23.812,
+      "grad_norm": 1.0048329830169678,
+      "learning_rate": 2e-05,
+      "loss": 0.03731166,
+      "step": 11906
+    },
+    {
+      "epoch": 23.814,
+      "grad_norm": 1.414959192276001,
+      "learning_rate": 2e-05,
+      "loss": 0.06279035,
+      "step": 11907
+    },
+    {
+      "epoch": 23.816,
+      "grad_norm": 1.7334407567977905,
+      "learning_rate": 2e-05,
+      "loss": 0.03839599,
+      "step": 11908
+    },
+    {
+      "epoch": 23.818,
+      "grad_norm": 1.2090424299240112,
+      "learning_rate": 2e-05,
+      "loss": 0.03566598,
+      "step": 11909
+    },
+    {
+      "epoch": 23.82,
+      "grad_norm": 1.0905237197875977,
+      "learning_rate": 2e-05,
+      "loss": 0.04220146,
+      "step": 11910
+    },
+    {
+      "epoch": 23.822,
+      "grad_norm": 2.646314859390259,
+      "learning_rate": 2e-05,
+      "loss": 0.06015548,
+      "step": 11911
+    },
+    {
+      "epoch": 23.824,
+      "grad_norm": 1.4104152917861938,
+      "learning_rate": 2e-05,
+      "loss": 0.03940243,
+      "step": 11912
+    },
+    {
+      "epoch": 23.826,
+      "grad_norm": 2.7767820358276367,
+      "learning_rate": 2e-05,
+      "loss": 0.04243702,
+      "step": 11913
+    },
+    {
+      "epoch": 23.828,
+      "grad_norm": 1.3988125324249268,
+      "learning_rate": 2e-05,
+      "loss": 0.03724855,
+      "step": 11914
+    },
+    {
+      "epoch": 23.83,
+      "grad_norm": 1.0905022621154785,
+      "learning_rate": 2e-05,
+      "loss": 0.02765215,
+      "step": 11915
+    },
+    {
+      "epoch": 23.832,
+      "grad_norm": 1.3085315227508545,
+      "learning_rate": 2e-05,
+      "loss": 0.02965268,
+      "step": 11916
+    },
+    {
+      "epoch": 23.834,
+      "grad_norm": 2.0017828941345215,
+      "learning_rate": 2e-05,
+      "loss": 0.05234018,
+      "step": 11917
+    },
+    {
+      "epoch": 23.836,
+      "grad_norm": 1.6834447383880615,
+      "learning_rate": 2e-05,
+      "loss": 0.04721446,
+      "step": 11918
+    },
+    {
+      "epoch": 23.838,
+      "grad_norm": 1.4597904682159424,
+      "learning_rate": 2e-05,
+      "loss": 0.05151546,
+      "step": 11919
+    },
+    {
+      "epoch": 23.84,
+      "grad_norm": 2.2075035572052,
+      "learning_rate": 2e-05,
+      "loss": 0.04451163,
+      "step": 11920
+    },
+    {
+      "epoch": 23.842,
+      "grad_norm": 1.4260821342468262,
+      "learning_rate": 2e-05,
+      "loss": 0.03413969,
+      "step": 11921
+    },
+    {
+      "epoch": 23.844,
+      "grad_norm": 1.2742419242858887,
+      "learning_rate": 2e-05,
+      "loss": 0.0354282,
+      "step": 11922
+    },
+    {
+      "epoch": 23.846,
+      "grad_norm": 1.3900338411331177,
+      "learning_rate": 2e-05,
+      "loss": 0.03880811,
+      "step": 11923
+    },
+    {
+      "epoch": 23.848,
+      "grad_norm": 1.640749216079712,
+      "learning_rate": 2e-05,
+      "loss": 0.03321562,
+      "step": 11924
+    },
+    {
+      "epoch": 23.85,
+      "grad_norm": 1.2059239149093628,
+      "learning_rate": 2e-05,
+      "loss": 0.03351336,
+      "step": 11925
+    },
+    {
+      "epoch": 23.852,
+      "grad_norm": 1.6385819911956787,
+      "learning_rate": 2e-05,
+      "loss": 0.03568233,
+      "step": 11926
+    },
+    {
+      "epoch": 23.854,
+      "grad_norm": 1.6820650100708008,
+      "learning_rate": 2e-05,
+      "loss": 0.04155786,
+      "step": 11927
+    },
+    {
+      "epoch": 23.856,
+      "grad_norm": 1.2074123620986938,
+      "learning_rate": 2e-05,
+      "loss": 0.04882422,
+      "step": 11928
+    },
+    {
+      "epoch": 23.858,
+      "grad_norm": 1.130673885345459,
+      "learning_rate": 2e-05,
+      "loss": 0.03696358,
+      "step": 11929
+    },
+    {
+      "epoch": 23.86,
+      "grad_norm": 1.832159161567688,
+      "learning_rate": 2e-05,
+      "loss": 0.02932361,
+      "step": 11930
+    },
+    {
+      "epoch": 23.862,
+      "grad_norm": 1.9174375534057617,
+      "learning_rate": 2e-05,
+      "loss": 0.05652925,
+      "step": 11931
+    },
+    {
+      "epoch": 23.864,
+      "grad_norm": 1.8890222311019897,
+      "learning_rate": 2e-05,
+      "loss": 0.03878344,
+      "step": 11932
+    },
+    {
+      "epoch": 23.866,
+      "grad_norm": 1.596886157989502,
+      "learning_rate": 2e-05,
+      "loss": 0.047834,
+      "step": 11933
+    },
+    {
+      "epoch": 23.868,
+      "grad_norm": 1.2127768993377686,
+      "learning_rate": 2e-05,
+      "loss": 0.04310586,
+      "step": 11934
+    },
+    {
+      "epoch": 23.87,
+      "grad_norm": 1.5679130554199219,
+      "learning_rate": 2e-05,
+      "loss": 0.05463172,
+      "step": 11935
+    },
+    {
+      "epoch": 23.872,
+      "grad_norm": 2.985722303390503,
+      "learning_rate": 2e-05,
+      "loss": 0.05015947,
+      "step": 11936
+    },
+    {
+      "epoch": 23.874,
+      "grad_norm": 1.1504042148590088,
+      "learning_rate": 2e-05,
+      "loss": 0.04618395,
+      "step": 11937
+    },
+    {
+      "epoch": 23.876,
+      "grad_norm": 1.0759702920913696,
+      "learning_rate": 2e-05,
+      "loss": 0.03113378,
+      "step": 11938
+    },
+    {
+      "epoch": 23.878,
+      "grad_norm": 0.9902017116546631,
+      "learning_rate": 2e-05,
+      "loss": 0.03239877,
+      "step": 11939
+    },
+    {
+      "epoch": 23.88,
+      "grad_norm": 2.066293954849243,
+      "learning_rate": 2e-05,
+      "loss": 0.0501477,
+      "step": 11940
+    },
+    {
+      "epoch": 23.882,
+      "grad_norm": 1.0674852132797241,
+      "learning_rate": 2e-05,
+      "loss": 0.02988044,
+      "step": 11941
+    },
+    {
+      "epoch": 23.884,
+      "grad_norm": 1.0986425876617432,
+      "learning_rate": 2e-05,
+      "loss": 0.03874698,
+      "step": 11942
+    },
+    {
+      "epoch": 23.886,
+      "grad_norm": 2.511347770690918,
+      "learning_rate": 2e-05,
+      "loss": 0.04582223,
+      "step": 11943
+    },
+    {
+      "epoch": 23.888,
+      "grad_norm": 1.30561101436615,
+      "learning_rate": 2e-05,
+      "loss": 0.04915758,
+      "step": 11944
+    },
+    {
+      "epoch": 23.89,
+      "grad_norm": 1.0842196941375732,
+      "learning_rate": 2e-05,
+      "loss": 0.03920197,
+      "step": 11945
+    },
+    {
+      "epoch": 23.892,
+      "grad_norm": 2.458435535430908,
+      "learning_rate": 2e-05,
+      "loss": 0.06143365,
+      "step": 11946
+    },
+    {
+      "epoch": 23.894,
+      "grad_norm": 1.3508713245391846,
+      "learning_rate": 2e-05,
+      "loss": 0.04859567,
+      "step": 11947
+    },
+    {
+      "epoch": 23.896,
+      "grad_norm": 1.2235263586044312,
+      "learning_rate": 2e-05,
+      "loss": 0.04123156,
+      "step": 11948
+    },
+    {
+      "epoch": 23.898,
+      "grad_norm": 1.3440524339675903,
+      "learning_rate": 2e-05,
+      "loss": 0.05537596,
+      "step": 11949
+    },
+    {
+      "epoch": 23.9,
+      "grad_norm": 1.618463158607483,
+      "learning_rate": 2e-05,
+      "loss": 0.05320452,
+      "step": 11950
+    },
+    {
+      "epoch": 23.902,
+      "grad_norm": 1.1047896146774292,
+      "learning_rate": 2e-05,
+      "loss": 0.03042508,
+      "step": 11951
+    },
+    {
+      "epoch": 23.904,
+      "grad_norm": 1.4379069805145264,
+      "learning_rate": 2e-05,
+      "loss": 0.03938862,
+      "step": 11952
+    },
+    {
+      "epoch": 23.906,
+      "grad_norm": 1.2280552387237549,
+      "learning_rate": 2e-05,
+      "loss": 0.03531722,
+      "step": 11953
+    },
+    {
+      "epoch": 23.908,
+      "grad_norm": 1.7554948329925537,
+      "learning_rate": 2e-05,
+      "loss": 0.0457234,
+      "step": 11954
+    },
+    {
+      "epoch": 23.91,
+      "grad_norm": 1.4118913412094116,
+      "learning_rate": 2e-05,
+      "loss": 0.04296067,
+      "step": 11955
+    },
+    {
+      "epoch": 23.912,
+      "grad_norm": 1.3764146566390991,
+      "learning_rate": 2e-05,
+      "loss": 0.04469619,
+      "step": 11956
+    },
+    {
+      "epoch": 23.914,
+      "grad_norm": 1.2633867263793945,
+      "learning_rate": 2e-05,
+      "loss": 0.03698818,
+      "step": 11957
+    },
+    {
+      "epoch": 23.916,
+      "grad_norm": 0.9480093717575073,
+      "learning_rate": 2e-05,
+      "loss": 0.02930729,
+      "step": 11958
+    },
+    {
+      "epoch": 23.918,
+      "grad_norm": 1.1288198232650757,
+      "learning_rate": 2e-05,
+      "loss": 0.04000931,
+      "step": 11959
+    },
+    {
+      "epoch": 23.92,
+      "grad_norm": 1.1547532081604004,
+      "learning_rate": 2e-05,
+      "loss": 0.04611331,
+      "step": 11960
+    },
+    {
+      "epoch": 23.922,
+      "grad_norm": 1.3261470794677734,
+      "learning_rate": 2e-05,
+      "loss": 0.05121126,
+      "step": 11961
+    },
+    {
+      "epoch": 23.924,
+      "grad_norm": 1.592149019241333,
+      "learning_rate": 2e-05,
+      "loss": 0.03919889,
+      "step": 11962
+    },
+    {
+      "epoch": 23.926,
+      "grad_norm": 1.4399209022521973,
+      "learning_rate": 2e-05,
+      "loss": 0.0399067,
+      "step": 11963
+    },
+    {
+      "epoch": 23.928,
+      "grad_norm": 2.2152695655822754,
+      "learning_rate": 2e-05,
+      "loss": 0.05575511,
+      "step": 11964
+    },
+    {
+      "epoch": 23.93,
+      "grad_norm": 1.7007336616516113,
+      "learning_rate": 2e-05,
+      "loss": 0.04545318,
+      "step": 11965
+    },
+    {
+      "epoch": 23.932,
+      "grad_norm": 1.149708867073059,
+      "learning_rate": 2e-05,
+      "loss": 0.04423167,
+      "step": 11966
+    },
+    {
+      "epoch": 23.934,
+      "grad_norm": 1.4653066396713257,
+      "learning_rate": 2e-05,
+      "loss": 0.03144412,
+      "step": 11967
+    },
+    {
+      "epoch": 23.936,
+      "grad_norm": 1.416612982749939,
+      "learning_rate": 2e-05,
+      "loss": 0.0496158,
+      "step": 11968
+    },
+    {
+      "epoch": 23.938,
+      "grad_norm": 3.0709216594696045,
+      "learning_rate": 2e-05,
+      "loss": 0.03679985,
+      "step": 11969
+    },
+    {
+      "epoch": 23.94,
+      "grad_norm": 1.1610190868377686,
+      "learning_rate": 2e-05,
+      "loss": 0.02437357,
+      "step": 11970
+    },
+    {
+      "epoch": 23.942,
+      "grad_norm": 1.3058899641036987,
+      "learning_rate": 2e-05,
+      "loss": 0.03944077,
+      "step": 11971
+    },
+    {
+      "epoch": 23.944,
+      "grad_norm": 1.279998779296875,
+      "learning_rate": 2e-05,
+      "loss": 0.03772661,
+      "step": 11972
+    },
+    {
+      "epoch": 23.946,
+      "grad_norm": 3.4212002754211426,
+      "learning_rate": 2e-05,
+      "loss": 0.06490073,
+      "step": 11973
+    },
+    {
+      "epoch": 23.948,
+      "grad_norm": 1.8969491720199585,
+      "learning_rate": 2e-05,
+      "loss": 0.05955943,
+      "step": 11974
+    },
+    {
+      "epoch": 23.95,
+      "grad_norm": 1.3523305654525757,
+      "learning_rate": 2e-05,
+      "loss": 0.04351056,
+      "step": 11975
+    },
+    {
+      "epoch": 23.951999999999998,
+      "grad_norm": 2.597954273223877,
+      "learning_rate": 2e-05,
+      "loss": 0.06529416,
+      "step": 11976
+    },
+    {
+      "epoch": 23.954,
+      "grad_norm": 1.3287073373794556,
+      "learning_rate": 2e-05,
+      "loss": 0.02537181,
+      "step": 11977
+    },
+    {
+      "epoch": 23.956,
+      "grad_norm": 1.4415262937545776,
+      "learning_rate": 2e-05,
+      "loss": 0.04575014,
+      "step": 11978
+    },
+    {
+      "epoch": 23.958,
+      "grad_norm": 2.070765256881714,
+      "learning_rate": 2e-05,
+      "loss": 0.04959387,
+      "step": 11979
+    },
+    {
+      "epoch": 23.96,
+      "grad_norm": 1.198920726776123,
+      "learning_rate": 2e-05,
+      "loss": 0.04554899,
+      "step": 11980
+    },
+    {
+      "epoch": 23.962,
+      "grad_norm": 1.5052765607833862,
+      "learning_rate": 2e-05,
+      "loss": 0.0560311,
+      "step": 11981
+    },
+    {
+      "epoch": 23.964,
+      "grad_norm": 1.083158016204834,
+      "learning_rate": 2e-05,
+      "loss": 0.03436514,
+      "step": 11982
+    },
+    {
+      "epoch": 23.966,
+      "grad_norm": 1.2290107011795044,
+      "learning_rate": 2e-05,
+      "loss": 0.02736303,
+      "step": 11983
+    },
+    {
+      "epoch": 23.968,
+      "grad_norm": 2.753286361694336,
+      "learning_rate": 2e-05,
+      "loss": 0.05253476,
+      "step": 11984
+    },
+    {
+      "epoch": 23.97,
+      "grad_norm": 1.206758975982666,
+      "learning_rate": 2e-05,
+      "loss": 0.02033206,
+      "step": 11985
+    },
+    {
+      "epoch": 23.972,
+      "grad_norm": 2.5052928924560547,
+      "learning_rate": 2e-05,
+      "loss": 0.05282067,
+      "step": 11986
+    },
+    {
+      "epoch": 23.974,
+      "grad_norm": 1.4807002544403076,
+      "learning_rate": 2e-05,
+      "loss": 0.04118428,
+      "step": 11987
+    },
+    {
+      "epoch": 23.976,
+      "grad_norm": 0.930181622505188,
+      "learning_rate": 2e-05,
+      "loss": 0.0332891,
+      "step": 11988
+    },
+    {
+      "epoch": 23.978,
+      "grad_norm": 1.6639894247055054,
+      "learning_rate": 2e-05,
+      "loss": 0.04549947,
+      "step": 11989
+    },
+    {
+      "epoch": 23.98,
+      "grad_norm": 1.3116426467895508,
+      "learning_rate": 2e-05,
+      "loss": 0.03784752,
+      "step": 11990
+    },
+    {
+      "epoch": 23.982,
+      "grad_norm": 0.9380161166191101,
+      "learning_rate": 2e-05,
+      "loss": 0.02113626,
+      "step": 11991
+    },
+    {
+      "epoch": 23.984,
+      "grad_norm": 1.9898308515548706,
+      "learning_rate": 2e-05,
+      "loss": 0.05268819,
+      "step": 11992
+    },
+    {
+      "epoch": 23.986,
+      "grad_norm": 0.75873202085495,
+      "learning_rate": 2e-05,
+      "loss": 0.01814926,
+      "step": 11993
+    },
+    {
+      "epoch": 23.988,
+      "grad_norm": 1.49936842918396,
+      "learning_rate": 2e-05,
+      "loss": 0.03649104,
+      "step": 11994
+    },
+    {
+      "epoch": 23.99,
+      "grad_norm": 1.114337682723999,
+      "learning_rate": 2e-05,
+      "loss": 0.04498353,
+      "step": 11995
+    },
+    {
+      "epoch": 23.992,
+      "grad_norm": 1.0039514303207397,
+      "learning_rate": 2e-05,
+      "loss": 0.03449545,
+      "step": 11996
+    },
+    {
+      "epoch": 23.994,
+      "grad_norm": 0.9935739040374756,
+      "learning_rate": 2e-05,
+      "loss": 0.03035527,
+      "step": 11997
+    },
+    {
+      "epoch": 23.996,
+      "grad_norm": 1.7863175868988037,
+      "learning_rate": 2e-05,
+      "loss": 0.044944,
+      "step": 11998
+    },
+    {
+      "epoch": 23.998,
+      "grad_norm": 2.193424701690674,
+      "learning_rate": 2e-05,
+      "loss": 0.03365751,
+      "step": 11999
+    },
+    {
+      "epoch": 24.0,
+      "grad_norm": 1.198980450630188,
+      "learning_rate": 2e-05,
+      "loss": 0.02690567,
+      "step": 12000
+    },
+    {
+      "epoch": 24.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.978,
+        "AngleClassification_2": 0.998,
+        "AngleClassification_3": 0.9700598802395209,
+        "Equal_1": 0.996,
+        "Equal_2": 0.9800399201596807,
+        "Equal_3": 0.9800399201596807,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.998003992015968,
+        "Parallel_1": 0.9899799599198397,
+        "Parallel_2": 0.9959919839679359,
+        "Parallel_3": 0.992,
+        "Perpendicular_1": 1.0,
+        "Perpendicular_2": 0.988,
+        "Perpendicular_3": 0.8697394789579158,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 0.9976666666666667,
+        "PointLiesOnCircle_3": 0.986,
+        "PointLiesOnLine_1": 0.9959919839679359,
+        "PointLiesOnLine_2": 0.9919839679358717,
+        "PointLiesOnLine_3": 0.9940119760479041
+      },
+      "eval_runtime": 320.967,
+      "eval_samples_per_second": 32.714,
+      "eval_steps_per_second": 0.654,
+      "step": 12000
+    },
+    {
+      "epoch": 24.002,
+      "grad_norm": 1.232427716255188,
+      "learning_rate": 2e-05,
+      "loss": 0.0302996,
+      "step": 12001
+    },
+    {
+      "epoch": 24.004,
+      "grad_norm": 1.1565972566604614,
+      "learning_rate": 2e-05,
+      "loss": 0.0323018,
+      "step": 12002
+    },
+    {
+      "epoch": 24.006,
+      "grad_norm": 2.3447890281677246,
+      "learning_rate": 2e-05,
+      "loss": 0.04620555,
+      "step": 12003
+    },
+    {
+      "epoch": 24.008,
+      "grad_norm": 1.4546306133270264,
+      "learning_rate": 2e-05,
+      "loss": 0.04248685,
+      "step": 12004
+    },
+    {
+      "epoch": 24.01,
+      "grad_norm": 2.750286340713501,
+      "learning_rate": 2e-05,
+      "loss": 0.05949475,
+      "step": 12005
+    },
+    {
+      "epoch": 24.012,
+      "grad_norm": 2.553252935409546,
+      "learning_rate": 2e-05,
+      "loss": 0.05698572,
+      "step": 12006
+    },
+    {
+      "epoch": 24.014,
+      "grad_norm": 1.3687764406204224,
+      "learning_rate": 2e-05,
+      "loss": 0.05525066,
+      "step": 12007
+    },
+    {
+      "epoch": 24.016,
+      "grad_norm": 1.277368187904358,
+      "learning_rate": 2e-05,
+      "loss": 0.05651337,
+      "step": 12008
+    },
+    {
+      "epoch": 24.018,
+      "grad_norm": 1.1808255910873413,
+      "learning_rate": 2e-05,
+      "loss": 0.04304896,
+      "step": 12009
+    },
+    {
+      "epoch": 24.02,
+      "grad_norm": 1.2423754930496216,
+      "learning_rate": 2e-05,
+      "loss": 0.04889252,
+      "step": 12010
+    },
+    {
+      "epoch": 24.022,
+      "grad_norm": 2.5050649642944336,
+      "learning_rate": 2e-05,
+      "loss": 0.05278215,
+      "step": 12011
+    },
+    {
+      "epoch": 24.024,
+      "grad_norm": 2.068812131881714,
+      "learning_rate": 2e-05,
+      "loss": 0.04776769,
+      "step": 12012
+    },
+    {
+      "epoch": 24.026,
+      "grad_norm": 1.2705119848251343,
+      "learning_rate": 2e-05,
+      "loss": 0.06349517,
+      "step": 12013
+    },
+    {
+      "epoch": 24.028,
+      "grad_norm": 1.3783538341522217,
+      "learning_rate": 2e-05,
+      "loss": 0.04796961,
+      "step": 12014
+    },
+    {
+      "epoch": 24.03,
+      "grad_norm": 1.4065525531768799,
+      "learning_rate": 2e-05,
+      "loss": 0.05514029,
+      "step": 12015
+    },
+    {
+      "epoch": 24.032,
+      "grad_norm": 2.559398889541626,
+      "learning_rate": 2e-05,
+      "loss": 0.04297279,
+      "step": 12016
+    },
+    {
+      "epoch": 24.034,
+      "grad_norm": 1.3823760747909546,
+      "learning_rate": 2e-05,
+      "loss": 0.05167028,
+      "step": 12017
+    },
+    {
+      "epoch": 24.036,
+      "grad_norm": 0.9577856659889221,
+      "learning_rate": 2e-05,
+      "loss": 0.02910194,
+      "step": 12018
+    },
+    {
+      "epoch": 24.038,
+      "grad_norm": 1.0815929174423218,
+      "learning_rate": 2e-05,
+      "loss": 0.03755962,
+      "step": 12019
+    },
+    {
+      "epoch": 24.04,
+      "grad_norm": 1.190253734588623,
+      "learning_rate": 2e-05,
+      "loss": 0.03743257,
+      "step": 12020
+    },
+    {
+      "epoch": 24.042,
+      "grad_norm": 2.5249247550964355,
+      "learning_rate": 2e-05,
+      "loss": 0.05092564,
+      "step": 12021
+    },
+    {
+      "epoch": 24.044,
+      "grad_norm": 1.0962779521942139,
+      "learning_rate": 2e-05,
+      "loss": 0.04152154,
+      "step": 12022
+    },
+    {
+      "epoch": 24.046,
+      "grad_norm": 1.5275741815567017,
+      "learning_rate": 2e-05,
+      "loss": 0.04184947,
+      "step": 12023
+    },
+    {
+      "epoch": 24.048,
+      "grad_norm": 1.0383355617523193,
+      "learning_rate": 2e-05,
+      "loss": 0.03165521,
+      "step": 12024
+    },
+    {
+      "epoch": 24.05,
+      "grad_norm": 1.5432151556015015,
+      "learning_rate": 2e-05,
+      "loss": 0.05171941,
+      "step": 12025
+    },
+    {
+      "epoch": 24.052,
+      "grad_norm": 1.249215841293335,
+      "learning_rate": 2e-05,
+      "loss": 0.04950016,
+      "step": 12026
+    },
+    {
+      "epoch": 24.054,
+      "grad_norm": 1.309938907623291,
+      "learning_rate": 2e-05,
+      "loss": 0.04960479,
+      "step": 12027
+    },
+    {
+      "epoch": 24.056,
+      "grad_norm": 0.9974091053009033,
+      "learning_rate": 2e-05,
+      "loss": 0.03317707,
+      "step": 12028
+    },
+    {
+      "epoch": 24.058,
+      "grad_norm": 1.5228265523910522,
+      "learning_rate": 2e-05,
+      "loss": 0.04856505,
+      "step": 12029
+    },
+    {
+      "epoch": 24.06,
+      "grad_norm": 1.8886607885360718,
+      "learning_rate": 2e-05,
+      "loss": 0.05328999,
+      "step": 12030
+    },
+    {
+      "epoch": 24.062,
+      "grad_norm": 1.02639639377594,
+      "learning_rate": 2e-05,
+      "loss": 0.02694909,
+      "step": 12031
+    },
+    {
+      "epoch": 24.064,
+      "grad_norm": 1.0701245069503784,
+      "learning_rate": 2e-05,
+      "loss": 0.04255382,
+      "step": 12032
+    },
+    {
+      "epoch": 24.066,
+      "grad_norm": 1.3096833229064941,
+      "learning_rate": 2e-05,
+      "loss": 0.04552224,
+      "step": 12033
+    },
+    {
+      "epoch": 24.068,
+      "grad_norm": 1.0245039463043213,
+      "learning_rate": 2e-05,
+      "loss": 0.03121547,
+      "step": 12034
+    },
+    {
+      "epoch": 24.07,
+      "grad_norm": 2.104153871536255,
+      "learning_rate": 2e-05,
+      "loss": 0.05049267,
+      "step": 12035
+    },
+    {
+      "epoch": 24.072,
+      "grad_norm": 0.9794881343841553,
+      "learning_rate": 2e-05,
+      "loss": 0.04013614,
+      "step": 12036
+    },
+    {
+      "epoch": 24.074,
+      "grad_norm": 1.3545254468917847,
+      "learning_rate": 2e-05,
+      "loss": 0.04678908,
+      "step": 12037
+    },
+    {
+      "epoch": 24.076,
+      "grad_norm": 1.7231847047805786,
+      "learning_rate": 2e-05,
+      "loss": 0.05604774,
+      "step": 12038
+    },
+    {
+      "epoch": 24.078,
+      "grad_norm": 1.0859684944152832,
+      "learning_rate": 2e-05,
+      "loss": 0.02874695,
+      "step": 12039
+    },
+    {
+      "epoch": 24.08,
+      "grad_norm": 1.4772136211395264,
+      "learning_rate": 2e-05,
+      "loss": 0.04590698,
+      "step": 12040
+    },
+    {
+      "epoch": 24.082,
+      "grad_norm": 1.1300466060638428,
+      "learning_rate": 2e-05,
+      "loss": 0.03713751,
+      "step": 12041
+    },
+    {
+      "epoch": 24.084,
+      "grad_norm": 1.4452584981918335,
+      "learning_rate": 2e-05,
+      "loss": 0.04535814,
+      "step": 12042
+    },
+    {
+      "epoch": 24.086,
+      "grad_norm": 1.703972578048706,
+      "learning_rate": 2e-05,
+      "loss": 0.05336364,
+      "step": 12043
+    },
+    {
+      "epoch": 24.088,
+      "grad_norm": 1.5008318424224854,
+      "learning_rate": 2e-05,
+      "loss": 0.0530287,
+      "step": 12044
+    },
+    {
+      "epoch": 24.09,
+      "grad_norm": 1.0992947816848755,
+      "learning_rate": 2e-05,
+      "loss": 0.05150533,
+      "step": 12045
+    },
+    {
+      "epoch": 24.092,
+      "grad_norm": 1.1373698711395264,
+      "learning_rate": 2e-05,
+      "loss": 0.04210654,
+      "step": 12046
+    },
+    {
+      "epoch": 24.094,
+      "grad_norm": 1.3362079858779907,
+      "learning_rate": 2e-05,
+      "loss": 0.04046137,
+      "step": 12047
+    },
+    {
+      "epoch": 24.096,
+      "grad_norm": 2.110414743423462,
+      "learning_rate": 2e-05,
+      "loss": 0.05383376,
+      "step": 12048
+    },
+    {
+      "epoch": 24.098,
+      "grad_norm": 2.22654128074646,
+      "learning_rate": 2e-05,
+      "loss": 0.04562001,
+      "step": 12049
+    },
+    {
+      "epoch": 24.1,
+      "grad_norm": 1.5892013311386108,
+      "learning_rate": 2e-05,
+      "loss": 0.04545607,
+      "step": 12050
+    },
+    {
+      "epoch": 24.102,
+      "grad_norm": 1.7397476434707642,
+      "learning_rate": 2e-05,
+      "loss": 0.05527458,
+      "step": 12051
+    },
+    {
+      "epoch": 24.104,
+      "grad_norm": 1.3350776433944702,
+      "learning_rate": 2e-05,
+      "loss": 0.04359471,
+      "step": 12052
+    },
+    {
+      "epoch": 24.106,
+      "grad_norm": 1.240904450416565,
+      "learning_rate": 2e-05,
+      "loss": 0.04825516,
+      "step": 12053
+    },
+    {
+      "epoch": 24.108,
+      "grad_norm": 1.1014329195022583,
+      "learning_rate": 2e-05,
+      "loss": 0.0354472,
+      "step": 12054
+    },
+    {
+      "epoch": 24.11,
+      "grad_norm": 1.3946998119354248,
+      "learning_rate": 2e-05,
+      "loss": 0.04494639,
+      "step": 12055
+    },
+    {
+      "epoch": 24.112,
+      "grad_norm": 1.7110241651535034,
+      "learning_rate": 2e-05,
+      "loss": 0.05996033,
+      "step": 12056
+    },
+    {
+      "epoch": 24.114,
+      "grad_norm": 1.3246296644210815,
+      "learning_rate": 2e-05,
+      "loss": 0.0507584,
+      "step": 12057
+    },
+    {
+      "epoch": 24.116,
+      "grad_norm": 1.423532247543335,
+      "learning_rate": 2e-05,
+      "loss": 0.04155703,
+      "step": 12058
+    },
+    {
+      "epoch": 24.118,
+      "grad_norm": 1.6252057552337646,
+      "learning_rate": 2e-05,
+      "loss": 0.05566594,
+      "step": 12059
+    },
+    {
+      "epoch": 24.12,
+      "grad_norm": 1.978572964668274,
+      "learning_rate": 2e-05,
+      "loss": 0.06226615,
+      "step": 12060
+    },
+    {
+      "epoch": 24.122,
+      "grad_norm": 1.053499460220337,
+      "learning_rate": 2e-05,
+      "loss": 0.03881479,
+      "step": 12061
+    },
+    {
+      "epoch": 24.124,
+      "grad_norm": 1.3819252252578735,
+      "learning_rate": 2e-05,
+      "loss": 0.05036075,
+      "step": 12062
+    },
+    {
+      "epoch": 24.126,
+      "grad_norm": 1.7073538303375244,
+      "learning_rate": 2e-05,
+      "loss": 0.05705138,
+      "step": 12063
+    },
+    {
+      "epoch": 24.128,
+      "grad_norm": 1.606673002243042,
+      "learning_rate": 2e-05,
+      "loss": 0.04539119,
+      "step": 12064
+    },
+    {
+      "epoch": 24.13,
+      "grad_norm": 1.0602710247039795,
+      "learning_rate": 2e-05,
+      "loss": 0.04797367,
+      "step": 12065
+    },
+    {
+      "epoch": 24.132,
+      "grad_norm": 1.3240618705749512,
+      "learning_rate": 2e-05,
+      "loss": 0.05344753,
+      "step": 12066
+    },
+    {
+      "epoch": 24.134,
+      "grad_norm": 1.652435541152954,
+      "learning_rate": 2e-05,
+      "loss": 0.03491877,
+      "step": 12067
+    },
+    {
+      "epoch": 24.136,
+      "grad_norm": 2.6440341472625732,
+      "learning_rate": 2e-05,
+      "loss": 0.05096871,
+      "step": 12068
+    },
+    {
+      "epoch": 24.138,
+      "grad_norm": 1.0971581935882568,
+      "learning_rate": 2e-05,
+      "loss": 0.04188203,
+      "step": 12069
+    },
+    {
+      "epoch": 24.14,
+      "grad_norm": 0.9165200591087341,
+      "learning_rate": 2e-05,
+      "loss": 0.03046097,
+      "step": 12070
+    },
+    {
+      "epoch": 24.142,
+      "grad_norm": 2.975809097290039,
+      "learning_rate": 2e-05,
+      "loss": 0.05606359,
+      "step": 12071
+    },
+    {
+      "epoch": 24.144,
+      "grad_norm": 1.127575397491455,
+      "learning_rate": 2e-05,
+      "loss": 0.0311743,
+      "step": 12072
+    },
+    {
+      "epoch": 24.146,
+      "grad_norm": 1.337193489074707,
+      "learning_rate": 2e-05,
+      "loss": 0.03485664,
+      "step": 12073
+    },
+    {
+      "epoch": 24.148,
+      "grad_norm": 2.253185510635376,
+      "learning_rate": 2e-05,
+      "loss": 0.04143036,
+      "step": 12074
+    },
+    {
+      "epoch": 24.15,
+      "grad_norm": 2.4968769550323486,
+      "learning_rate": 2e-05,
+      "loss": 0.0586257,
+      "step": 12075
+    },
+    {
+      "epoch": 24.152,
+      "grad_norm": 1.3548872470855713,
+      "learning_rate": 2e-05,
+      "loss": 0.04275392,
+      "step": 12076
+    },
+    {
+      "epoch": 24.154,
+      "grad_norm": 1.05968177318573,
+      "learning_rate": 2e-05,
+      "loss": 0.03807945,
+      "step": 12077
+    },
+    {
+      "epoch": 24.156,
+      "grad_norm": 1.2706948518753052,
+      "learning_rate": 2e-05,
+      "loss": 0.05273522,
+      "step": 12078
+    },
+    {
+      "epoch": 24.158,
+      "grad_norm": 1.2641117572784424,
+      "learning_rate": 2e-05,
+      "loss": 0.05271117,
+      "step": 12079
+    },
+    {
+      "epoch": 24.16,
+      "grad_norm": 1.2767136096954346,
+      "learning_rate": 2e-05,
+      "loss": 0.03717405,
+      "step": 12080
+    },
+    {
+      "epoch": 24.162,
+      "grad_norm": 1.8129770755767822,
+      "learning_rate": 2e-05,
+      "loss": 0.05442118,
+      "step": 12081
+    },
+    {
+      "epoch": 24.164,
+      "grad_norm": 1.6097712516784668,
+      "learning_rate": 2e-05,
+      "loss": 0.02836596,
+      "step": 12082
+    },
+    {
+      "epoch": 24.166,
+      "grad_norm": 1.6396865844726562,
+      "learning_rate": 2e-05,
+      "loss": 0.0487038,
+      "step": 12083
+    },
+    {
+      "epoch": 24.168,
+      "grad_norm": 5.180088520050049,
+      "learning_rate": 2e-05,
+      "loss": 0.07057432,
+      "step": 12084
+    },
+    {
+      "epoch": 24.17,
+      "grad_norm": 3.231645345687866,
+      "learning_rate": 2e-05,
+      "loss": 0.0679914,
+      "step": 12085
+    },
+    {
+      "epoch": 24.172,
+      "grad_norm": 2.2453713417053223,
+      "learning_rate": 2e-05,
+      "loss": 0.0494958,
+      "step": 12086
+    },
+    {
+      "epoch": 24.174,
+      "grad_norm": 1.019257664680481,
+      "learning_rate": 2e-05,
+      "loss": 0.03262502,
+      "step": 12087
+    },
+    {
+      "epoch": 24.176,
+      "grad_norm": 1.8194106817245483,
+      "learning_rate": 2e-05,
+      "loss": 0.04817811,
+      "step": 12088
+    },
+    {
+      "epoch": 24.178,
+      "grad_norm": 1.1825400590896606,
+      "learning_rate": 2e-05,
+      "loss": 0.04405992,
+      "step": 12089
+    },
+    {
+      "epoch": 24.18,
+      "grad_norm": 0.9297054409980774,
+      "learning_rate": 2e-05,
+      "loss": 0.03029897,
+      "step": 12090
+    },
+    {
+      "epoch": 24.182,
+      "grad_norm": 1.7777191400527954,
+      "learning_rate": 2e-05,
+      "loss": 0.04615546,
+      "step": 12091
+    },
+    {
+      "epoch": 24.184,
+      "grad_norm": 3.235630750656128,
+      "learning_rate": 2e-05,
+      "loss": 0.05811172,
+      "step": 12092
+    },
+    {
+      "epoch": 24.186,
+      "grad_norm": 1.1768110990524292,
+      "learning_rate": 2e-05,
+      "loss": 0.03787137,
+      "step": 12093
+    },
+    {
+      "epoch": 24.188,
+      "grad_norm": 4.949007034301758,
+      "learning_rate": 2e-05,
+      "loss": 0.03465776,
+      "step": 12094
+    },
+    {
+      "epoch": 24.19,
+      "grad_norm": 1.4985238313674927,
+      "learning_rate": 2e-05,
+      "loss": 0.04126935,
+      "step": 12095
+    },
+    {
+      "epoch": 24.192,
+      "grad_norm": 2.3714089393615723,
+      "learning_rate": 2e-05,
+      "loss": 0.03277493,
+      "step": 12096
+    },
+    {
+      "epoch": 24.194,
+      "grad_norm": 1.5087124109268188,
+      "learning_rate": 2e-05,
+      "loss": 0.05334678,
+      "step": 12097
+    },
+    {
+      "epoch": 24.196,
+      "grad_norm": 1.2616825103759766,
+      "learning_rate": 2e-05,
+      "loss": 0.04551134,
+      "step": 12098
+    },
+    {
+      "epoch": 24.198,
+      "grad_norm": 1.158451795578003,
+      "learning_rate": 2e-05,
+      "loss": 0.04630031,
+      "step": 12099
+    },
+    {
+      "epoch": 24.2,
+      "grad_norm": 1.520171880722046,
+      "learning_rate": 2e-05,
+      "loss": 0.03448167,
+      "step": 12100
+    },
+    {
+      "epoch": 24.202,
+      "grad_norm": 5.375433921813965,
+      "learning_rate": 2e-05,
+      "loss": 0.05557346,
+      "step": 12101
+    },
+    {
+      "epoch": 24.204,
+      "grad_norm": 2.1798291206359863,
+      "learning_rate": 2e-05,
+      "loss": 0.05976859,
+      "step": 12102
+    },
+    {
+      "epoch": 24.206,
+      "grad_norm": 1.5132516622543335,
+      "learning_rate": 2e-05,
+      "loss": 0.04973812,
+      "step": 12103
+    },
+    {
+      "epoch": 24.208,
+      "grad_norm": 1.4551432132720947,
+      "learning_rate": 2e-05,
+      "loss": 0.03925546,
+      "step": 12104
+    },
+    {
+      "epoch": 24.21,
+      "grad_norm": 1.4387096166610718,
+      "learning_rate": 2e-05,
+      "loss": 0.04110787,
+      "step": 12105
+    },
+    {
+      "epoch": 24.212,
+      "grad_norm": 0.9817237854003906,
+      "learning_rate": 2e-05,
+      "loss": 0.043275,
+      "step": 12106
+    },
+    {
+      "epoch": 24.214,
+      "grad_norm": 0.899992048740387,
+      "learning_rate": 2e-05,
+      "loss": 0.02760095,
+      "step": 12107
+    },
+    {
+      "epoch": 24.216,
+      "grad_norm": 1.9981087446212769,
+      "learning_rate": 2e-05,
+      "loss": 0.08154525,
+      "step": 12108
+    },
+    {
+      "epoch": 24.218,
+      "grad_norm": 1.6760551929473877,
+      "learning_rate": 2e-05,
+      "loss": 0.05679325,
+      "step": 12109
+    },
+    {
+      "epoch": 24.22,
+      "grad_norm": 1.3508455753326416,
+      "learning_rate": 2e-05,
+      "loss": 0.05091976,
+      "step": 12110
+    },
+    {
+      "epoch": 24.222,
+      "grad_norm": 1.8912901878356934,
+      "learning_rate": 2e-05,
+      "loss": 0.06227458,
+      "step": 12111
+    },
+    {
+      "epoch": 24.224,
+      "grad_norm": 1.3273926973342896,
+      "learning_rate": 2e-05,
+      "loss": 0.05052402,
+      "step": 12112
+    },
+    {
+      "epoch": 24.226,
+      "grad_norm": 1.2464828491210938,
+      "learning_rate": 2e-05,
+      "loss": 0.05788217,
+      "step": 12113
+    },
+    {
+      "epoch": 24.228,
+      "grad_norm": 2.049316883087158,
+      "learning_rate": 2e-05,
+      "loss": 0.06217295,
+      "step": 12114
+    },
+    {
+      "epoch": 24.23,
+      "grad_norm": 1.428179144859314,
+      "learning_rate": 2e-05,
+      "loss": 0.05017955,
+      "step": 12115
+    },
+    {
+      "epoch": 24.232,
+      "grad_norm": 1.767197847366333,
+      "learning_rate": 2e-05,
+      "loss": 0.05077636,
+      "step": 12116
+    },
+    {
+      "epoch": 24.234,
+      "grad_norm": 1.231481671333313,
+      "learning_rate": 2e-05,
+      "loss": 0.05873109,
+      "step": 12117
+    },
+    {
+      "epoch": 24.236,
+      "grad_norm": 1.6732200384140015,
+      "learning_rate": 2e-05,
+      "loss": 0.04653981,
+      "step": 12118
+    },
+    {
+      "epoch": 24.238,
+      "grad_norm": 1.0597010850906372,
+      "learning_rate": 2e-05,
+      "loss": 0.03581889,
+      "step": 12119
+    },
+    {
+      "epoch": 24.24,
+      "grad_norm": 1.2121529579162598,
+      "learning_rate": 2e-05,
+      "loss": 0.04930365,
+      "step": 12120
+    },
+    {
+      "epoch": 24.242,
+      "grad_norm": 1.3998339176177979,
+      "learning_rate": 2e-05,
+      "loss": 0.03907991,
+      "step": 12121
+    },
+    {
+      "epoch": 24.244,
+      "grad_norm": 1.2382348775863647,
+      "learning_rate": 2e-05,
+      "loss": 0.05211905,
+      "step": 12122
+    },
+    {
+      "epoch": 24.246,
+      "grad_norm": 1.1277509927749634,
+      "learning_rate": 2e-05,
+      "loss": 0.03834852,
+      "step": 12123
+    },
+    {
+      "epoch": 24.248,
+      "grad_norm": 1.2743759155273438,
+      "learning_rate": 2e-05,
+      "loss": 0.0476411,
+      "step": 12124
+    },
+    {
+      "epoch": 24.25,
+      "grad_norm": 1.2436480522155762,
+      "learning_rate": 2e-05,
+      "loss": 0.0411014,
+      "step": 12125
+    },
+    {
+      "epoch": 24.252,
+      "grad_norm": 1.9588987827301025,
+      "learning_rate": 2e-05,
+      "loss": 0.06562616,
+      "step": 12126
+    },
+    {
+      "epoch": 24.254,
+      "grad_norm": 1.318472146987915,
+      "learning_rate": 2e-05,
+      "loss": 0.04503114,
+      "step": 12127
+    },
+    {
+      "epoch": 24.256,
+      "grad_norm": 1.3602678775787354,
+      "learning_rate": 2e-05,
+      "loss": 0.02978888,
+      "step": 12128
+    },
+    {
+      "epoch": 24.258,
+      "grad_norm": 1.2801728248596191,
+      "learning_rate": 2e-05,
+      "loss": 0.03404569,
+      "step": 12129
+    },
+    {
+      "epoch": 24.26,
+      "grad_norm": 1.8510991334915161,
+      "learning_rate": 2e-05,
+      "loss": 0.041315,
+      "step": 12130
+    },
+    {
+      "epoch": 24.262,
+      "grad_norm": 1.1254820823669434,
+      "learning_rate": 2e-05,
+      "loss": 0.04042017,
+      "step": 12131
+    },
+    {
+      "epoch": 24.264,
+      "grad_norm": 1.4146277904510498,
+      "learning_rate": 2e-05,
+      "loss": 0.04251551,
+      "step": 12132
+    },
+    {
+      "epoch": 24.266,
+      "grad_norm": 1.4054226875305176,
+      "learning_rate": 2e-05,
+      "loss": 0.03718261,
+      "step": 12133
+    },
+    {
+      "epoch": 24.268,
+      "grad_norm": 1.2703016996383667,
+      "learning_rate": 2e-05,
+      "loss": 0.0467224,
+      "step": 12134
+    },
+    {
+      "epoch": 24.27,
+      "grad_norm": 1.6173110008239746,
+      "learning_rate": 2e-05,
+      "loss": 0.05684683,
+      "step": 12135
+    },
+    {
+      "epoch": 24.272,
+      "grad_norm": 2.035446882247925,
+      "learning_rate": 2e-05,
+      "loss": 0.04400192,
+      "step": 12136
+    },
+    {
+      "epoch": 24.274,
+      "grad_norm": 1.105391502380371,
+      "learning_rate": 2e-05,
+      "loss": 0.04434478,
+      "step": 12137
+    },
+    {
+      "epoch": 24.276,
+      "grad_norm": 1.2330455780029297,
+      "learning_rate": 2e-05,
+      "loss": 0.05438062,
+      "step": 12138
+    },
+    {
+      "epoch": 24.278,
+      "grad_norm": 1.1647205352783203,
+      "learning_rate": 2e-05,
+      "loss": 0.03314302,
+      "step": 12139
+    },
+    {
+      "epoch": 24.28,
+      "grad_norm": 1.1791774034500122,
+      "learning_rate": 2e-05,
+      "loss": 0.05466758,
+      "step": 12140
+    },
+    {
+      "epoch": 24.282,
+      "grad_norm": 1.4208977222442627,
+      "learning_rate": 2e-05,
+      "loss": 0.03269932,
+      "step": 12141
+    },
+    {
+      "epoch": 24.284,
+      "grad_norm": 1.1843621730804443,
+      "learning_rate": 2e-05,
+      "loss": 0.04443068,
+      "step": 12142
+    },
+    {
+      "epoch": 24.286,
+      "grad_norm": 1.522578239440918,
+      "learning_rate": 2e-05,
+      "loss": 0.06013615,
+      "step": 12143
+    },
+    {
+      "epoch": 24.288,
+      "grad_norm": 1.4927306175231934,
+      "learning_rate": 2e-05,
+      "loss": 0.04813029,
+      "step": 12144
+    },
+    {
+      "epoch": 24.29,
+      "grad_norm": 1.2117818593978882,
+      "learning_rate": 2e-05,
+      "loss": 0.03942649,
+      "step": 12145
+    },
+    {
+      "epoch": 24.292,
+      "grad_norm": 1.0251680612564087,
+      "learning_rate": 2e-05,
+      "loss": 0.03021972,
+      "step": 12146
+    },
+    {
+      "epoch": 24.294,
+      "grad_norm": 1.3790630102157593,
+      "learning_rate": 2e-05,
+      "loss": 0.05024281,
+      "step": 12147
+    },
+    {
+      "epoch": 24.296,
+      "grad_norm": 1.5437334775924683,
+      "learning_rate": 2e-05,
+      "loss": 0.03962575,
+      "step": 12148
+    },
+    {
+      "epoch": 24.298,
+      "grad_norm": 1.204317569732666,
+      "learning_rate": 2e-05,
+      "loss": 0.03893788,
+      "step": 12149
+    },
+    {
+      "epoch": 24.3,
+      "grad_norm": 1.1474257707595825,
+      "learning_rate": 2e-05,
+      "loss": 0.04675932,
+      "step": 12150
+    },
+    {
+      "epoch": 24.302,
+      "grad_norm": 0.9916406869888306,
+      "learning_rate": 2e-05,
+      "loss": 0.03100633,
+      "step": 12151
+    },
+    {
+      "epoch": 24.304,
+      "grad_norm": 1.9832628965377808,
+      "learning_rate": 2e-05,
+      "loss": 0.05169064,
+      "step": 12152
+    },
+    {
+      "epoch": 24.306,
+      "grad_norm": 1.811354398727417,
+      "learning_rate": 2e-05,
+      "loss": 0.05321472,
+      "step": 12153
+    },
+    {
+      "epoch": 24.308,
+      "grad_norm": 1.277738094329834,
+      "learning_rate": 2e-05,
+      "loss": 0.04673781,
+      "step": 12154
+    },
+    {
+      "epoch": 24.31,
+      "grad_norm": 1.3241530656814575,
+      "learning_rate": 2e-05,
+      "loss": 0.04513844,
+      "step": 12155
+    },
+    {
+      "epoch": 24.312,
+      "grad_norm": 2.2016355991363525,
+      "learning_rate": 2e-05,
+      "loss": 0.03816313,
+      "step": 12156
+    },
+    {
+      "epoch": 24.314,
+      "grad_norm": 1.972180962562561,
+      "learning_rate": 2e-05,
+      "loss": 0.04108272,
+      "step": 12157
+    },
+    {
+      "epoch": 24.316,
+      "grad_norm": 1.1562265157699585,
+      "learning_rate": 2e-05,
+      "loss": 0.04686077,
+      "step": 12158
+    },
+    {
+      "epoch": 24.318,
+      "grad_norm": 1.2996026277542114,
+      "learning_rate": 2e-05,
+      "loss": 0.04213938,
+      "step": 12159
+    },
+    {
+      "epoch": 24.32,
+      "grad_norm": 2.028224468231201,
+      "learning_rate": 2e-05,
+      "loss": 0.05159696,
+      "step": 12160
+    },
+    {
+      "epoch": 24.322,
+      "grad_norm": 1.2881364822387695,
+      "learning_rate": 2e-05,
+      "loss": 0.04292353,
+      "step": 12161
+    },
+    {
+      "epoch": 24.324,
+      "grad_norm": 1.4620164632797241,
+      "learning_rate": 2e-05,
+      "loss": 0.04024596,
+      "step": 12162
+    },
+    {
+      "epoch": 24.326,
+      "grad_norm": 1.6173001527786255,
+      "learning_rate": 2e-05,
+      "loss": 0.04233917,
+      "step": 12163
+    },
+    {
+      "epoch": 24.328,
+      "grad_norm": 1.7425227165222168,
+      "learning_rate": 2e-05,
+      "loss": 0.03686478,
+      "step": 12164
+    },
+    {
+      "epoch": 24.33,
+      "grad_norm": 1.112245798110962,
+      "learning_rate": 2e-05,
+      "loss": 0.04378101,
+      "step": 12165
+    },
+    {
+      "epoch": 24.332,
+      "grad_norm": 2.1608293056488037,
+      "learning_rate": 2e-05,
+      "loss": 0.04491881,
+      "step": 12166
+    },
+    {
+      "epoch": 24.334,
+      "grad_norm": 1.1910284757614136,
+      "learning_rate": 2e-05,
+      "loss": 0.03951531,
+      "step": 12167
+    },
+    {
+      "epoch": 24.336,
+      "grad_norm": 1.9994529485702515,
+      "learning_rate": 2e-05,
+      "loss": 0.04223923,
+      "step": 12168
+    },
+    {
+      "epoch": 24.338,
+      "grad_norm": 1.0348660945892334,
+      "learning_rate": 2e-05,
+      "loss": 0.03475592,
+      "step": 12169
+    },
+    {
+      "epoch": 24.34,
+      "grad_norm": 3.0013294219970703,
+      "learning_rate": 2e-05,
+      "loss": 0.06690484,
+      "step": 12170
+    },
+    {
+      "epoch": 24.342,
+      "grad_norm": 1.6031739711761475,
+      "learning_rate": 2e-05,
+      "loss": 0.04361062,
+      "step": 12171
+    },
+    {
+      "epoch": 24.344,
+      "grad_norm": 2.190584182739258,
+      "learning_rate": 2e-05,
+      "loss": 0.04059751,
+      "step": 12172
+    },
+    {
+      "epoch": 24.346,
+      "grad_norm": 1.5847196578979492,
+      "learning_rate": 2e-05,
+      "loss": 0.03874936,
+      "step": 12173
+    },
+    {
+      "epoch": 24.348,
+      "grad_norm": 2.4007716178894043,
+      "learning_rate": 2e-05,
+      "loss": 0.04349686,
+      "step": 12174
+    },
+    {
+      "epoch": 24.35,
+      "grad_norm": 3.5807485580444336,
+      "learning_rate": 2e-05,
+      "loss": 0.04612588,
+      "step": 12175
+    },
+    {
+      "epoch": 24.352,
+      "grad_norm": 1.5115255117416382,
+      "learning_rate": 2e-05,
+      "loss": 0.05002192,
+      "step": 12176
+    },
+    {
+      "epoch": 24.354,
+      "grad_norm": 1.3225903511047363,
+      "learning_rate": 2e-05,
+      "loss": 0.04832003,
+      "step": 12177
+    },
+    {
+      "epoch": 24.356,
+      "grad_norm": 1.0971873998641968,
+      "learning_rate": 2e-05,
+      "loss": 0.038155,
+      "step": 12178
+    },
+    {
+      "epoch": 24.358,
+      "grad_norm": 0.9766896367073059,
+      "learning_rate": 2e-05,
+      "loss": 0.0364832,
+      "step": 12179
+    },
+    {
+      "epoch": 24.36,
+      "grad_norm": 2.5646822452545166,
+      "learning_rate": 2e-05,
+      "loss": 0.04737934,
+      "step": 12180
+    },
+    {
+      "epoch": 24.362,
+      "grad_norm": 1.9678560495376587,
+      "learning_rate": 2e-05,
+      "loss": 0.03837503,
+      "step": 12181
+    },
+    {
+      "epoch": 24.364,
+      "grad_norm": 1.550510048866272,
+      "learning_rate": 2e-05,
+      "loss": 0.05680725,
+      "step": 12182
+    },
+    {
+      "epoch": 24.366,
+      "grad_norm": 1.2245664596557617,
+      "learning_rate": 2e-05,
+      "loss": 0.04190246,
+      "step": 12183
+    },
+    {
+      "epoch": 24.368,
+      "grad_norm": 1.0684999227523804,
+      "learning_rate": 2e-05,
+      "loss": 0.04685017,
+      "step": 12184
+    },
+    {
+      "epoch": 24.37,
+      "grad_norm": 1.253795862197876,
+      "learning_rate": 2e-05,
+      "loss": 0.05009492,
+      "step": 12185
+    },
+    {
+      "epoch": 24.372,
+      "grad_norm": 1.1181055307388306,
+      "learning_rate": 2e-05,
+      "loss": 0.03659803,
+      "step": 12186
+    },
+    {
+      "epoch": 24.374,
+      "grad_norm": 2.1159071922302246,
+      "learning_rate": 2e-05,
+      "loss": 0.04460203,
+      "step": 12187
+    },
+    {
+      "epoch": 24.376,
+      "grad_norm": 1.5088142156600952,
+      "learning_rate": 2e-05,
+      "loss": 0.05556968,
+      "step": 12188
+    },
+    {
+      "epoch": 24.378,
+      "grad_norm": 2.524197578430176,
+      "learning_rate": 2e-05,
+      "loss": 0.05749575,
+      "step": 12189
+    },
+    {
+      "epoch": 24.38,
+      "grad_norm": 1.4844391345977783,
+      "learning_rate": 2e-05,
+      "loss": 0.04623716,
+      "step": 12190
+    },
+    {
+      "epoch": 24.382,
+      "grad_norm": 1.9306976795196533,
+      "learning_rate": 2e-05,
+      "loss": 0.04222405,
+      "step": 12191
+    },
+    {
+      "epoch": 24.384,
+      "grad_norm": 1.4491407871246338,
+      "learning_rate": 2e-05,
+      "loss": 0.03950711,
+      "step": 12192
+    },
+    {
+      "epoch": 24.386,
+      "grad_norm": 3.8366308212280273,
+      "learning_rate": 2e-05,
+      "loss": 0.04016727,
+      "step": 12193
+    },
+    {
+      "epoch": 24.388,
+      "grad_norm": 2.6945362091064453,
+      "learning_rate": 2e-05,
+      "loss": 0.05264809,
+      "step": 12194
+    },
+    {
+      "epoch": 24.39,
+      "grad_norm": 1.028458833694458,
+      "learning_rate": 2e-05,
+      "loss": 0.02927229,
+      "step": 12195
+    },
+    {
+      "epoch": 24.392,
+      "grad_norm": 1.1725099086761475,
+      "learning_rate": 2e-05,
+      "loss": 0.04898581,
+      "step": 12196
+    },
+    {
+      "epoch": 24.394,
+      "grad_norm": 1.7470937967300415,
+      "learning_rate": 2e-05,
+      "loss": 0.03954092,
+      "step": 12197
+    },
+    {
+      "epoch": 24.396,
+      "grad_norm": 2.421860456466675,
+      "learning_rate": 2e-05,
+      "loss": 0.04920001,
+      "step": 12198
+    },
+    {
+      "epoch": 24.398,
+      "grad_norm": 1.3461695909500122,
+      "learning_rate": 2e-05,
+      "loss": 0.04510684,
+      "step": 12199
+    },
+    {
+      "epoch": 24.4,
+      "grad_norm": 1.7747408151626587,
+      "learning_rate": 2e-05,
+      "loss": 0.05545141,
+      "step": 12200
+    },
+    {
+      "epoch": 24.402,
+      "grad_norm": 1.768431544303894,
+      "learning_rate": 2e-05,
+      "loss": 0.03842364,
+      "step": 12201
+    },
+    {
+      "epoch": 24.404,
+      "grad_norm": 1.4380724430084229,
+      "learning_rate": 2e-05,
+      "loss": 0.03090051,
+      "step": 12202
+    },
+    {
+      "epoch": 24.406,
+      "grad_norm": 1.3252431154251099,
+      "learning_rate": 2e-05,
+      "loss": 0.04328402,
+      "step": 12203
+    },
+    {
+      "epoch": 24.408,
+      "grad_norm": 1.5969432592391968,
+      "learning_rate": 2e-05,
+      "loss": 0.04301886,
+      "step": 12204
+    },
+    {
+      "epoch": 24.41,
+      "grad_norm": 1.354119062423706,
+      "learning_rate": 2e-05,
+      "loss": 0.03353261,
+      "step": 12205
+    },
+    {
+      "epoch": 24.412,
+      "grad_norm": 1.791136622428894,
+      "learning_rate": 2e-05,
+      "loss": 0.06121269,
+      "step": 12206
+    },
+    {
+      "epoch": 24.414,
+      "grad_norm": 3.968132495880127,
+      "learning_rate": 2e-05,
+      "loss": 0.04851215,
+      "step": 12207
+    },
+    {
+      "epoch": 24.416,
+      "grad_norm": 1.5197120904922485,
+      "learning_rate": 2e-05,
+      "loss": 0.04968302,
+      "step": 12208
+    },
+    {
+      "epoch": 24.418,
+      "grad_norm": 2.272183418273926,
+      "learning_rate": 2e-05,
+      "loss": 0.04752434,
+      "step": 12209
+    },
+    {
+      "epoch": 24.42,
+      "grad_norm": 2.935992479324341,
+      "learning_rate": 2e-05,
+      "loss": 0.05976545,
+      "step": 12210
+    },
+    {
+      "epoch": 24.422,
+      "grad_norm": 1.0861903429031372,
+      "learning_rate": 2e-05,
+      "loss": 0.04118664,
+      "step": 12211
+    },
+    {
+      "epoch": 24.424,
+      "grad_norm": 1.050560474395752,
+      "learning_rate": 2e-05,
+      "loss": 0.03079762,
+      "step": 12212
+    },
+    {
+      "epoch": 24.426,
+      "grad_norm": 2.5487446784973145,
+      "learning_rate": 2e-05,
+      "loss": 0.04272437,
+      "step": 12213
+    },
+    {
+      "epoch": 24.428,
+      "grad_norm": 1.3722039461135864,
+      "learning_rate": 2e-05,
+      "loss": 0.0525198,
+      "step": 12214
+    },
+    {
+      "epoch": 24.43,
+      "grad_norm": 1.176689863204956,
+      "learning_rate": 2e-05,
+      "loss": 0.0387351,
+      "step": 12215
+    },
+    {
+      "epoch": 24.432,
+      "grad_norm": 1.4874681234359741,
+      "learning_rate": 2e-05,
+      "loss": 0.05370562,
+      "step": 12216
+    },
+    {
+      "epoch": 24.434,
+      "grad_norm": 1.266204595565796,
+      "learning_rate": 2e-05,
+      "loss": 0.04619331,
+      "step": 12217
+    },
+    {
+      "epoch": 24.436,
+      "grad_norm": 1.2195080518722534,
+      "learning_rate": 2e-05,
+      "loss": 0.03758291,
+      "step": 12218
+    },
+    {
+      "epoch": 24.438,
+      "grad_norm": 2.4579455852508545,
+      "learning_rate": 2e-05,
+      "loss": 0.04582806,
+      "step": 12219
+    },
+    {
+      "epoch": 24.44,
+      "grad_norm": 1.3026827573776245,
+      "learning_rate": 2e-05,
+      "loss": 0.04520482,
+      "step": 12220
+    },
+    {
+      "epoch": 24.442,
+      "grad_norm": 1.2211519479751587,
+      "learning_rate": 2e-05,
+      "loss": 0.0372187,
+      "step": 12221
+    },
+    {
+      "epoch": 24.444,
+      "grad_norm": 1.1291041374206543,
+      "learning_rate": 2e-05,
+      "loss": 0.03202847,
+      "step": 12222
+    },
+    {
+      "epoch": 24.446,
+      "grad_norm": 1.0536681413650513,
+      "learning_rate": 2e-05,
+      "loss": 0.03556249,
+      "step": 12223
+    },
+    {
+      "epoch": 24.448,
+      "grad_norm": 1.5793570280075073,
+      "learning_rate": 2e-05,
+      "loss": 0.04041308,
+      "step": 12224
+    },
+    {
+      "epoch": 24.45,
+      "grad_norm": 1.9870901107788086,
+      "learning_rate": 2e-05,
+      "loss": 0.04744115,
+      "step": 12225
+    },
+    {
+      "epoch": 24.452,
+      "grad_norm": 1.1192922592163086,
+      "learning_rate": 2e-05,
+      "loss": 0.03771604,
+      "step": 12226
+    },
+    {
+      "epoch": 24.454,
+      "grad_norm": 1.5271708965301514,
+      "learning_rate": 2e-05,
+      "loss": 0.04509822,
+      "step": 12227
+    },
+    {
+      "epoch": 24.456,
+      "grad_norm": 1.09177565574646,
+      "learning_rate": 2e-05,
+      "loss": 0.03703531,
+      "step": 12228
+    },
+    {
+      "epoch": 24.458,
+      "grad_norm": 1.141800880432129,
+      "learning_rate": 2e-05,
+      "loss": 0.03585979,
+      "step": 12229
+    },
+    {
+      "epoch": 24.46,
+      "grad_norm": 0.9765757322311401,
+      "learning_rate": 2e-05,
+      "loss": 0.03288185,
+      "step": 12230
+    },
+    {
+      "epoch": 24.462,
+      "grad_norm": 1.7331185340881348,
+      "learning_rate": 2e-05,
+      "loss": 0.06162242,
+      "step": 12231
+    },
+    {
+      "epoch": 24.464,
+      "grad_norm": 1.0068011283874512,
+      "learning_rate": 2e-05,
+      "loss": 0.03150559,
+      "step": 12232
+    },
+    {
+      "epoch": 24.466,
+      "grad_norm": 2.214324951171875,
+      "learning_rate": 2e-05,
+      "loss": 0.05624455,
+      "step": 12233
+    },
+    {
+      "epoch": 24.468,
+      "grad_norm": 1.3637458086013794,
+      "learning_rate": 2e-05,
+      "loss": 0.04643622,
+      "step": 12234
+    },
+    {
+      "epoch": 24.47,
+      "grad_norm": 1.8071335554122925,
+      "learning_rate": 2e-05,
+      "loss": 0.05237973,
+      "step": 12235
+    },
+    {
+      "epoch": 24.472,
+      "grad_norm": 1.4858529567718506,
+      "learning_rate": 2e-05,
+      "loss": 0.04988906,
+      "step": 12236
+    },
+    {
+      "epoch": 24.474,
+      "grad_norm": 2.615488052368164,
+      "learning_rate": 2e-05,
+      "loss": 0.05420715,
+      "step": 12237
+    },
+    {
+      "epoch": 24.476,
+      "grad_norm": 1.3015847206115723,
+      "learning_rate": 2e-05,
+      "loss": 0.04124864,
+      "step": 12238
+    },
+    {
+      "epoch": 24.478,
+      "grad_norm": 1.670394778251648,
+      "learning_rate": 2e-05,
+      "loss": 0.05579087,
+      "step": 12239
+    },
+    {
+      "epoch": 24.48,
+      "grad_norm": 0.9430084228515625,
+      "learning_rate": 2e-05,
+      "loss": 0.03753918,
+      "step": 12240
+    },
+    {
+      "epoch": 24.482,
+      "grad_norm": 1.3326176404953003,
+      "learning_rate": 2e-05,
+      "loss": 0.05490822,
+      "step": 12241
+    },
+    {
+      "epoch": 24.484,
+      "grad_norm": 1.2883130311965942,
+      "learning_rate": 2e-05,
+      "loss": 0.0533482,
+      "step": 12242
+    },
+    {
+      "epoch": 24.486,
+      "grad_norm": 1.2082029581069946,
+      "learning_rate": 2e-05,
+      "loss": 0.04799801,
+      "step": 12243
+    },
+    {
+      "epoch": 24.488,
+      "grad_norm": 1.6594882011413574,
+      "learning_rate": 2e-05,
+      "loss": 0.04107316,
+      "step": 12244
+    },
+    {
+      "epoch": 24.49,
+      "grad_norm": 1.1790004968643188,
+      "learning_rate": 2e-05,
+      "loss": 0.03800457,
+      "step": 12245
+    },
+    {
+      "epoch": 24.492,
+      "grad_norm": 1.3230494260787964,
+      "learning_rate": 2e-05,
+      "loss": 0.05228403,
+      "step": 12246
+    },
+    {
+      "epoch": 24.494,
+      "grad_norm": 2.5752084255218506,
+      "learning_rate": 2e-05,
+      "loss": 0.04225325,
+      "step": 12247
+    },
+    {
+      "epoch": 24.496,
+      "grad_norm": 1.1051968336105347,
+      "learning_rate": 2e-05,
+      "loss": 0.03597275,
+      "step": 12248
+    },
+    {
+      "epoch": 24.498,
+      "grad_norm": 1.0105959177017212,
+      "learning_rate": 2e-05,
+      "loss": 0.03709402,
+      "step": 12249
+    },
+    {
+      "epoch": 24.5,
+      "grad_norm": 1.417772650718689,
+      "learning_rate": 2e-05,
+      "loss": 0.05351973,
+      "step": 12250
+    },
+    {
+      "epoch": 24.502,
+      "grad_norm": 1.1224205493927002,
+      "learning_rate": 2e-05,
+      "loss": 0.03034319,
+      "step": 12251
+    },
+    {
+      "epoch": 24.504,
+      "grad_norm": 1.2055881023406982,
+      "learning_rate": 2e-05,
+      "loss": 0.04971508,
+      "step": 12252
+    },
+    {
+      "epoch": 24.506,
+      "grad_norm": 2.210259199142456,
+      "learning_rate": 2e-05,
+      "loss": 0.05114106,
+      "step": 12253
+    },
+    {
+      "epoch": 24.508,
+      "grad_norm": 1.4939137697219849,
+      "learning_rate": 2e-05,
+      "loss": 0.05452298,
+      "step": 12254
+    },
+    {
+      "epoch": 24.51,
+      "grad_norm": 0.9842581748962402,
+      "learning_rate": 2e-05,
+      "loss": 0.03696226,
+      "step": 12255
+    },
+    {
+      "epoch": 24.512,
+      "grad_norm": 1.253783941268921,
+      "learning_rate": 2e-05,
+      "loss": 0.03539809,
+      "step": 12256
+    },
+    {
+      "epoch": 24.514,
+      "grad_norm": 3.1072165966033936,
+      "learning_rate": 2e-05,
+      "loss": 0.05696538,
+      "step": 12257
+    },
+    {
+      "epoch": 24.516,
+      "grad_norm": 1.1666399240493774,
+      "learning_rate": 2e-05,
+      "loss": 0.04510186,
+      "step": 12258
+    },
+    {
+      "epoch": 24.518,
+      "grad_norm": 1.8516592979431152,
+      "learning_rate": 2e-05,
+      "loss": 0.06507455,
+      "step": 12259
+    },
+    {
+      "epoch": 24.52,
+      "grad_norm": 2.5476531982421875,
+      "learning_rate": 2e-05,
+      "loss": 0.04386877,
+      "step": 12260
+    },
+    {
+      "epoch": 24.522,
+      "grad_norm": 1.530977487564087,
+      "learning_rate": 2e-05,
+      "loss": 0.05524298,
+      "step": 12261
+    },
+    {
+      "epoch": 24.524,
+      "grad_norm": 0.8967766761779785,
+      "learning_rate": 2e-05,
+      "loss": 0.02332542,
+      "step": 12262
+    },
+    {
+      "epoch": 24.526,
+      "grad_norm": 1.7736743688583374,
+      "learning_rate": 2e-05,
+      "loss": 0.05415568,
+      "step": 12263
+    },
+    {
+      "epoch": 24.528,
+      "grad_norm": 1.1440763473510742,
+      "learning_rate": 2e-05,
+      "loss": 0.0214408,
+      "step": 12264
+    },
+    {
+      "epoch": 24.53,
+      "grad_norm": 1.4621922969818115,
+      "learning_rate": 2e-05,
+      "loss": 0.04975047,
+      "step": 12265
+    },
+    {
+      "epoch": 24.532,
+      "grad_norm": 2.4430623054504395,
+      "learning_rate": 2e-05,
+      "loss": 0.0487452,
+      "step": 12266
+    },
+    {
+      "epoch": 24.534,
+      "grad_norm": 1.67803955078125,
+      "learning_rate": 2e-05,
+      "loss": 0.046294,
+      "step": 12267
+    },
+    {
+      "epoch": 24.536,
+      "grad_norm": 3.9234485626220703,
+      "learning_rate": 2e-05,
+      "loss": 0.05363451,
+      "step": 12268
+    },
+    {
+      "epoch": 24.538,
+      "grad_norm": 1.0378974676132202,
+      "learning_rate": 2e-05,
+      "loss": 0.03485204,
+      "step": 12269
+    },
+    {
+      "epoch": 24.54,
+      "grad_norm": 2.17879581451416,
+      "learning_rate": 2e-05,
+      "loss": 0.05576217,
+      "step": 12270
+    },
+    {
+      "epoch": 24.542,
+      "grad_norm": 1.3864517211914062,
+      "learning_rate": 2e-05,
+      "loss": 0.0414337,
+      "step": 12271
+    },
+    {
+      "epoch": 24.544,
+      "grad_norm": 1.147458553314209,
+      "learning_rate": 2e-05,
+      "loss": 0.03207621,
+      "step": 12272
+    },
+    {
+      "epoch": 24.546,
+      "grad_norm": 1.8686847686767578,
+      "learning_rate": 2e-05,
+      "loss": 0.06861442,
+      "step": 12273
+    },
+    {
+      "epoch": 24.548000000000002,
+      "grad_norm": 1.162184715270996,
+      "learning_rate": 2e-05,
+      "loss": 0.03643592,
+      "step": 12274
+    },
+    {
+      "epoch": 24.55,
+      "grad_norm": 2.051259994506836,
+      "learning_rate": 2e-05,
+      "loss": 0.04812557,
+      "step": 12275
+    },
+    {
+      "epoch": 24.552,
+      "grad_norm": 3.596480131149292,
+      "learning_rate": 2e-05,
+      "loss": 0.04335639,
+      "step": 12276
+    },
+    {
+      "epoch": 24.554,
+      "grad_norm": 1.473477840423584,
+      "learning_rate": 2e-05,
+      "loss": 0.05357858,
+      "step": 12277
+    },
+    {
+      "epoch": 24.556,
+      "grad_norm": 1.5587390661239624,
+      "learning_rate": 2e-05,
+      "loss": 0.04699494,
+      "step": 12278
+    },
+    {
+      "epoch": 24.558,
+      "grad_norm": 1.1098556518554688,
+      "learning_rate": 2e-05,
+      "loss": 0.04609701,
+      "step": 12279
+    },
+    {
+      "epoch": 24.56,
+      "grad_norm": 2.259882688522339,
+      "learning_rate": 2e-05,
+      "loss": 0.05510319,
+      "step": 12280
+    },
+    {
+      "epoch": 24.562,
+      "grad_norm": 1.759789228439331,
+      "learning_rate": 2e-05,
+      "loss": 0.04140832,
+      "step": 12281
+    },
+    {
+      "epoch": 24.564,
+      "grad_norm": 1.605728268623352,
+      "learning_rate": 2e-05,
+      "loss": 0.04575901,
+      "step": 12282
+    },
+    {
+      "epoch": 24.566,
+      "grad_norm": 1.5841501951217651,
+      "learning_rate": 2e-05,
+      "loss": 0.06220622,
+      "step": 12283
+    },
+    {
+      "epoch": 24.568,
+      "grad_norm": 1.8580799102783203,
+      "learning_rate": 2e-05,
+      "loss": 0.0317122,
+      "step": 12284
+    },
+    {
+      "epoch": 24.57,
+      "grad_norm": 1.2249244451522827,
+      "learning_rate": 2e-05,
+      "loss": 0.04554923,
+      "step": 12285
+    },
+    {
+      "epoch": 24.572,
+      "grad_norm": 1.4103236198425293,
+      "learning_rate": 2e-05,
+      "loss": 0.05672751,
+      "step": 12286
+    },
+    {
+      "epoch": 24.574,
+      "grad_norm": 1.2100307941436768,
+      "learning_rate": 2e-05,
+      "loss": 0.04418559,
+      "step": 12287
+    },
+    {
+      "epoch": 24.576,
+      "grad_norm": 1.7110328674316406,
+      "learning_rate": 2e-05,
+      "loss": 0.05983176,
+      "step": 12288
+    },
+    {
+      "epoch": 24.578,
+      "grad_norm": 4.228270053863525,
+      "learning_rate": 2e-05,
+      "loss": 0.05921287,
+      "step": 12289
+    },
+    {
+      "epoch": 24.58,
+      "grad_norm": 1.1426968574523926,
+      "learning_rate": 2e-05,
+      "loss": 0.04294404,
+      "step": 12290
+    },
+    {
+      "epoch": 24.582,
+      "grad_norm": 2.854487419128418,
+      "learning_rate": 2e-05,
+      "loss": 0.04818879,
+      "step": 12291
+    },
+    {
+      "epoch": 24.584,
+      "grad_norm": 1.464268445968628,
+      "learning_rate": 2e-05,
+      "loss": 0.05259855,
+      "step": 12292
+    },
+    {
+      "epoch": 24.586,
+      "grad_norm": 1.6875348091125488,
+      "learning_rate": 2e-05,
+      "loss": 0.04386838,
+      "step": 12293
+    },
+    {
+      "epoch": 24.588,
+      "grad_norm": 1.284960150718689,
+      "learning_rate": 2e-05,
+      "loss": 0.03831192,
+      "step": 12294
+    },
+    {
+      "epoch": 24.59,
+      "grad_norm": 1.7952815294265747,
+      "learning_rate": 2e-05,
+      "loss": 0.05706906,
+      "step": 12295
+    },
+    {
+      "epoch": 24.592,
+      "grad_norm": 1.6755213737487793,
+      "learning_rate": 2e-05,
+      "loss": 0.05718254,
+      "step": 12296
+    },
+    {
+      "epoch": 24.594,
+      "grad_norm": 1.6490305662155151,
+      "learning_rate": 2e-05,
+      "loss": 0.04554712,
+      "step": 12297
+    },
+    {
+      "epoch": 24.596,
+      "grad_norm": 1.4170808792114258,
+      "learning_rate": 2e-05,
+      "loss": 0.04948672,
+      "step": 12298
+    },
+    {
+      "epoch": 24.598,
+      "grad_norm": 0.9543237686157227,
+      "learning_rate": 2e-05,
+      "loss": 0.03450447,
+      "step": 12299
+    },
+    {
+      "epoch": 24.6,
+      "grad_norm": 1.3354976177215576,
+      "learning_rate": 2e-05,
+      "loss": 0.04121848,
+      "step": 12300
+    },
+    {
+      "epoch": 24.602,
+      "grad_norm": 1.322101354598999,
+      "learning_rate": 2e-05,
+      "loss": 0.03823018,
+      "step": 12301
+    },
+    {
+      "epoch": 24.604,
+      "grad_norm": 3.014385938644409,
+      "learning_rate": 2e-05,
+      "loss": 0.04420658,
+      "step": 12302
+    },
+    {
+      "epoch": 24.606,
+      "grad_norm": 1.7275763750076294,
+      "learning_rate": 2e-05,
+      "loss": 0.05624251,
+      "step": 12303
+    },
+    {
+      "epoch": 24.608,
+      "grad_norm": 1.091782569885254,
+      "learning_rate": 2e-05,
+      "loss": 0.0310542,
+      "step": 12304
+    },
+    {
+      "epoch": 24.61,
+      "grad_norm": 1.6847692728042603,
+      "learning_rate": 2e-05,
+      "loss": 0.04774918,
+      "step": 12305
+    },
+    {
+      "epoch": 24.612,
+      "grad_norm": 1.1882266998291016,
+      "learning_rate": 2e-05,
+      "loss": 0.04614477,
+      "step": 12306
+    },
+    {
+      "epoch": 24.614,
+      "grad_norm": 2.3512465953826904,
+      "learning_rate": 2e-05,
+      "loss": 0.07600546,
+      "step": 12307
+    },
+    {
+      "epoch": 24.616,
+      "grad_norm": 1.3159834146499634,
+      "learning_rate": 2e-05,
+      "loss": 0.05892436,
+      "step": 12308
+    },
+    {
+      "epoch": 24.618,
+      "grad_norm": 1.2765941619873047,
+      "learning_rate": 2e-05,
+      "loss": 0.03476145,
+      "step": 12309
+    },
+    {
+      "epoch": 24.62,
+      "grad_norm": 1.3784600496292114,
+      "learning_rate": 2e-05,
+      "loss": 0.03707298,
+      "step": 12310
+    },
+    {
+      "epoch": 24.622,
+      "grad_norm": 1.248801350593567,
+      "learning_rate": 2e-05,
+      "loss": 0.04117713,
+      "step": 12311
+    },
+    {
+      "epoch": 24.624,
+      "grad_norm": 1.257441759109497,
+      "learning_rate": 2e-05,
+      "loss": 0.03224612,
+      "step": 12312
+    },
+    {
+      "epoch": 24.626,
+      "grad_norm": 1.3691023588180542,
+      "learning_rate": 2e-05,
+      "loss": 0.04922335,
+      "step": 12313
+    },
+    {
+      "epoch": 24.628,
+      "grad_norm": 1.3157551288604736,
+      "learning_rate": 2e-05,
+      "loss": 0.03670745,
+      "step": 12314
+    },
+    {
+      "epoch": 24.63,
+      "grad_norm": 1.2883113622665405,
+      "learning_rate": 2e-05,
+      "loss": 0.03211619,
+      "step": 12315
+    },
+    {
+      "epoch": 24.632,
+      "grad_norm": 1.0646746158599854,
+      "learning_rate": 2e-05,
+      "loss": 0.03154656,
+      "step": 12316
+    },
+    {
+      "epoch": 24.634,
+      "grad_norm": 2.252239227294922,
+      "learning_rate": 2e-05,
+      "loss": 0.0534423,
+      "step": 12317
+    },
+    {
+      "epoch": 24.636,
+      "grad_norm": 1.5512782335281372,
+      "learning_rate": 2e-05,
+      "loss": 0.05046772,
+      "step": 12318
+    },
+    {
+      "epoch": 24.638,
+      "grad_norm": 1.1680291891098022,
+      "learning_rate": 2e-05,
+      "loss": 0.04530792,
+      "step": 12319
+    },
+    {
+      "epoch": 24.64,
+      "grad_norm": 1.2998766899108887,
+      "learning_rate": 2e-05,
+      "loss": 0.0484796,
+      "step": 12320
+    },
+    {
+      "epoch": 24.642,
+      "grad_norm": 2.491387128829956,
+      "learning_rate": 2e-05,
+      "loss": 0.04955789,
+      "step": 12321
+    },
+    {
+      "epoch": 24.644,
+      "grad_norm": 2.7813923358917236,
+      "learning_rate": 2e-05,
+      "loss": 0.05465161,
+      "step": 12322
+    },
+    {
+      "epoch": 24.646,
+      "grad_norm": 3.847205400466919,
+      "learning_rate": 2e-05,
+      "loss": 0.07493284,
+      "step": 12323
+    },
+    {
+      "epoch": 24.648,
+      "grad_norm": 1.364022135734558,
+      "learning_rate": 2e-05,
+      "loss": 0.03194257,
+      "step": 12324
+    },
+    {
+      "epoch": 24.65,
+      "grad_norm": 1.3578834533691406,
+      "learning_rate": 2e-05,
+      "loss": 0.05537183,
+      "step": 12325
+    },
+    {
+      "epoch": 24.652,
+      "grad_norm": 1.1270604133605957,
+      "learning_rate": 2e-05,
+      "loss": 0.04368369,
+      "step": 12326
+    },
+    {
+      "epoch": 24.654,
+      "grad_norm": 2.827782154083252,
+      "learning_rate": 2e-05,
+      "loss": 0.05406433,
+      "step": 12327
+    },
+    {
+      "epoch": 24.656,
+      "grad_norm": 1.6303904056549072,
+      "learning_rate": 2e-05,
+      "loss": 0.05844283,
+      "step": 12328
+    },
+    {
+      "epoch": 24.658,
+      "grad_norm": 1.267575740814209,
+      "learning_rate": 2e-05,
+      "loss": 0.04553982,
+      "step": 12329
+    },
+    {
+      "epoch": 24.66,
+      "grad_norm": 2.1354129314422607,
+      "learning_rate": 2e-05,
+      "loss": 0.05353948,
+      "step": 12330
+    },
+    {
+      "epoch": 24.662,
+      "grad_norm": 3.700606346130371,
+      "learning_rate": 2e-05,
+      "loss": 0.03866877,
+      "step": 12331
+    },
+    {
+      "epoch": 24.664,
+      "grad_norm": 3.4093546867370605,
+      "learning_rate": 2e-05,
+      "loss": 0.03720096,
+      "step": 12332
+    },
+    {
+      "epoch": 24.666,
+      "grad_norm": 1.756603479385376,
+      "learning_rate": 2e-05,
+      "loss": 0.04626716,
+      "step": 12333
+    },
+    {
+      "epoch": 24.668,
+      "grad_norm": 2.8935391902923584,
+      "learning_rate": 2e-05,
+      "loss": 0.04296909,
+      "step": 12334
+    },
+    {
+      "epoch": 24.67,
+      "grad_norm": 2.631392002105713,
+      "learning_rate": 2e-05,
+      "loss": 0.05087268,
+      "step": 12335
+    },
+    {
+      "epoch": 24.672,
+      "grad_norm": 1.9597891569137573,
+      "learning_rate": 2e-05,
+      "loss": 0.04416256,
+      "step": 12336
+    },
+    {
+      "epoch": 24.674,
+      "grad_norm": 1.3087323904037476,
+      "learning_rate": 2e-05,
+      "loss": 0.05481458,
+      "step": 12337
+    },
+    {
+      "epoch": 24.676,
+      "grad_norm": 172.79251098632812,
+      "learning_rate": 2e-05,
+      "loss": 0.04823511,
+      "step": 12338
+    },
+    {
+      "epoch": 24.678,
+      "grad_norm": 1.6188311576843262,
+      "learning_rate": 2e-05,
+      "loss": 0.05058477,
+      "step": 12339
+    },
+    {
+      "epoch": 24.68,
+      "grad_norm": 1.4070014953613281,
+      "learning_rate": 2e-05,
+      "loss": 0.05267171,
+      "step": 12340
+    },
+    {
+      "epoch": 24.682,
+      "grad_norm": 1.056060552597046,
+      "learning_rate": 2e-05,
+      "loss": 0.04122014,
+      "step": 12341
+    },
+    {
+      "epoch": 24.684,
+      "grad_norm": 1.0840667486190796,
+      "learning_rate": 2e-05,
+      "loss": 0.03836139,
+      "step": 12342
+    },
+    {
+      "epoch": 24.686,
+      "grad_norm": 1.6224464178085327,
+      "learning_rate": 2e-05,
+      "loss": 0.03195419,
+      "step": 12343
+    },
+    {
+      "epoch": 24.688,
+      "grad_norm": 1.4936470985412598,
+      "learning_rate": 2e-05,
+      "loss": 0.07893459,
+      "step": 12344
+    },
+    {
+      "epoch": 24.69,
+      "grad_norm": 1.8073782920837402,
+      "learning_rate": 2e-05,
+      "loss": 0.05727108,
+      "step": 12345
+    },
+    {
+      "epoch": 24.692,
+      "grad_norm": 1.218421220779419,
+      "learning_rate": 2e-05,
+      "loss": 0.046979,
+      "step": 12346
+    },
+    {
+      "epoch": 24.694,
+      "grad_norm": 1.725235939025879,
+      "learning_rate": 2e-05,
+      "loss": 0.03699161,
+      "step": 12347
+    },
+    {
+      "epoch": 24.696,
+      "grad_norm": 2.4789557456970215,
+      "learning_rate": 2e-05,
+      "loss": 0.04929878,
+      "step": 12348
+    },
+    {
+      "epoch": 24.698,
+      "grad_norm": 3.441601276397705,
+      "learning_rate": 2e-05,
+      "loss": 0.0512137,
+      "step": 12349
+    },
+    {
+      "epoch": 24.7,
+      "grad_norm": 1.1820622682571411,
+      "learning_rate": 2e-05,
+      "loss": 0.0380182,
+      "step": 12350
+    },
+    {
+      "epoch": 24.701999999999998,
+      "grad_norm": 1.0114235877990723,
+      "learning_rate": 2e-05,
+      "loss": 0.03822229,
+      "step": 12351
+    },
+    {
+      "epoch": 24.704,
+      "grad_norm": 1.1000703573226929,
+      "learning_rate": 2e-05,
+      "loss": 0.03765395,
+      "step": 12352
+    },
+    {
+      "epoch": 24.706,
+      "grad_norm": 1.2947152853012085,
+      "learning_rate": 2e-05,
+      "loss": 0.03386806,
+      "step": 12353
+    },
+    {
+      "epoch": 24.708,
+      "grad_norm": 1.8047212362289429,
+      "learning_rate": 2e-05,
+      "loss": 0.05379967,
+      "step": 12354
+    },
+    {
+      "epoch": 24.71,
+      "grad_norm": 2.945786476135254,
+      "learning_rate": 2e-05,
+      "loss": 0.05634,
+      "step": 12355
+    },
+    {
+      "epoch": 24.712,
+      "grad_norm": 2.3145387172698975,
+      "learning_rate": 2e-05,
+      "loss": 0.06896135,
+      "step": 12356
+    },
+    {
+      "epoch": 24.714,
+      "grad_norm": 1.7058087587356567,
+      "learning_rate": 2e-05,
+      "loss": 0.04002954,
+      "step": 12357
+    },
+    {
+      "epoch": 24.716,
+      "grad_norm": 1.0332450866699219,
+      "learning_rate": 2e-05,
+      "loss": 0.04152671,
+      "step": 12358
+    },
+    {
+      "epoch": 24.718,
+      "grad_norm": 1.0253223180770874,
+      "learning_rate": 2e-05,
+      "loss": 0.05022538,
+      "step": 12359
+    },
+    {
+      "epoch": 24.72,
+      "grad_norm": 1.4460958242416382,
+      "learning_rate": 2e-05,
+      "loss": 0.04726547,
+      "step": 12360
+    },
+    {
+      "epoch": 24.722,
+      "grad_norm": 2.94338321685791,
+      "learning_rate": 2e-05,
+      "loss": 0.04982202,
+      "step": 12361
+    },
+    {
+      "epoch": 24.724,
+      "grad_norm": 1.2132338285446167,
+      "learning_rate": 2e-05,
+      "loss": 0.0424381,
+      "step": 12362
+    },
+    {
+      "epoch": 24.726,
+      "grad_norm": 1.2509801387786865,
+      "learning_rate": 2e-05,
+      "loss": 0.03625492,
+      "step": 12363
+    },
+    {
+      "epoch": 24.728,
+      "grad_norm": 1.001670241355896,
+      "learning_rate": 2e-05,
+      "loss": 0.03844631,
+      "step": 12364
+    },
+    {
+      "epoch": 24.73,
+      "grad_norm": 1.2094303369522095,
+      "learning_rate": 2e-05,
+      "loss": 0.05520956,
+      "step": 12365
+    },
+    {
+      "epoch": 24.732,
+      "grad_norm": 1.0332337617874146,
+      "learning_rate": 2e-05,
+      "loss": 0.03330713,
+      "step": 12366
+    },
+    {
+      "epoch": 24.734,
+      "grad_norm": 1.8160288333892822,
+      "learning_rate": 2e-05,
+      "loss": 0.06366383,
+      "step": 12367
+    },
+    {
+      "epoch": 24.736,
+      "grad_norm": 1.0962200164794922,
+      "learning_rate": 2e-05,
+      "loss": 0.03358031,
+      "step": 12368
+    },
+    {
+      "epoch": 24.738,
+      "grad_norm": 1.2256057262420654,
+      "learning_rate": 2e-05,
+      "loss": 0.04108488,
+      "step": 12369
+    },
+    {
+      "epoch": 24.74,
+      "grad_norm": 2.2090930938720703,
+      "learning_rate": 2e-05,
+      "loss": 0.06522129,
+      "step": 12370
+    },
+    {
+      "epoch": 24.742,
+      "grad_norm": 1.1580830812454224,
+      "learning_rate": 2e-05,
+      "loss": 0.03839488,
+      "step": 12371
+    },
+    {
+      "epoch": 24.744,
+      "grad_norm": 1.1396980285644531,
+      "learning_rate": 2e-05,
+      "loss": 0.04421246,
+      "step": 12372
+    },
+    {
+      "epoch": 24.746,
+      "grad_norm": 1.0483407974243164,
+      "learning_rate": 2e-05,
+      "loss": 0.03586071,
+      "step": 12373
+    },
+    {
+      "epoch": 24.748,
+      "grad_norm": 1.1575627326965332,
+      "learning_rate": 2e-05,
+      "loss": 0.02652693,
+      "step": 12374
+    },
+    {
+      "epoch": 24.75,
+      "grad_norm": 1.2578696012496948,
+      "learning_rate": 2e-05,
+      "loss": 0.04251358,
+      "step": 12375
+    },
+    {
+      "epoch": 24.752,
+      "grad_norm": 1.2931718826293945,
+      "learning_rate": 2e-05,
+      "loss": 0.04332025,
+      "step": 12376
+    },
+    {
+      "epoch": 24.754,
+      "grad_norm": 1.2202832698822021,
+      "learning_rate": 2e-05,
+      "loss": 0.0368543,
+      "step": 12377
+    },
+    {
+      "epoch": 24.756,
+      "grad_norm": 2.3942205905914307,
+      "learning_rate": 2e-05,
+      "loss": 0.05159555,
+      "step": 12378
+    },
+    {
+      "epoch": 24.758,
+      "grad_norm": 1.4237806797027588,
+      "learning_rate": 2e-05,
+      "loss": 0.04190379,
+      "step": 12379
+    },
+    {
+      "epoch": 24.76,
+      "grad_norm": 2.9148285388946533,
+      "learning_rate": 2e-05,
+      "loss": 0.0515301,
+      "step": 12380
+    },
+    {
+      "epoch": 24.762,
+      "grad_norm": 1.355830192565918,
+      "learning_rate": 2e-05,
+      "loss": 0.04525499,
+      "step": 12381
+    },
+    {
+      "epoch": 24.764,
+      "grad_norm": 1.7303012609481812,
+      "learning_rate": 2e-05,
+      "loss": 0.05111047,
+      "step": 12382
+    },
+    {
+      "epoch": 24.766,
+      "grad_norm": 1.722062110900879,
+      "learning_rate": 2e-05,
+      "loss": 0.05258113,
+      "step": 12383
+    },
+    {
+      "epoch": 24.768,
+      "grad_norm": 0.8899957537651062,
+      "learning_rate": 2e-05,
+      "loss": 0.02432646,
+      "step": 12384
+    },
+    {
+      "epoch": 24.77,
+      "grad_norm": 1.543811559677124,
+      "learning_rate": 2e-05,
+      "loss": 0.04550616,
+      "step": 12385
+    },
+    {
+      "epoch": 24.772,
+      "grad_norm": 1.6278526782989502,
+      "learning_rate": 2e-05,
+      "loss": 0.05577083,
+      "step": 12386
+    },
+    {
+      "epoch": 24.774,
+      "grad_norm": 1.1185393333435059,
+      "learning_rate": 2e-05,
+      "loss": 0.0319613,
+      "step": 12387
+    },
+    {
+      "epoch": 24.776,
+      "grad_norm": 1.4521170854568481,
+      "learning_rate": 2e-05,
+      "loss": 0.05207948,
+      "step": 12388
+    },
+    {
+      "epoch": 24.778,
+      "grad_norm": 1.3518644571304321,
+      "learning_rate": 2e-05,
+      "loss": 0.04421254,
+      "step": 12389
+    },
+    {
+      "epoch": 24.78,
+      "grad_norm": 1.6394786834716797,
+      "learning_rate": 2e-05,
+      "loss": 0.05696079,
+      "step": 12390
+    },
+    {
+      "epoch": 24.782,
+      "grad_norm": 1.0339347124099731,
+      "learning_rate": 2e-05,
+      "loss": 0.03561187,
+      "step": 12391
+    },
+    {
+      "epoch": 24.784,
+      "grad_norm": 2.129795789718628,
+      "learning_rate": 2e-05,
+      "loss": 0.04448505,
+      "step": 12392
+    },
+    {
+      "epoch": 24.786,
+      "grad_norm": 1.024999976158142,
+      "learning_rate": 2e-05,
+      "loss": 0.05013253,
+      "step": 12393
+    },
+    {
+      "epoch": 24.788,
+      "grad_norm": 1.413780927658081,
+      "learning_rate": 2e-05,
+      "loss": 0.03977506,
+      "step": 12394
+    },
+    {
+      "epoch": 24.79,
+      "grad_norm": 1.4434248208999634,
+      "learning_rate": 2e-05,
+      "loss": 0.05382534,
+      "step": 12395
+    },
+    {
+      "epoch": 24.792,
+      "grad_norm": 1.4509685039520264,
+      "learning_rate": 2e-05,
+      "loss": 0.04992918,
+      "step": 12396
+    },
+    {
+      "epoch": 24.794,
+      "grad_norm": 1.47963547706604,
+      "learning_rate": 2e-05,
+      "loss": 0.0391894,
+      "step": 12397
+    },
+    {
+      "epoch": 24.796,
+      "grad_norm": 1.1163206100463867,
+      "learning_rate": 2e-05,
+      "loss": 0.03056799,
+      "step": 12398
+    },
+    {
+      "epoch": 24.798000000000002,
+      "grad_norm": 1.147055745124817,
+      "learning_rate": 2e-05,
+      "loss": 0.03964867,
+      "step": 12399
+    },
+    {
+      "epoch": 24.8,
+      "grad_norm": 0.9283363819122314,
+      "learning_rate": 2e-05,
+      "loss": 0.02905277,
+      "step": 12400
+    },
+    {
+      "epoch": 24.802,
+      "grad_norm": 0.967431902885437,
+      "learning_rate": 2e-05,
+      "loss": 0.03618809,
+      "step": 12401
+    },
+    {
+      "epoch": 24.804,
+      "grad_norm": 1.2811535596847534,
+      "learning_rate": 2e-05,
+      "loss": 0.04123492,
+      "step": 12402
+    },
+    {
+      "epoch": 24.806,
+      "grad_norm": 0.7825071215629578,
+      "learning_rate": 2e-05,
+      "loss": 0.02419277,
+      "step": 12403
+    },
+    {
+      "epoch": 24.808,
+      "grad_norm": 3.262420654296875,
+      "learning_rate": 2e-05,
+      "loss": 0.04532979,
+      "step": 12404
+    },
+    {
+      "epoch": 24.81,
+      "grad_norm": 3.3228957653045654,
+      "learning_rate": 2e-05,
+      "loss": 0.04625534,
+      "step": 12405
+    },
+    {
+      "epoch": 24.812,
+      "grad_norm": 2.1002676486968994,
+      "learning_rate": 2e-05,
+      "loss": 0.04704674,
+      "step": 12406
+    },
+    {
+      "epoch": 24.814,
+      "grad_norm": 2.0813817977905273,
+      "learning_rate": 2e-05,
+      "loss": 0.06468686,
+      "step": 12407
+    },
+    {
+      "epoch": 24.816,
+      "grad_norm": 1.4478918313980103,
+      "learning_rate": 2e-05,
+      "loss": 0.04282294,
+      "step": 12408
+    },
+    {
+      "epoch": 24.818,
+      "grad_norm": 1.2476028203964233,
+      "learning_rate": 2e-05,
+      "loss": 0.04784694,
+      "step": 12409
+    },
+    {
+      "epoch": 24.82,
+      "grad_norm": 1.4642516374588013,
+      "learning_rate": 2e-05,
+      "loss": 0.05254901,
+      "step": 12410
+    },
+    {
+      "epoch": 24.822,
+      "grad_norm": 0.8719230890274048,
+      "learning_rate": 2e-05,
+      "loss": 0.0320285,
+      "step": 12411
+    },
+    {
+      "epoch": 24.824,
+      "grad_norm": 3.0026919841766357,
+      "learning_rate": 2e-05,
+      "loss": 0.05441289,
+      "step": 12412
+    },
+    {
+      "epoch": 24.826,
+      "grad_norm": 3.024451732635498,
+      "learning_rate": 2e-05,
+      "loss": 0.04576012,
+      "step": 12413
+    },
+    {
+      "epoch": 24.828,
+      "grad_norm": 1.2607303857803345,
+      "learning_rate": 2e-05,
+      "loss": 0.05559196,
+      "step": 12414
+    },
+    {
+      "epoch": 24.83,
+      "grad_norm": 1.9243049621582031,
+      "learning_rate": 2e-05,
+      "loss": 0.04724313,
+      "step": 12415
+    },
+    {
+      "epoch": 24.832,
+      "grad_norm": 1.1702837944030762,
+      "learning_rate": 2e-05,
+      "loss": 0.04239753,
+      "step": 12416
+    },
+    {
+      "epoch": 24.834,
+      "grad_norm": 1.2738158702850342,
+      "learning_rate": 2e-05,
+      "loss": 0.04667337,
+      "step": 12417
+    },
+    {
+      "epoch": 24.836,
+      "grad_norm": 1.4549773931503296,
+      "learning_rate": 2e-05,
+      "loss": 0.03520815,
+      "step": 12418
+    },
+    {
+      "epoch": 24.838,
+      "grad_norm": 1.89911687374115,
+      "learning_rate": 2e-05,
+      "loss": 0.05533872,
+      "step": 12419
+    },
+    {
+      "epoch": 24.84,
+      "grad_norm": 1.3373122215270996,
+      "learning_rate": 2e-05,
+      "loss": 0.04861745,
+      "step": 12420
+    },
+    {
+      "epoch": 24.842,
+      "grad_norm": 1.1704626083374023,
+      "learning_rate": 2e-05,
+      "loss": 0.04655242,
+      "step": 12421
+    },
+    {
+      "epoch": 24.844,
+      "grad_norm": 1.7334611415863037,
+      "learning_rate": 2e-05,
+      "loss": 0.04279828,
+      "step": 12422
+    },
+    {
+      "epoch": 24.846,
+      "grad_norm": 1.5228956937789917,
+      "learning_rate": 2e-05,
+      "loss": 0.06574546,
+      "step": 12423
+    },
+    {
+      "epoch": 24.848,
+      "grad_norm": 1.5824440717697144,
+      "learning_rate": 2e-05,
+      "loss": 0.03519925,
+      "step": 12424
+    },
+    {
+      "epoch": 24.85,
+      "grad_norm": 1.4289437532424927,
+      "learning_rate": 2e-05,
+      "loss": 0.03780733,
+      "step": 12425
+    },
+    {
+      "epoch": 24.852,
+      "grad_norm": 1.498123288154602,
+      "learning_rate": 2e-05,
+      "loss": 0.03797191,
+      "step": 12426
+    },
+    {
+      "epoch": 24.854,
+      "grad_norm": 1.006430745124817,
+      "learning_rate": 2e-05,
+      "loss": 0.03351758,
+      "step": 12427
+    },
+    {
+      "epoch": 24.856,
+      "grad_norm": 1.0861419439315796,
+      "learning_rate": 2e-05,
+      "loss": 0.0464669,
+      "step": 12428
+    },
+    {
+      "epoch": 24.858,
+      "grad_norm": 1.2548224925994873,
+      "learning_rate": 2e-05,
+      "loss": 0.02788034,
+      "step": 12429
+    },
+    {
+      "epoch": 24.86,
+      "grad_norm": 1.3924577236175537,
+      "learning_rate": 2e-05,
+      "loss": 0.02896036,
+      "step": 12430
+    },
+    {
+      "epoch": 24.862,
+      "grad_norm": 2.9513254165649414,
+      "learning_rate": 2e-05,
+      "loss": 0.05202112,
+      "step": 12431
+    },
+    {
+      "epoch": 24.864,
+      "grad_norm": 1.3894107341766357,
+      "learning_rate": 2e-05,
+      "loss": 0.05267946,
+      "step": 12432
+    },
+    {
+      "epoch": 24.866,
+      "grad_norm": 1.1392052173614502,
+      "learning_rate": 2e-05,
+      "loss": 0.04014817,
+      "step": 12433
+    },
+    {
+      "epoch": 24.868,
+      "grad_norm": 1.3228307962417603,
+      "learning_rate": 2e-05,
+      "loss": 0.03538731,
+      "step": 12434
+    },
+    {
+      "epoch": 24.87,
+      "grad_norm": 1.11819589138031,
+      "learning_rate": 2e-05,
+      "loss": 0.03386147,
+      "step": 12435
+    },
+    {
+      "epoch": 24.872,
+      "grad_norm": 1.2249038219451904,
+      "learning_rate": 2e-05,
+      "loss": 0.0489039,
+      "step": 12436
+    },
+    {
+      "epoch": 24.874,
+      "grad_norm": 1.4754010438919067,
+      "learning_rate": 2e-05,
+      "loss": 0.04698762,
+      "step": 12437
+    },
+    {
+      "epoch": 24.876,
+      "grad_norm": 3.302642583847046,
+      "learning_rate": 2e-05,
+      "loss": 0.05646276,
+      "step": 12438
+    },
+    {
+      "epoch": 24.878,
+      "grad_norm": 1.419040560722351,
+      "learning_rate": 2e-05,
+      "loss": 0.04880114,
+      "step": 12439
+    },
+    {
+      "epoch": 24.88,
+      "grad_norm": 1.1326826810836792,
+      "learning_rate": 2e-05,
+      "loss": 0.04119021,
+      "step": 12440
+    },
+    {
+      "epoch": 24.882,
+      "grad_norm": 1.429632306098938,
+      "learning_rate": 2e-05,
+      "loss": 0.0510744,
+      "step": 12441
+    },
+    {
+      "epoch": 24.884,
+      "grad_norm": 2.021416425704956,
+      "learning_rate": 2e-05,
+      "loss": 0.05398323,
+      "step": 12442
+    },
+    {
+      "epoch": 24.886,
+      "grad_norm": 1.2925662994384766,
+      "learning_rate": 2e-05,
+      "loss": 0.0487259,
+      "step": 12443
+    },
+    {
+      "epoch": 24.888,
+      "grad_norm": 1.3188891410827637,
+      "learning_rate": 2e-05,
+      "loss": 0.04359906,
+      "step": 12444
+    },
+    {
+      "epoch": 24.89,
+      "grad_norm": 1.8213289976119995,
+      "learning_rate": 2e-05,
+      "loss": 0.05419963,
+      "step": 12445
+    },
+    {
+      "epoch": 24.892,
+      "grad_norm": 1.8820993900299072,
+      "learning_rate": 2e-05,
+      "loss": 0.04406717,
+      "step": 12446
+    },
+    {
+      "epoch": 24.894,
+      "grad_norm": 1.1095833778381348,
+      "learning_rate": 2e-05,
+      "loss": 0.04703937,
+      "step": 12447
+    },
+    {
+      "epoch": 24.896,
+      "grad_norm": 1.145403265953064,
+      "learning_rate": 2e-05,
+      "loss": 0.04098237,
+      "step": 12448
+    },
+    {
+      "epoch": 24.898,
+      "grad_norm": 4.939516067504883,
+      "learning_rate": 2e-05,
+      "loss": 0.04888795,
+      "step": 12449
+    },
+    {
+      "epoch": 24.9,
+      "grad_norm": 1.2902284860610962,
+      "learning_rate": 2e-05,
+      "loss": 0.04742273,
+      "step": 12450
+    },
+    {
+      "epoch": 24.902,
+      "grad_norm": 1.15613853931427,
+      "learning_rate": 2e-05,
+      "loss": 0.02914495,
+      "step": 12451
+    },
+    {
+      "epoch": 24.904,
+      "grad_norm": 2.327373504638672,
+      "learning_rate": 2e-05,
+      "loss": 0.05553031,
+      "step": 12452
+    },
+    {
+      "epoch": 24.906,
+      "grad_norm": 1.4978739023208618,
+      "learning_rate": 2e-05,
+      "loss": 0.03483084,
+      "step": 12453
+    },
+    {
+      "epoch": 24.908,
+      "grad_norm": 1.7437061071395874,
+      "learning_rate": 2e-05,
+      "loss": 0.05667341,
+      "step": 12454
+    },
+    {
+      "epoch": 24.91,
+      "grad_norm": 2.045104742050171,
+      "learning_rate": 2e-05,
+      "loss": 0.0504937,
+      "step": 12455
+    },
+    {
+      "epoch": 24.912,
+      "grad_norm": 2.127146005630493,
+      "learning_rate": 2e-05,
+      "loss": 0.04464479,
+      "step": 12456
+    },
+    {
+      "epoch": 24.914,
+      "grad_norm": 1.9251153469085693,
+      "learning_rate": 2e-05,
+      "loss": 0.04323905,
+      "step": 12457
+    },
+    {
+      "epoch": 24.916,
+      "grad_norm": 1.2294851541519165,
+      "learning_rate": 2e-05,
+      "loss": 0.03710897,
+      "step": 12458
+    },
+    {
+      "epoch": 24.918,
+      "grad_norm": 1.0860822200775146,
+      "learning_rate": 2e-05,
+      "loss": 0.04871416,
+      "step": 12459
+    },
+    {
+      "epoch": 24.92,
+      "grad_norm": 1.0837758779525757,
+      "learning_rate": 2e-05,
+      "loss": 0.03727125,
+      "step": 12460
+    },
+    {
+      "epoch": 24.922,
+      "grad_norm": 1.4478741884231567,
+      "learning_rate": 2e-05,
+      "loss": 0.04571364,
+      "step": 12461
+    },
+    {
+      "epoch": 24.924,
+      "grad_norm": 1.5994067192077637,
+      "learning_rate": 2e-05,
+      "loss": 0.04567996,
+      "step": 12462
+    },
+    {
+      "epoch": 24.926,
+      "grad_norm": 2.3954379558563232,
+      "learning_rate": 2e-05,
+      "loss": 0.05547336,
+      "step": 12463
+    },
+    {
+      "epoch": 24.928,
+      "grad_norm": 2.130188465118408,
+      "learning_rate": 2e-05,
+      "loss": 0.0510396,
+      "step": 12464
+    },
+    {
+      "epoch": 24.93,
+      "grad_norm": 19.31425666809082,
+      "learning_rate": 2e-05,
+      "loss": 0.09202991,
+      "step": 12465
+    },
+    {
+      "epoch": 24.932,
+      "grad_norm": 1.2676620483398438,
+      "learning_rate": 2e-05,
+      "loss": 0.03964383,
+      "step": 12466
+    },
+    {
+      "epoch": 24.934,
+      "grad_norm": 1.5841865539550781,
+      "learning_rate": 2e-05,
+      "loss": 0.03884972,
+      "step": 12467
+    },
+    {
+      "epoch": 24.936,
+      "grad_norm": 1.8690932989120483,
+      "learning_rate": 2e-05,
+      "loss": 0.05343563,
+      "step": 12468
+    },
+    {
+      "epoch": 24.938,
+      "grad_norm": 1.262275218963623,
+      "learning_rate": 2e-05,
+      "loss": 0.03402144,
+      "step": 12469
+    },
+    {
+      "epoch": 24.94,
+      "grad_norm": 1.1100060939788818,
+      "learning_rate": 2e-05,
+      "loss": 0.04514051,
+      "step": 12470
+    },
+    {
+      "epoch": 24.942,
+      "grad_norm": 1.4891170263290405,
+      "learning_rate": 2e-05,
+      "loss": 0.04965903,
+      "step": 12471
+    },
+    {
+      "epoch": 24.944,
+      "grad_norm": 1.1756887435913086,
+      "learning_rate": 2e-05,
+      "loss": 0.03555219,
+      "step": 12472
+    },
+    {
+      "epoch": 24.946,
+      "grad_norm": 1.38705575466156,
+      "learning_rate": 2e-05,
+      "loss": 0.05496265,
+      "step": 12473
+    },
+    {
+      "epoch": 24.948,
+      "grad_norm": 1.8380342721939087,
+      "learning_rate": 2e-05,
+      "loss": 0.05392865,
+      "step": 12474
+    },
+    {
+      "epoch": 24.95,
+      "grad_norm": 0.9538768529891968,
+      "learning_rate": 2e-05,
+      "loss": 0.03763831,
+      "step": 12475
+    },
+    {
+      "epoch": 24.951999999999998,
+      "grad_norm": 1.4392766952514648,
+      "learning_rate": 2e-05,
+      "loss": 0.03219794,
+      "step": 12476
+    },
+    {
+      "epoch": 24.954,
+      "grad_norm": 1.1344821453094482,
+      "learning_rate": 2e-05,
+      "loss": 0.04084168,
+      "step": 12477
+    },
+    {
+      "epoch": 24.956,
+      "grad_norm": 1.4016164541244507,
+      "learning_rate": 2e-05,
+      "loss": 0.04202775,
+      "step": 12478
+    },
+    {
+      "epoch": 24.958,
+      "grad_norm": 1.3038735389709473,
+      "learning_rate": 2e-05,
+      "loss": 0.05586255,
+      "step": 12479
+    },
+    {
+      "epoch": 24.96,
+      "grad_norm": 1.2745096683502197,
+      "learning_rate": 2e-05,
+      "loss": 0.04563636,
+      "step": 12480
+    },
+    {
+      "epoch": 24.962,
+      "grad_norm": 1.8629027605056763,
+      "learning_rate": 2e-05,
+      "loss": 0.04409952,
+      "step": 12481
+    },
+    {
+      "epoch": 24.964,
+      "grad_norm": 1.1554667949676514,
+      "learning_rate": 2e-05,
+      "loss": 0.04890275,
+      "step": 12482
+    },
+    {
+      "epoch": 24.966,
+      "grad_norm": 1.329375147819519,
+      "learning_rate": 2e-05,
+      "loss": 0.04882819,
+      "step": 12483
+    },
+    {
+      "epoch": 24.968,
+      "grad_norm": 1.0821406841278076,
+      "learning_rate": 2e-05,
+      "loss": 0.03480864,
+      "step": 12484
+    },
+    {
+      "epoch": 24.97,
+      "grad_norm": 1.5325900316238403,
+      "learning_rate": 2e-05,
+      "loss": 0.04224129,
+      "step": 12485
+    },
+    {
+      "epoch": 24.972,
+      "grad_norm": 1.1319386959075928,
+      "learning_rate": 2e-05,
+      "loss": 0.03097108,
+      "step": 12486
+    },
+    {
+      "epoch": 24.974,
+      "grad_norm": 1.6647253036499023,
+      "learning_rate": 2e-05,
+      "loss": 0.05010362,
+      "step": 12487
+    },
+    {
+      "epoch": 24.976,
+      "grad_norm": 1.5213617086410522,
+      "learning_rate": 2e-05,
+      "loss": 0.05155212,
+      "step": 12488
+    },
+    {
+      "epoch": 24.978,
+      "grad_norm": 1.181519865989685,
+      "learning_rate": 2e-05,
+      "loss": 0.05764193,
+      "step": 12489
+    },
+    {
+      "epoch": 24.98,
+      "grad_norm": 1.0516437292099,
+      "learning_rate": 2e-05,
+      "loss": 0.04471372,
+      "step": 12490
+    },
+    {
+      "epoch": 24.982,
+      "grad_norm": 1.2531055212020874,
+      "learning_rate": 2e-05,
+      "loss": 0.04816005,
+      "step": 12491
+    },
+    {
+      "epoch": 24.984,
+      "grad_norm": 2.094572067260742,
+      "learning_rate": 2e-05,
+      "loss": 0.0658077,
+      "step": 12492
+    },
+    {
+      "epoch": 24.986,
+      "grad_norm": 1.2252720594406128,
+      "learning_rate": 2e-05,
+      "loss": 0.04415631,
+      "step": 12493
+    },
+    {
+      "epoch": 24.988,
+      "grad_norm": 1.2434792518615723,
+      "learning_rate": 2e-05,
+      "loss": 0.04903311,
+      "step": 12494
+    },
+    {
+      "epoch": 24.99,
+      "grad_norm": 0.994875967502594,
+      "learning_rate": 2e-05,
+      "loss": 0.03526482,
+      "step": 12495
+    },
+    {
+      "epoch": 24.992,
+      "grad_norm": 1.4973106384277344,
+      "learning_rate": 2e-05,
+      "loss": 0.03749626,
+      "step": 12496
+    },
+    {
+      "epoch": 24.994,
+      "grad_norm": 0.9715427160263062,
+      "learning_rate": 2e-05,
+      "loss": 0.0366944,
+      "step": 12497
+    },
+    {
+      "epoch": 24.996,
+      "grad_norm": 1.1738646030426025,
+      "learning_rate": 2e-05,
+      "loss": 0.03677404,
+      "step": 12498
+    },
+    {
+      "epoch": 24.998,
+      "grad_norm": 3.087301731109619,
+      "learning_rate": 2e-05,
+      "loss": 0.06307375,
+      "step": 12499
+    },
+    {
+      "epoch": 25.0,
+      "grad_norm": 1.783982276916504,
+      "learning_rate": 2e-05,
+      "loss": 0.04853023,
+      "step": 12500
+    },
+    {
+      "epoch": 25.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9720558882235529,
+        "Equal_1": 0.996,
+        "Equal_2": 0.9840319361277445,
+        "Equal_3": 0.9840319361277445,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.998003992015968,
+        "Parallel_1": 0.9899799599198397,
+        "Parallel_2": 0.9959919839679359,
+        "Parallel_3": 0.988,
+        "Perpendicular_1": 0.992,
+        "Perpendicular_2": 0.984,
+        "Perpendicular_3": 0.8697394789579158,
+        "PointLiesOnCircle_1": 0.9959919839679359,
+        "PointLiesOnCircle_2": 0.9916666666666667,
+        "PointLiesOnCircle_3": 0.9912000000000001,
+        "PointLiesOnLine_1": 0.9959919839679359,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9800399201596807
+      },
+      "eval_runtime": 320.1532,
+      "eval_samples_per_second": 32.797,
+      "eval_steps_per_second": 0.656,
+      "step": 12500
+    },
+    {
+      "epoch": 25.002,
+      "grad_norm": 2.553316354751587,
+      "learning_rate": 2e-05,
+      "loss": 0.04658042,
+      "step": 12501
+    },
+    {
+      "epoch": 25.004,
+      "grad_norm": 1.2298393249511719,
+      "learning_rate": 2e-05,
+      "loss": 0.04537143,
+      "step": 12502
+    },
+    {
+      "epoch": 25.006,
+      "grad_norm": 1.3164474964141846,
+      "learning_rate": 2e-05,
+      "loss": 0.06407714,
+      "step": 12503
+    },
+    {
+      "epoch": 25.008,
+      "grad_norm": 1.3401498794555664,
+      "learning_rate": 2e-05,
+      "loss": 0.04077845,
+      "step": 12504
+    },
+    {
+      "epoch": 25.01,
+      "grad_norm": 1.0891218185424805,
+      "learning_rate": 2e-05,
+      "loss": 0.02724287,
+      "step": 12505
+    },
+    {
+      "epoch": 25.012,
+      "grad_norm": 1.2074103355407715,
+      "learning_rate": 2e-05,
+      "loss": 0.05306327,
+      "step": 12506
+    },
+    {
+      "epoch": 25.014,
+      "grad_norm": 1.1338789463043213,
+      "learning_rate": 2e-05,
+      "loss": 0.04252898,
+      "step": 12507
+    },
+    {
+      "epoch": 25.016,
+      "grad_norm": 1.2618156671524048,
+      "learning_rate": 2e-05,
+      "loss": 0.0498926,
+      "step": 12508
+    },
+    {
+      "epoch": 25.018,
+      "grad_norm": 1.14939546585083,
+      "learning_rate": 2e-05,
+      "loss": 0.04680283,
+      "step": 12509
+    },
+    {
+      "epoch": 25.02,
+      "grad_norm": 1.1815592050552368,
+      "learning_rate": 2e-05,
+      "loss": 0.04917777,
+      "step": 12510
+    },
+    {
+      "epoch": 25.022,
+      "grad_norm": 1.619907259941101,
+      "learning_rate": 2e-05,
+      "loss": 0.05033459,
+      "step": 12511
+    },
+    {
+      "epoch": 25.024,
+      "grad_norm": 1.0121499300003052,
+      "learning_rate": 2e-05,
+      "loss": 0.03902668,
+      "step": 12512
+    },
+    {
+      "epoch": 25.026,
+      "grad_norm": 1.4635061025619507,
+      "learning_rate": 2e-05,
+      "loss": 0.03216758,
+      "step": 12513
+    },
+    {
+      "epoch": 25.028,
+      "grad_norm": 1.818674087524414,
+      "learning_rate": 2e-05,
+      "loss": 0.05529295,
+      "step": 12514
+    },
+    {
+      "epoch": 25.03,
+      "grad_norm": 1.5501385927200317,
+      "learning_rate": 2e-05,
+      "loss": 0.05194931,
+      "step": 12515
+    },
+    {
+      "epoch": 25.032,
+      "grad_norm": 1.735026240348816,
+      "learning_rate": 2e-05,
+      "loss": 0.0563397,
+      "step": 12516
+    },
+    {
+      "epoch": 25.034,
+      "grad_norm": 1.9267089366912842,
+      "learning_rate": 2e-05,
+      "loss": 0.05115256,
+      "step": 12517
+    },
+    {
+      "epoch": 25.036,
+      "grad_norm": 1.591173768043518,
+      "learning_rate": 2e-05,
+      "loss": 0.04865324,
+      "step": 12518
+    },
+    {
+      "epoch": 25.038,
+      "grad_norm": 1.233425259590149,
+      "learning_rate": 2e-05,
+      "loss": 0.04186363,
+      "step": 12519
+    },
+    {
+      "epoch": 25.04,
+      "grad_norm": 1.3910762071609497,
+      "learning_rate": 2e-05,
+      "loss": 0.05607095,
+      "step": 12520
+    },
+    {
+      "epoch": 25.042,
+      "grad_norm": 1.911817193031311,
+      "learning_rate": 2e-05,
+      "loss": 0.04746141,
+      "step": 12521
+    },
+    {
+      "epoch": 25.044,
+      "grad_norm": 1.5743073225021362,
+      "learning_rate": 2e-05,
+      "loss": 0.0310831,
+      "step": 12522
+    },
+    {
+      "epoch": 25.046,
+      "grad_norm": 2.0004522800445557,
+      "learning_rate": 2e-05,
+      "loss": 0.05432073,
+      "step": 12523
+    },
+    {
+      "epoch": 25.048,
+      "grad_norm": 1.304870367050171,
+      "learning_rate": 2e-05,
+      "loss": 0.05191491,
+      "step": 12524
+    },
+    {
+      "epoch": 25.05,
+      "grad_norm": 1.6975371837615967,
+      "learning_rate": 2e-05,
+      "loss": 0.04251885,
+      "step": 12525
+    },
+    {
+      "epoch": 25.052,
+      "grad_norm": 1.4703913927078247,
+      "learning_rate": 2e-05,
+      "loss": 0.04141282,
+      "step": 12526
+    },
+    {
+      "epoch": 25.054,
+      "grad_norm": 1.6704292297363281,
+      "learning_rate": 2e-05,
+      "loss": 0.05951826,
+      "step": 12527
+    },
+    {
+      "epoch": 25.056,
+      "grad_norm": 1.4327316284179688,
+      "learning_rate": 2e-05,
+      "loss": 0.05041571,
+      "step": 12528
+    },
+    {
+      "epoch": 25.058,
+      "grad_norm": 2.054896116256714,
+      "learning_rate": 2e-05,
+      "loss": 0.05156599,
+      "step": 12529
+    },
+    {
+      "epoch": 25.06,
+      "grad_norm": 1.0578651428222656,
+      "learning_rate": 2e-05,
+      "loss": 0.03640224,
+      "step": 12530
+    },
+    {
+      "epoch": 25.062,
+      "grad_norm": 1.2330236434936523,
+      "learning_rate": 2e-05,
+      "loss": 0.03764113,
+      "step": 12531
+    },
+    {
+      "epoch": 25.064,
+      "grad_norm": 1.5884158611297607,
+      "learning_rate": 2e-05,
+      "loss": 0.04582855,
+      "step": 12532
+    },
+    {
+      "epoch": 25.066,
+      "grad_norm": 1.3398418426513672,
+      "learning_rate": 2e-05,
+      "loss": 0.04462974,
+      "step": 12533
+    },
+    {
+      "epoch": 25.068,
+      "grad_norm": 1.4082545042037964,
+      "learning_rate": 2e-05,
+      "loss": 0.03973638,
+      "step": 12534
+    },
+    {
+      "epoch": 25.07,
+      "grad_norm": 3.2086148262023926,
+      "learning_rate": 2e-05,
+      "loss": 0.0822015,
+      "step": 12535
+    },
+    {
+      "epoch": 25.072,
+      "grad_norm": 1.2738802433013916,
+      "learning_rate": 2e-05,
+      "loss": 0.0383213,
+      "step": 12536
+    },
+    {
+      "epoch": 25.074,
+      "grad_norm": 1.359592318534851,
+      "learning_rate": 2e-05,
+      "loss": 0.04718591,
+      "step": 12537
+    },
+    {
+      "epoch": 25.076,
+      "grad_norm": 2.490873336791992,
+      "learning_rate": 2e-05,
+      "loss": 0.04099283,
+      "step": 12538
+    },
+    {
+      "epoch": 25.078,
+      "grad_norm": 1.3400729894638062,
+      "learning_rate": 2e-05,
+      "loss": 0.05246966,
+      "step": 12539
+    },
+    {
+      "epoch": 25.08,
+      "grad_norm": 1.2105579376220703,
+      "learning_rate": 2e-05,
+      "loss": 0.04603817,
+      "step": 12540
+    },
+    {
+      "epoch": 25.082,
+      "grad_norm": 1.2426912784576416,
+      "learning_rate": 2e-05,
+      "loss": 0.04203958,
+      "step": 12541
+    },
+    {
+      "epoch": 25.084,
+      "grad_norm": 1.232663869857788,
+      "learning_rate": 2e-05,
+      "loss": 0.03772744,
+      "step": 12542
+    },
+    {
+      "epoch": 25.086,
+      "grad_norm": 1.6726223230361938,
+      "learning_rate": 2e-05,
+      "loss": 0.04957581,
+      "step": 12543
+    },
+    {
+      "epoch": 25.088,
+      "grad_norm": 1.9126125574111938,
+      "learning_rate": 2e-05,
+      "loss": 0.05544265,
+      "step": 12544
+    },
+    {
+      "epoch": 25.09,
+      "grad_norm": 1.0108952522277832,
+      "learning_rate": 2e-05,
+      "loss": 0.02635121,
+      "step": 12545
+    },
+    {
+      "epoch": 25.092,
+      "grad_norm": 1.3411767482757568,
+      "learning_rate": 2e-05,
+      "loss": 0.03911927,
+      "step": 12546
+    },
+    {
+      "epoch": 25.094,
+      "grad_norm": 2.594507932662964,
+      "learning_rate": 2e-05,
+      "loss": 0.05281266,
+      "step": 12547
+    },
+    {
+      "epoch": 25.096,
+      "grad_norm": 1.0433748960494995,
+      "learning_rate": 2e-05,
+      "loss": 0.04605139,
+      "step": 12548
+    },
+    {
+      "epoch": 25.098,
+      "grad_norm": 1.9768903255462646,
+      "learning_rate": 2e-05,
+      "loss": 0.0549066,
+      "step": 12549
+    },
+    {
+      "epoch": 25.1,
+      "grad_norm": 2.1046414375305176,
+      "learning_rate": 2e-05,
+      "loss": 0.05329197,
+      "step": 12550
+    },
+    {
+      "epoch": 25.102,
+      "grad_norm": 1.1665661334991455,
+      "learning_rate": 2e-05,
+      "loss": 0.04764657,
+      "step": 12551
+    },
+    {
+      "epoch": 25.104,
+      "grad_norm": 2.242833137512207,
+      "learning_rate": 2e-05,
+      "loss": 0.04974656,
+      "step": 12552
+    },
+    {
+      "epoch": 25.106,
+      "grad_norm": 1.416796326637268,
+      "learning_rate": 2e-05,
+      "loss": 0.04945759,
+      "step": 12553
+    },
+    {
+      "epoch": 25.108,
+      "grad_norm": 1.1873366832733154,
+      "learning_rate": 2e-05,
+      "loss": 0.05103811,
+      "step": 12554
+    },
+    {
+      "epoch": 25.11,
+      "grad_norm": 1.926827311515808,
+      "learning_rate": 2e-05,
+      "loss": 0.05267503,
+      "step": 12555
+    },
+    {
+      "epoch": 25.112,
+      "grad_norm": 0.9494261741638184,
+      "learning_rate": 2e-05,
+      "loss": 0.03307255,
+      "step": 12556
+    },
+    {
+      "epoch": 25.114,
+      "grad_norm": 1.8937228918075562,
+      "learning_rate": 2e-05,
+      "loss": 0.06592927,
+      "step": 12557
+    },
+    {
+      "epoch": 25.116,
+      "grad_norm": 2.0769922733306885,
+      "learning_rate": 2e-05,
+      "loss": 0.05289575,
+      "step": 12558
+    },
+    {
+      "epoch": 25.118,
+      "grad_norm": 1.1254668235778809,
+      "learning_rate": 2e-05,
+      "loss": 0.03823429,
+      "step": 12559
+    },
+    {
+      "epoch": 25.12,
+      "grad_norm": 2.4549336433410645,
+      "learning_rate": 2e-05,
+      "loss": 0.04544316,
+      "step": 12560
+    },
+    {
+      "epoch": 25.122,
+      "grad_norm": 1.7756479978561401,
+      "learning_rate": 2e-05,
+      "loss": 0.0518619,
+      "step": 12561
+    },
+    {
+      "epoch": 25.124,
+      "grad_norm": 1.4617094993591309,
+      "learning_rate": 2e-05,
+      "loss": 0.05234993,
+      "step": 12562
+    },
+    {
+      "epoch": 25.126,
+      "grad_norm": 1.5612094402313232,
+      "learning_rate": 2e-05,
+      "loss": 0.04799661,
+      "step": 12563
+    },
+    {
+      "epoch": 25.128,
+      "grad_norm": 1.2895623445510864,
+      "learning_rate": 2e-05,
+      "loss": 0.04298138,
+      "step": 12564
+    },
+    {
+      "epoch": 25.13,
+      "grad_norm": 1.085653305053711,
+      "learning_rate": 2e-05,
+      "loss": 0.03644668,
+      "step": 12565
+    },
+    {
+      "epoch": 25.132,
+      "grad_norm": 2.4465861320495605,
+      "learning_rate": 2e-05,
+      "loss": 0.04523169,
+      "step": 12566
+    },
+    {
+      "epoch": 25.134,
+      "grad_norm": 1.476048231124878,
+      "learning_rate": 2e-05,
+      "loss": 0.02938263,
+      "step": 12567
+    },
+    {
+      "epoch": 25.136,
+      "grad_norm": 0.9077489376068115,
+      "learning_rate": 2e-05,
+      "loss": 0.02951773,
+      "step": 12568
+    },
+    {
+      "epoch": 25.138,
+      "grad_norm": 1.348293662071228,
+      "learning_rate": 2e-05,
+      "loss": 0.04014814,
+      "step": 12569
+    },
+    {
+      "epoch": 25.14,
+      "grad_norm": 1.2255805730819702,
+      "learning_rate": 2e-05,
+      "loss": 0.04978126,
+      "step": 12570
+    },
+    {
+      "epoch": 25.142,
+      "grad_norm": 1.4020949602127075,
+      "learning_rate": 2e-05,
+      "loss": 0.05750665,
+      "step": 12571
+    },
+    {
+      "epoch": 25.144,
+      "grad_norm": 1.526517629623413,
+      "learning_rate": 2e-05,
+      "loss": 0.04812055,
+      "step": 12572
+    },
+    {
+      "epoch": 25.146,
+      "grad_norm": 1.3054735660552979,
+      "learning_rate": 2e-05,
+      "loss": 0.04948972,
+      "step": 12573
+    },
+    {
+      "epoch": 25.148,
+      "grad_norm": 0.968127429485321,
+      "learning_rate": 2e-05,
+      "loss": 0.03364644,
+      "step": 12574
+    },
+    {
+      "epoch": 25.15,
+      "grad_norm": 2.5585010051727295,
+      "learning_rate": 2e-05,
+      "loss": 0.06127685,
+      "step": 12575
+    },
+    {
+      "epoch": 25.152,
+      "grad_norm": 1.267968773841858,
+      "learning_rate": 2e-05,
+      "loss": 0.05282283,
+      "step": 12576
+    },
+    {
+      "epoch": 25.154,
+      "grad_norm": 1.3801593780517578,
+      "learning_rate": 2e-05,
+      "loss": 0.0542292,
+      "step": 12577
+    },
+    {
+      "epoch": 25.156,
+      "grad_norm": 1.1124210357666016,
+      "learning_rate": 2e-05,
+      "loss": 0.03990389,
+      "step": 12578
+    },
+    {
+      "epoch": 25.158,
+      "grad_norm": 1.1204739809036255,
+      "learning_rate": 2e-05,
+      "loss": 0.04882549,
+      "step": 12579
+    },
+    {
+      "epoch": 25.16,
+      "grad_norm": 1.5411254167556763,
+      "learning_rate": 2e-05,
+      "loss": 0.03279787,
+      "step": 12580
+    },
+    {
+      "epoch": 25.162,
+      "grad_norm": 1.0646896362304688,
+      "learning_rate": 2e-05,
+      "loss": 0.04206897,
+      "step": 12581
+    },
+    {
+      "epoch": 25.164,
+      "grad_norm": 1.4947314262390137,
+      "learning_rate": 2e-05,
+      "loss": 0.04689228,
+      "step": 12582
+    },
+    {
+      "epoch": 25.166,
+      "grad_norm": 1.3432666063308716,
+      "learning_rate": 2e-05,
+      "loss": 0.03807723,
+      "step": 12583
+    },
+    {
+      "epoch": 25.168,
+      "grad_norm": 1.3646643161773682,
+      "learning_rate": 2e-05,
+      "loss": 0.05141376,
+      "step": 12584
+    },
+    {
+      "epoch": 25.17,
+      "grad_norm": 2.0790557861328125,
+      "learning_rate": 2e-05,
+      "loss": 0.04800526,
+      "step": 12585
+    },
+    {
+      "epoch": 25.172,
+      "grad_norm": 1.5363324880599976,
+      "learning_rate": 2e-05,
+      "loss": 0.03496822,
+      "step": 12586
+    },
+    {
+      "epoch": 25.174,
+      "grad_norm": 1.1385252475738525,
+      "learning_rate": 2e-05,
+      "loss": 0.04525664,
+      "step": 12587
+    },
+    {
+      "epoch": 25.176,
+      "grad_norm": 1.0282803773880005,
+      "learning_rate": 2e-05,
+      "loss": 0.03413811,
+      "step": 12588
+    },
+    {
+      "epoch": 25.178,
+      "grad_norm": 1.563729166984558,
+      "learning_rate": 2e-05,
+      "loss": 0.04510281,
+      "step": 12589
+    },
+    {
+      "epoch": 25.18,
+      "grad_norm": 2.2484776973724365,
+      "learning_rate": 2e-05,
+      "loss": 0.05236363,
+      "step": 12590
+    },
+    {
+      "epoch": 25.182,
+      "grad_norm": 1.737161636352539,
+      "learning_rate": 2e-05,
+      "loss": 0.04524446,
+      "step": 12591
+    },
+    {
+      "epoch": 25.184,
+      "grad_norm": 1.867302656173706,
+      "learning_rate": 2e-05,
+      "loss": 0.06257963,
+      "step": 12592
+    },
+    {
+      "epoch": 25.186,
+      "grad_norm": 1.426841139793396,
+      "learning_rate": 2e-05,
+      "loss": 0.04176682,
+      "step": 12593
+    },
+    {
+      "epoch": 25.188,
+      "grad_norm": 1.0782243013381958,
+      "learning_rate": 2e-05,
+      "loss": 0.03285508,
+      "step": 12594
+    },
+    {
+      "epoch": 25.19,
+      "grad_norm": 3.013762950897217,
+      "learning_rate": 2e-05,
+      "loss": 0.05704542,
+      "step": 12595
+    },
+    {
+      "epoch": 25.192,
+      "grad_norm": 1.8257044553756714,
+      "learning_rate": 2e-05,
+      "loss": 0.04196394,
+      "step": 12596
+    },
+    {
+      "epoch": 25.194,
+      "grad_norm": 0.9650020599365234,
+      "learning_rate": 2e-05,
+      "loss": 0.03296274,
+      "step": 12597
+    },
+    {
+      "epoch": 25.196,
+      "grad_norm": 2.078765630722046,
+      "learning_rate": 2e-05,
+      "loss": 0.04554818,
+      "step": 12598
+    },
+    {
+      "epoch": 25.198,
+      "grad_norm": 1.3087246417999268,
+      "learning_rate": 2e-05,
+      "loss": 0.05761782,
+      "step": 12599
+    },
+    {
+      "epoch": 25.2,
+      "grad_norm": 1.3766916990280151,
+      "learning_rate": 2e-05,
+      "loss": 0.052761,
+      "step": 12600
+    },
+    {
+      "epoch": 25.202,
+      "grad_norm": 1.2516851425170898,
+      "learning_rate": 2e-05,
+      "loss": 0.04158796,
+      "step": 12601
+    },
+    {
+      "epoch": 25.204,
+      "grad_norm": 1.0545592308044434,
+      "learning_rate": 2e-05,
+      "loss": 0.03375325,
+      "step": 12602
+    },
+    {
+      "epoch": 25.206,
+      "grad_norm": 1.732971429824829,
+      "learning_rate": 2e-05,
+      "loss": 0.05470074,
+      "step": 12603
+    },
+    {
+      "epoch": 25.208,
+      "grad_norm": 1.1150429248809814,
+      "learning_rate": 2e-05,
+      "loss": 0.03985225,
+      "step": 12604
+    },
+    {
+      "epoch": 25.21,
+      "grad_norm": 1.1678208112716675,
+      "learning_rate": 2e-05,
+      "loss": 0.04104099,
+      "step": 12605
+    },
+    {
+      "epoch": 25.212,
+      "grad_norm": 1.085423469543457,
+      "learning_rate": 2e-05,
+      "loss": 0.04031212,
+      "step": 12606
+    },
+    {
+      "epoch": 25.214,
+      "grad_norm": 1.413877248764038,
+      "learning_rate": 2e-05,
+      "loss": 0.04183791,
+      "step": 12607
+    },
+    {
+      "epoch": 25.216,
+      "grad_norm": 1.270094871520996,
+      "learning_rate": 2e-05,
+      "loss": 0.03035654,
+      "step": 12608
+    },
+    {
+      "epoch": 25.218,
+      "grad_norm": 1.045606017112732,
+      "learning_rate": 2e-05,
+      "loss": 0.03988522,
+      "step": 12609
+    },
+    {
+      "epoch": 25.22,
+      "grad_norm": 1.19292151927948,
+      "learning_rate": 2e-05,
+      "loss": 0.03305714,
+      "step": 12610
+    },
+    {
+      "epoch": 25.222,
+      "grad_norm": 1.377384901046753,
+      "learning_rate": 2e-05,
+      "loss": 0.0412189,
+      "step": 12611
+    },
+    {
+      "epoch": 25.224,
+      "grad_norm": 1.903836727142334,
+      "learning_rate": 2e-05,
+      "loss": 0.05559286,
+      "step": 12612
+    },
+    {
+      "epoch": 25.226,
+      "grad_norm": 1.2150182723999023,
+      "learning_rate": 2e-05,
+      "loss": 0.04249961,
+      "step": 12613
+    },
+    {
+      "epoch": 25.228,
+      "grad_norm": 1.5361416339874268,
+      "learning_rate": 2e-05,
+      "loss": 0.05546279,
+      "step": 12614
+    },
+    {
+      "epoch": 25.23,
+      "grad_norm": 1.1674758195877075,
+      "learning_rate": 2e-05,
+      "loss": 0.03651671,
+      "step": 12615
+    },
+    {
+      "epoch": 25.232,
+      "grad_norm": 1.5208848714828491,
+      "learning_rate": 2e-05,
+      "loss": 0.06328288,
+      "step": 12616
+    },
+    {
+      "epoch": 25.234,
+      "grad_norm": 1.5309934616088867,
+      "learning_rate": 2e-05,
+      "loss": 0.06355253,
+      "step": 12617
+    },
+    {
+      "epoch": 25.236,
+      "grad_norm": 1.5100221633911133,
+      "learning_rate": 2e-05,
+      "loss": 0.04278588,
+      "step": 12618
+    },
+    {
+      "epoch": 25.238,
+      "grad_norm": 1.191089153289795,
+      "learning_rate": 2e-05,
+      "loss": 0.04086279,
+      "step": 12619
+    },
+    {
+      "epoch": 25.24,
+      "grad_norm": 1.7226134538650513,
+      "learning_rate": 2e-05,
+      "loss": 0.04553945,
+      "step": 12620
+    },
+    {
+      "epoch": 25.242,
+      "grad_norm": 1.6444710493087769,
+      "learning_rate": 2e-05,
+      "loss": 0.04359776,
+      "step": 12621
+    },
+    {
+      "epoch": 25.244,
+      "grad_norm": 1.3565912246704102,
+      "learning_rate": 2e-05,
+      "loss": 0.04153623,
+      "step": 12622
+    },
+    {
+      "epoch": 25.246,
+      "grad_norm": 1.295622706413269,
+      "learning_rate": 2e-05,
+      "loss": 0.03719122,
+      "step": 12623
+    },
+    {
+      "epoch": 25.248,
+      "grad_norm": 1.560507893562317,
+      "learning_rate": 2e-05,
+      "loss": 0.04558239,
+      "step": 12624
+    },
+    {
+      "epoch": 25.25,
+      "grad_norm": 1.1936217546463013,
+      "learning_rate": 2e-05,
+      "loss": 0.04132167,
+      "step": 12625
+    },
+    {
+      "epoch": 25.252,
+      "grad_norm": 1.165188193321228,
+      "learning_rate": 2e-05,
+      "loss": 0.04114609,
+      "step": 12626
+    },
+    {
+      "epoch": 25.254,
+      "grad_norm": 1.5809382200241089,
+      "learning_rate": 2e-05,
+      "loss": 0.04351167,
+      "step": 12627
+    },
+    {
+      "epoch": 25.256,
+      "grad_norm": 1.449571132659912,
+      "learning_rate": 2e-05,
+      "loss": 0.0306358,
+      "step": 12628
+    },
+    {
+      "epoch": 25.258,
+      "grad_norm": 1.3308436870574951,
+      "learning_rate": 2e-05,
+      "loss": 0.02577632,
+      "step": 12629
+    },
+    {
+      "epoch": 25.26,
+      "grad_norm": 1.2263076305389404,
+      "learning_rate": 2e-05,
+      "loss": 0.03553704,
+      "step": 12630
+    },
+    {
+      "epoch": 25.262,
+      "grad_norm": 1.204790711402893,
+      "learning_rate": 2e-05,
+      "loss": 0.04875295,
+      "step": 12631
+    },
+    {
+      "epoch": 25.264,
+      "grad_norm": 1.111572504043579,
+      "learning_rate": 2e-05,
+      "loss": 0.04481274,
+      "step": 12632
+    },
+    {
+      "epoch": 25.266,
+      "grad_norm": 1.6431896686553955,
+      "learning_rate": 2e-05,
+      "loss": 0.05574667,
+      "step": 12633
+    },
+    {
+      "epoch": 25.268,
+      "grad_norm": 1.5125669240951538,
+      "learning_rate": 2e-05,
+      "loss": 0.05090974,
+      "step": 12634
+    },
+    {
+      "epoch": 25.27,
+      "grad_norm": 6.2000932693481445,
+      "learning_rate": 2e-05,
+      "loss": 0.04442751,
+      "step": 12635
+    },
+    {
+      "epoch": 25.272,
+      "grad_norm": 4.573184967041016,
+      "learning_rate": 2e-05,
+      "loss": 0.06365525,
+      "step": 12636
+    },
+    {
+      "epoch": 25.274,
+      "grad_norm": 2.154564142227173,
+      "learning_rate": 2e-05,
+      "loss": 0.03896628,
+      "step": 12637
+    },
+    {
+      "epoch": 25.276,
+      "grad_norm": 1.2963447570800781,
+      "learning_rate": 2e-05,
+      "loss": 0.04430124,
+      "step": 12638
+    },
+    {
+      "epoch": 25.278,
+      "grad_norm": 1.1166437864303589,
+      "learning_rate": 2e-05,
+      "loss": 0.03765798,
+      "step": 12639
+    },
+    {
+      "epoch": 25.28,
+      "grad_norm": 1.1233835220336914,
+      "learning_rate": 2e-05,
+      "loss": 0.03647223,
+      "step": 12640
+    },
+    {
+      "epoch": 25.282,
+      "grad_norm": 1.2153040170669556,
+      "learning_rate": 2e-05,
+      "loss": 0.04259994,
+      "step": 12641
+    },
+    {
+      "epoch": 25.284,
+      "grad_norm": 1.6373004913330078,
+      "learning_rate": 2e-05,
+      "loss": 0.05044476,
+      "step": 12642
+    },
+    {
+      "epoch": 25.286,
+      "grad_norm": 1.4357736110687256,
+      "learning_rate": 2e-05,
+      "loss": 0.04274999,
+      "step": 12643
+    },
+    {
+      "epoch": 25.288,
+      "grad_norm": 1.0987167358398438,
+      "learning_rate": 2e-05,
+      "loss": 0.04832689,
+      "step": 12644
+    },
+    {
+      "epoch": 25.29,
+      "grad_norm": 1.4695364236831665,
+      "learning_rate": 2e-05,
+      "loss": 0.04650634,
+      "step": 12645
+    },
+    {
+      "epoch": 25.292,
+      "grad_norm": 1.8129997253417969,
+      "learning_rate": 2e-05,
+      "loss": 0.03666451,
+      "step": 12646
+    },
+    {
+      "epoch": 25.294,
+      "grad_norm": 1.2179253101348877,
+      "learning_rate": 2e-05,
+      "loss": 0.03910381,
+      "step": 12647
+    },
+    {
+      "epoch": 25.296,
+      "grad_norm": 1.5008174180984497,
+      "learning_rate": 2e-05,
+      "loss": 0.05692586,
+      "step": 12648
+    },
+    {
+      "epoch": 25.298,
+      "grad_norm": 0.9341026544570923,
+      "learning_rate": 2e-05,
+      "loss": 0.03112362,
+      "step": 12649
+    },
+    {
+      "epoch": 25.3,
+      "grad_norm": 0.8844679594039917,
+      "learning_rate": 2e-05,
+      "loss": 0.02362539,
+      "step": 12650
+    },
+    {
+      "epoch": 25.302,
+      "grad_norm": 1.7629514932632446,
+      "learning_rate": 2e-05,
+      "loss": 0.03517685,
+      "step": 12651
+    },
+    {
+      "epoch": 25.304,
+      "grad_norm": 1.303821325302124,
+      "learning_rate": 2e-05,
+      "loss": 0.05173323,
+      "step": 12652
+    },
+    {
+      "epoch": 25.306,
+      "grad_norm": 1.4266308546066284,
+      "learning_rate": 2e-05,
+      "loss": 0.05540767,
+      "step": 12653
+    },
+    {
+      "epoch": 25.308,
+      "grad_norm": 1.2281945943832397,
+      "learning_rate": 2e-05,
+      "loss": 0.04114369,
+      "step": 12654
+    },
+    {
+      "epoch": 25.31,
+      "grad_norm": 1.3700491189956665,
+      "learning_rate": 2e-05,
+      "loss": 0.03522223,
+      "step": 12655
+    },
+    {
+      "epoch": 25.312,
+      "grad_norm": 1.426082968711853,
+      "learning_rate": 2e-05,
+      "loss": 0.06648382,
+      "step": 12656
+    },
+    {
+      "epoch": 25.314,
+      "grad_norm": 1.3388099670410156,
+      "learning_rate": 2e-05,
+      "loss": 0.05874391,
+      "step": 12657
+    },
+    {
+      "epoch": 25.316,
+      "grad_norm": 0.9311839938163757,
+      "learning_rate": 2e-05,
+      "loss": 0.0290327,
+      "step": 12658
+    },
+    {
+      "epoch": 25.318,
+      "grad_norm": 1.222132682800293,
+      "learning_rate": 2e-05,
+      "loss": 0.03191571,
+      "step": 12659
+    },
+    {
+      "epoch": 25.32,
+      "grad_norm": 1.1202843189239502,
+      "learning_rate": 2e-05,
+      "loss": 0.03595486,
+      "step": 12660
+    },
+    {
+      "epoch": 25.322,
+      "grad_norm": 1.4704622030258179,
+      "learning_rate": 2e-05,
+      "loss": 0.04287569,
+      "step": 12661
+    },
+    {
+      "epoch": 25.324,
+      "grad_norm": 0.9908611178398132,
+      "learning_rate": 2e-05,
+      "loss": 0.03052938,
+      "step": 12662
+    },
+    {
+      "epoch": 25.326,
+      "grad_norm": 1.0778824090957642,
+      "learning_rate": 2e-05,
+      "loss": 0.04454803,
+      "step": 12663
+    },
+    {
+      "epoch": 25.328,
+      "grad_norm": 1.4507533311843872,
+      "learning_rate": 2e-05,
+      "loss": 0.04375216,
+      "step": 12664
+    },
+    {
+      "epoch": 25.33,
+      "grad_norm": 1.0799167156219482,
+      "learning_rate": 2e-05,
+      "loss": 0.03894107,
+      "step": 12665
+    },
+    {
+      "epoch": 25.332,
+      "grad_norm": 1.2972630262374878,
+      "learning_rate": 2e-05,
+      "loss": 0.03249099,
+      "step": 12666
+    },
+    {
+      "epoch": 25.334,
+      "grad_norm": 1.2298357486724854,
+      "learning_rate": 2e-05,
+      "loss": 0.04411597,
+      "step": 12667
+    },
+    {
+      "epoch": 25.336,
+      "grad_norm": 1.2464599609375,
+      "learning_rate": 2e-05,
+      "loss": 0.03922657,
+      "step": 12668
+    },
+    {
+      "epoch": 25.338,
+      "grad_norm": 1.6811728477478027,
+      "learning_rate": 2e-05,
+      "loss": 0.05457025,
+      "step": 12669
+    },
+    {
+      "epoch": 25.34,
+      "grad_norm": 1.4204384088516235,
+      "learning_rate": 2e-05,
+      "loss": 0.03159765,
+      "step": 12670
+    },
+    {
+      "epoch": 25.342,
+      "grad_norm": 1.5988712310791016,
+      "learning_rate": 2e-05,
+      "loss": 0.04662495,
+      "step": 12671
+    },
+    {
+      "epoch": 25.344,
+      "grad_norm": 1.4310500621795654,
+      "learning_rate": 2e-05,
+      "loss": 0.05521808,
+      "step": 12672
+    },
+    {
+      "epoch": 25.346,
+      "grad_norm": 1.4796041250228882,
+      "learning_rate": 2e-05,
+      "loss": 0.05029026,
+      "step": 12673
+    },
+    {
+      "epoch": 25.348,
+      "grad_norm": 1.120280385017395,
+      "learning_rate": 2e-05,
+      "loss": 0.05406159,
+      "step": 12674
+    },
+    {
+      "epoch": 25.35,
+      "grad_norm": 1.5472899675369263,
+      "learning_rate": 2e-05,
+      "loss": 0.04434624,
+      "step": 12675
+    },
+    {
+      "epoch": 25.352,
+      "grad_norm": 1.5988304615020752,
+      "learning_rate": 2e-05,
+      "loss": 0.03512241,
+      "step": 12676
+    },
+    {
+      "epoch": 25.354,
+      "grad_norm": 2.166592836380005,
+      "learning_rate": 2e-05,
+      "loss": 0.05324619,
+      "step": 12677
+    },
+    {
+      "epoch": 25.356,
+      "grad_norm": 1.23325514793396,
+      "learning_rate": 2e-05,
+      "loss": 0.0458478,
+      "step": 12678
+    },
+    {
+      "epoch": 25.358,
+      "grad_norm": 1.305748462677002,
+      "learning_rate": 2e-05,
+      "loss": 0.04038317,
+      "step": 12679
+    },
+    {
+      "epoch": 25.36,
+      "grad_norm": 1.2116440534591675,
+      "learning_rate": 2e-05,
+      "loss": 0.04997266,
+      "step": 12680
+    },
+    {
+      "epoch": 25.362,
+      "grad_norm": 1.0483074188232422,
+      "learning_rate": 2e-05,
+      "loss": 0.03716281,
+      "step": 12681
+    },
+    {
+      "epoch": 25.364,
+      "grad_norm": 1.1586565971374512,
+      "learning_rate": 2e-05,
+      "loss": 0.03767997,
+      "step": 12682
+    },
+    {
+      "epoch": 25.366,
+      "grad_norm": 1.693536400794983,
+      "learning_rate": 2e-05,
+      "loss": 0.05087979,
+      "step": 12683
+    },
+    {
+      "epoch": 25.368,
+      "grad_norm": 1.4441686868667603,
+      "learning_rate": 2e-05,
+      "loss": 0.05642378,
+      "step": 12684
+    },
+    {
+      "epoch": 25.37,
+      "grad_norm": 1.18898344039917,
+      "learning_rate": 2e-05,
+      "loss": 0.04010397,
+      "step": 12685
+    },
+    {
+      "epoch": 25.372,
+      "grad_norm": 1.593220591545105,
+      "learning_rate": 2e-05,
+      "loss": 0.0599516,
+      "step": 12686
+    },
+    {
+      "epoch": 25.374,
+      "grad_norm": 2.1640501022338867,
+      "learning_rate": 2e-05,
+      "loss": 0.04657248,
+      "step": 12687
+    },
+    {
+      "epoch": 25.376,
+      "grad_norm": 1.3017292022705078,
+      "learning_rate": 2e-05,
+      "loss": 0.05840718,
+      "step": 12688
+    },
+    {
+      "epoch": 25.378,
+      "grad_norm": 1.1755383014678955,
+      "learning_rate": 2e-05,
+      "loss": 0.04789887,
+      "step": 12689
+    },
+    {
+      "epoch": 25.38,
+      "grad_norm": 1.9385019540786743,
+      "learning_rate": 2e-05,
+      "loss": 0.05613485,
+      "step": 12690
+    },
+    {
+      "epoch": 25.382,
+      "grad_norm": 1.2281224727630615,
+      "learning_rate": 2e-05,
+      "loss": 0.03748925,
+      "step": 12691
+    },
+    {
+      "epoch": 25.384,
+      "grad_norm": 1.4565155506134033,
+      "learning_rate": 2e-05,
+      "loss": 0.0405966,
+      "step": 12692
+    },
+    {
+      "epoch": 25.386,
+      "grad_norm": 2.12028169631958,
+      "learning_rate": 2e-05,
+      "loss": 0.04600952,
+      "step": 12693
+    },
+    {
+      "epoch": 25.388,
+      "grad_norm": 1.0148862600326538,
+      "learning_rate": 2e-05,
+      "loss": 0.02967937,
+      "step": 12694
+    },
+    {
+      "epoch": 25.39,
+      "grad_norm": 1.3207978010177612,
+      "learning_rate": 2e-05,
+      "loss": 0.06044834,
+      "step": 12695
+    },
+    {
+      "epoch": 25.392,
+      "grad_norm": 1.3644589185714722,
+      "learning_rate": 2e-05,
+      "loss": 0.03912506,
+      "step": 12696
+    },
+    {
+      "epoch": 25.394,
+      "grad_norm": 1.5914479494094849,
+      "learning_rate": 2e-05,
+      "loss": 0.0491058,
+      "step": 12697
+    },
+    {
+      "epoch": 25.396,
+      "grad_norm": 1.4540284872055054,
+      "learning_rate": 2e-05,
+      "loss": 0.04001951,
+      "step": 12698
+    },
+    {
+      "epoch": 25.398,
+      "grad_norm": 1.3614214658737183,
+      "learning_rate": 2e-05,
+      "loss": 0.04715081,
+      "step": 12699
+    },
+    {
+      "epoch": 25.4,
+      "grad_norm": 1.1477137804031372,
+      "learning_rate": 2e-05,
+      "loss": 0.0351803,
+      "step": 12700
+    },
+    {
+      "epoch": 25.402,
+      "grad_norm": 1.382109522819519,
+      "learning_rate": 2e-05,
+      "loss": 0.04908609,
+      "step": 12701
+    },
+    {
+      "epoch": 25.404,
+      "grad_norm": 1.372443675994873,
+      "learning_rate": 2e-05,
+      "loss": 0.06265668,
+      "step": 12702
+    },
+    {
+      "epoch": 25.406,
+      "grad_norm": 1.0587457418441772,
+      "learning_rate": 2e-05,
+      "loss": 0.04351508,
+      "step": 12703
+    },
+    {
+      "epoch": 25.408,
+      "grad_norm": 1.4110264778137207,
+      "learning_rate": 2e-05,
+      "loss": 0.05135802,
+      "step": 12704
+    },
+    {
+      "epoch": 25.41,
+      "grad_norm": 2.1139957904815674,
+      "learning_rate": 2e-05,
+      "loss": 0.03126623,
+      "step": 12705
+    },
+    {
+      "epoch": 25.412,
+      "grad_norm": 1.8020521402359009,
+      "learning_rate": 2e-05,
+      "loss": 0.04075658,
+      "step": 12706
+    },
+    {
+      "epoch": 25.414,
+      "grad_norm": 1.4906790256500244,
+      "learning_rate": 2e-05,
+      "loss": 0.03323084,
+      "step": 12707
+    },
+    {
+      "epoch": 25.416,
+      "grad_norm": 1.2605034112930298,
+      "learning_rate": 2e-05,
+      "loss": 0.04425685,
+      "step": 12708
+    },
+    {
+      "epoch": 25.418,
+      "grad_norm": 1.4229034185409546,
+      "learning_rate": 2e-05,
+      "loss": 0.06664809,
+      "step": 12709
+    },
+    {
+      "epoch": 25.42,
+      "grad_norm": 3.019768476486206,
+      "learning_rate": 2e-05,
+      "loss": 0.06653355,
+      "step": 12710
+    },
+    {
+      "epoch": 25.422,
+      "grad_norm": 1.1232490539550781,
+      "learning_rate": 2e-05,
+      "loss": 0.03858437,
+      "step": 12711
+    },
+    {
+      "epoch": 25.424,
+      "grad_norm": 1.0489226579666138,
+      "learning_rate": 2e-05,
+      "loss": 0.03429198,
+      "step": 12712
+    },
+    {
+      "epoch": 25.426,
+      "grad_norm": 2.398514747619629,
+      "learning_rate": 2e-05,
+      "loss": 0.07073388,
+      "step": 12713
+    },
+    {
+      "epoch": 25.428,
+      "grad_norm": 1.2204501628875732,
+      "learning_rate": 2e-05,
+      "loss": 0.04312594,
+      "step": 12714
+    },
+    {
+      "epoch": 25.43,
+      "grad_norm": 0.9238330721855164,
+      "learning_rate": 2e-05,
+      "loss": 0.03127215,
+      "step": 12715
+    },
+    {
+      "epoch": 25.432,
+      "grad_norm": 1.1325103044509888,
+      "learning_rate": 2e-05,
+      "loss": 0.04861875,
+      "step": 12716
+    },
+    {
+      "epoch": 25.434,
+      "grad_norm": 1.4087941646575928,
+      "learning_rate": 2e-05,
+      "loss": 0.04097962,
+      "step": 12717
+    },
+    {
+      "epoch": 25.436,
+      "grad_norm": 2.770249366760254,
+      "learning_rate": 2e-05,
+      "loss": 0.05767106,
+      "step": 12718
+    },
+    {
+      "epoch": 25.438,
+      "grad_norm": 1.5272942781448364,
+      "learning_rate": 2e-05,
+      "loss": 0.05232708,
+      "step": 12719
+    },
+    {
+      "epoch": 25.44,
+      "grad_norm": 1.8735677003860474,
+      "learning_rate": 2e-05,
+      "loss": 0.04702976,
+      "step": 12720
+    },
+    {
+      "epoch": 25.442,
+      "grad_norm": 1.5433956384658813,
+      "learning_rate": 2e-05,
+      "loss": 0.04802687,
+      "step": 12721
+    },
+    {
+      "epoch": 25.444,
+      "grad_norm": 1.2358498573303223,
+      "learning_rate": 2e-05,
+      "loss": 0.03744403,
+      "step": 12722
+    },
+    {
+      "epoch": 25.446,
+      "grad_norm": 1.9896537065505981,
+      "learning_rate": 2e-05,
+      "loss": 0.0489389,
+      "step": 12723
+    },
+    {
+      "epoch": 25.448,
+      "grad_norm": 1.2103878259658813,
+      "learning_rate": 2e-05,
+      "loss": 0.04206013,
+      "step": 12724
+    },
+    {
+      "epoch": 25.45,
+      "grad_norm": 1.4859042167663574,
+      "learning_rate": 2e-05,
+      "loss": 0.04817963,
+      "step": 12725
+    },
+    {
+      "epoch": 25.452,
+      "grad_norm": 1.0707011222839355,
+      "learning_rate": 2e-05,
+      "loss": 0.02487176,
+      "step": 12726
+    },
+    {
+      "epoch": 25.454,
+      "grad_norm": 0.8648288249969482,
+      "learning_rate": 2e-05,
+      "loss": 0.02900369,
+      "step": 12727
+    },
+    {
+      "epoch": 25.456,
+      "grad_norm": 1.057659387588501,
+      "learning_rate": 2e-05,
+      "loss": 0.03823693,
+      "step": 12728
+    },
+    {
+      "epoch": 25.458,
+      "grad_norm": 1.477248191833496,
+      "learning_rate": 2e-05,
+      "loss": 0.05023545,
+      "step": 12729
+    },
+    {
+      "epoch": 25.46,
+      "grad_norm": 1.2823160886764526,
+      "learning_rate": 2e-05,
+      "loss": 0.05811198,
+      "step": 12730
+    },
+    {
+      "epoch": 25.462,
+      "grad_norm": 1.3287951946258545,
+      "learning_rate": 2e-05,
+      "loss": 0.05152711,
+      "step": 12731
+    },
+    {
+      "epoch": 25.464,
+      "grad_norm": 1.0755188465118408,
+      "learning_rate": 2e-05,
+      "loss": 0.04283263,
+      "step": 12732
+    },
+    {
+      "epoch": 25.466,
+      "grad_norm": 1.084681510925293,
+      "learning_rate": 2e-05,
+      "loss": 0.03596375,
+      "step": 12733
+    },
+    {
+      "epoch": 25.468,
+      "grad_norm": 1.1877176761627197,
+      "learning_rate": 2e-05,
+      "loss": 0.04653904,
+      "step": 12734
+    },
+    {
+      "epoch": 25.47,
+      "grad_norm": 1.7374459505081177,
+      "learning_rate": 2e-05,
+      "loss": 0.03502556,
+      "step": 12735
+    },
+    {
+      "epoch": 25.472,
+      "grad_norm": 1.5863116979599,
+      "learning_rate": 2e-05,
+      "loss": 0.03652181,
+      "step": 12736
+    },
+    {
+      "epoch": 25.474,
+      "grad_norm": 1.1849809885025024,
+      "learning_rate": 2e-05,
+      "loss": 0.0374795,
+      "step": 12737
+    },
+    {
+      "epoch": 25.476,
+      "grad_norm": 1.6925612688064575,
+      "learning_rate": 2e-05,
+      "loss": 0.05677975,
+      "step": 12738
+    },
+    {
+      "epoch": 25.478,
+      "grad_norm": 1.0018976926803589,
+      "learning_rate": 2e-05,
+      "loss": 0.035368,
+      "step": 12739
+    },
+    {
+      "epoch": 25.48,
+      "grad_norm": 1.0738310813903809,
+      "learning_rate": 2e-05,
+      "loss": 0.04165709,
+      "step": 12740
+    },
+    {
+      "epoch": 25.482,
+      "grad_norm": 1.0751115083694458,
+      "learning_rate": 2e-05,
+      "loss": 0.04024215,
+      "step": 12741
+    },
+    {
+      "epoch": 25.484,
+      "grad_norm": 1.2548881769180298,
+      "learning_rate": 2e-05,
+      "loss": 0.06063127,
+      "step": 12742
+    },
+    {
+      "epoch": 25.486,
+      "grad_norm": 2.8296592235565186,
+      "learning_rate": 2e-05,
+      "loss": 0.04534397,
+      "step": 12743
+    },
+    {
+      "epoch": 25.488,
+      "grad_norm": 2.0588326454162598,
+      "learning_rate": 2e-05,
+      "loss": 0.07510805,
+      "step": 12744
+    },
+    {
+      "epoch": 25.49,
+      "grad_norm": 2.919177293777466,
+      "learning_rate": 2e-05,
+      "loss": 0.05336414,
+      "step": 12745
+    },
+    {
+      "epoch": 25.492,
+      "grad_norm": 1.1437351703643799,
+      "learning_rate": 2e-05,
+      "loss": 0.04661782,
+      "step": 12746
+    },
+    {
+      "epoch": 25.494,
+      "grad_norm": 1.2010008096694946,
+      "learning_rate": 2e-05,
+      "loss": 0.04713929,
+      "step": 12747
+    },
+    {
+      "epoch": 25.496,
+      "grad_norm": 1.2396228313446045,
+      "learning_rate": 2e-05,
+      "loss": 0.03676428,
+      "step": 12748
+    },
+    {
+      "epoch": 25.498,
+      "grad_norm": 1.1913061141967773,
+      "learning_rate": 2e-05,
+      "loss": 0.03432823,
+      "step": 12749
+    },
+    {
+      "epoch": 25.5,
+      "grad_norm": 1.0427284240722656,
+      "learning_rate": 2e-05,
+      "loss": 0.04304377,
+      "step": 12750
+    },
+    {
+      "epoch": 25.502,
+      "grad_norm": 0.9250820279121399,
+      "learning_rate": 2e-05,
+      "loss": 0.02991773,
+      "step": 12751
+    },
+    {
+      "epoch": 25.504,
+      "grad_norm": 1.7029353380203247,
+      "learning_rate": 2e-05,
+      "loss": 0.05027733,
+      "step": 12752
+    },
+    {
+      "epoch": 25.506,
+      "grad_norm": 1.2598884105682373,
+      "learning_rate": 2e-05,
+      "loss": 0.03872483,
+      "step": 12753
+    },
+    {
+      "epoch": 25.508,
+      "grad_norm": 1.1312626600265503,
+      "learning_rate": 2e-05,
+      "loss": 0.03387032,
+      "step": 12754
+    },
+    {
+      "epoch": 25.51,
+      "grad_norm": 1.4454662799835205,
+      "learning_rate": 2e-05,
+      "loss": 0.04541782,
+      "step": 12755
+    },
+    {
+      "epoch": 25.512,
+      "grad_norm": 1.2431281805038452,
+      "learning_rate": 2e-05,
+      "loss": 0.04072841,
+      "step": 12756
+    },
+    {
+      "epoch": 25.514,
+      "grad_norm": 1.0991272926330566,
+      "learning_rate": 2e-05,
+      "loss": 0.03918246,
+      "step": 12757
+    },
+    {
+      "epoch": 25.516,
+      "grad_norm": 1.9460620880126953,
+      "learning_rate": 2e-05,
+      "loss": 0.06519485,
+      "step": 12758
+    },
+    {
+      "epoch": 25.518,
+      "grad_norm": 1.238317608833313,
+      "learning_rate": 2e-05,
+      "loss": 0.05222324,
+      "step": 12759
+    },
+    {
+      "epoch": 25.52,
+      "grad_norm": 1.1859232187271118,
+      "learning_rate": 2e-05,
+      "loss": 0.03913081,
+      "step": 12760
+    },
+    {
+      "epoch": 25.522,
+      "grad_norm": 1.1361500024795532,
+      "learning_rate": 2e-05,
+      "loss": 0.05126172,
+      "step": 12761
+    },
+    {
+      "epoch": 25.524,
+      "grad_norm": 1.657873511314392,
+      "learning_rate": 2e-05,
+      "loss": 0.06139502,
+      "step": 12762
+    },
+    {
+      "epoch": 25.526,
+      "grad_norm": 0.9964461326599121,
+      "learning_rate": 2e-05,
+      "loss": 0.03220826,
+      "step": 12763
+    },
+    {
+      "epoch": 25.528,
+      "grad_norm": 1.958337426185608,
+      "learning_rate": 2e-05,
+      "loss": 0.03383143,
+      "step": 12764
+    },
+    {
+      "epoch": 25.53,
+      "grad_norm": 1.3577609062194824,
+      "learning_rate": 2e-05,
+      "loss": 0.04257483,
+      "step": 12765
+    },
+    {
+      "epoch": 25.532,
+      "grad_norm": 1.1528269052505493,
+      "learning_rate": 2e-05,
+      "loss": 0.04727619,
+      "step": 12766
+    },
+    {
+      "epoch": 25.534,
+      "grad_norm": 1.3304978609085083,
+      "learning_rate": 2e-05,
+      "loss": 0.04373534,
+      "step": 12767
+    },
+    {
+      "epoch": 25.536,
+      "grad_norm": 2.185859441757202,
+      "learning_rate": 2e-05,
+      "loss": 0.06034337,
+      "step": 12768
+    },
+    {
+      "epoch": 25.538,
+      "grad_norm": 1.6654530763626099,
+      "learning_rate": 2e-05,
+      "loss": 0.05115644,
+      "step": 12769
+    },
+    {
+      "epoch": 25.54,
+      "grad_norm": 1.3865916728973389,
+      "learning_rate": 2e-05,
+      "loss": 0.05147178,
+      "step": 12770
+    },
+    {
+      "epoch": 25.542,
+      "grad_norm": 1.5472357273101807,
+      "learning_rate": 2e-05,
+      "loss": 0.04129698,
+      "step": 12771
+    },
+    {
+      "epoch": 25.544,
+      "grad_norm": 2.8680312633514404,
+      "learning_rate": 2e-05,
+      "loss": 0.05492282,
+      "step": 12772
+    },
+    {
+      "epoch": 25.546,
+      "grad_norm": 1.9399982690811157,
+      "learning_rate": 2e-05,
+      "loss": 0.04655978,
+      "step": 12773
+    },
+    {
+      "epoch": 25.548000000000002,
+      "grad_norm": 1.1271389722824097,
+      "learning_rate": 2e-05,
+      "loss": 0.04596433,
+      "step": 12774
+    },
+    {
+      "epoch": 25.55,
+      "grad_norm": 1.3790403604507446,
+      "learning_rate": 2e-05,
+      "loss": 0.04871649,
+      "step": 12775
+    },
+    {
+      "epoch": 25.552,
+      "grad_norm": 1.8597426414489746,
+      "learning_rate": 2e-05,
+      "loss": 0.04528391,
+      "step": 12776
+    },
+    {
+      "epoch": 25.554,
+      "grad_norm": 1.6001335382461548,
+      "learning_rate": 2e-05,
+      "loss": 0.04731301,
+      "step": 12777
+    },
+    {
+      "epoch": 25.556,
+      "grad_norm": 1.0347784757614136,
+      "learning_rate": 2e-05,
+      "loss": 0.03880426,
+      "step": 12778
+    },
+    {
+      "epoch": 25.558,
+      "grad_norm": 1.1842137575149536,
+      "learning_rate": 2e-05,
+      "loss": 0.04206136,
+      "step": 12779
+    },
+    {
+      "epoch": 25.56,
+      "grad_norm": 1.1110947132110596,
+      "learning_rate": 2e-05,
+      "loss": 0.02964817,
+      "step": 12780
+    },
+    {
+      "epoch": 25.562,
+      "grad_norm": 1.3468905687332153,
+      "learning_rate": 2e-05,
+      "loss": 0.05464962,
+      "step": 12781
+    },
+    {
+      "epoch": 25.564,
+      "grad_norm": 1.660844326019287,
+      "learning_rate": 2e-05,
+      "loss": 0.05474025,
+      "step": 12782
+    },
+    {
+      "epoch": 25.566,
+      "grad_norm": 0.958869993686676,
+      "learning_rate": 2e-05,
+      "loss": 0.03442763,
+      "step": 12783
+    },
+    {
+      "epoch": 25.568,
+      "grad_norm": 1.2681729793548584,
+      "learning_rate": 2e-05,
+      "loss": 0.05398714,
+      "step": 12784
+    },
+    {
+      "epoch": 25.57,
+      "grad_norm": 1.6511400938034058,
+      "learning_rate": 2e-05,
+      "loss": 0.04100592,
+      "step": 12785
+    },
+    {
+      "epoch": 25.572,
+      "grad_norm": 1.7444071769714355,
+      "learning_rate": 2e-05,
+      "loss": 0.05146802,
+      "step": 12786
+    },
+    {
+      "epoch": 25.574,
+      "grad_norm": 1.5246992111206055,
+      "learning_rate": 2e-05,
+      "loss": 0.04002138,
+      "step": 12787
+    },
+    {
+      "epoch": 25.576,
+      "grad_norm": 1.2669793367385864,
+      "learning_rate": 2e-05,
+      "loss": 0.03664404,
+      "step": 12788
+    },
+    {
+      "epoch": 25.578,
+      "grad_norm": 1.2276067733764648,
+      "learning_rate": 2e-05,
+      "loss": 0.04909092,
+      "step": 12789
+    },
+    {
+      "epoch": 25.58,
+      "grad_norm": 1.8915936946868896,
+      "learning_rate": 2e-05,
+      "loss": 0.04786887,
+      "step": 12790
+    },
+    {
+      "epoch": 25.582,
+      "grad_norm": 1.526929259300232,
+      "learning_rate": 2e-05,
+      "loss": 0.0330558,
+      "step": 12791
+    },
+    {
+      "epoch": 25.584,
+      "grad_norm": 1.7936866283416748,
+      "learning_rate": 2e-05,
+      "loss": 0.04382351,
+      "step": 12792
+    },
+    {
+      "epoch": 25.586,
+      "grad_norm": 2.3560688495635986,
+      "learning_rate": 2e-05,
+      "loss": 0.03439542,
+      "step": 12793
+    },
+    {
+      "epoch": 25.588,
+      "grad_norm": 1.3485525846481323,
+      "learning_rate": 2e-05,
+      "loss": 0.03818742,
+      "step": 12794
+    },
+    {
+      "epoch": 25.59,
+      "grad_norm": 1.1782861948013306,
+      "learning_rate": 2e-05,
+      "loss": 0.03852155,
+      "step": 12795
+    },
+    {
+      "epoch": 25.592,
+      "grad_norm": 1.5356895923614502,
+      "learning_rate": 2e-05,
+      "loss": 0.05399141,
+      "step": 12796
+    },
+    {
+      "epoch": 25.594,
+      "grad_norm": 1.246084451675415,
+      "learning_rate": 2e-05,
+      "loss": 0.04823676,
+      "step": 12797
+    },
+    {
+      "epoch": 25.596,
+      "grad_norm": 1.3950222730636597,
+      "learning_rate": 2e-05,
+      "loss": 0.02780427,
+      "step": 12798
+    },
+    {
+      "epoch": 25.598,
+      "grad_norm": 3.066758632659912,
+      "learning_rate": 2e-05,
+      "loss": 0.05022492,
+      "step": 12799
+    },
+    {
+      "epoch": 25.6,
+      "grad_norm": 0.9001642465591431,
+      "learning_rate": 2e-05,
+      "loss": 0.02825335,
+      "step": 12800
+    },
+    {
+      "epoch": 25.602,
+      "grad_norm": 1.1426812410354614,
+      "learning_rate": 2e-05,
+      "loss": 0.02332602,
+      "step": 12801
+    },
+    {
+      "epoch": 25.604,
+      "grad_norm": 1.341883897781372,
+      "learning_rate": 2e-05,
+      "loss": 0.05295313,
+      "step": 12802
+    },
+    {
+      "epoch": 25.606,
+      "grad_norm": 1.4535542726516724,
+      "learning_rate": 2e-05,
+      "loss": 0.04531492,
+      "step": 12803
+    },
+    {
+      "epoch": 25.608,
+      "grad_norm": 1.222536563873291,
+      "learning_rate": 2e-05,
+      "loss": 0.03721589,
+      "step": 12804
+    },
+    {
+      "epoch": 25.61,
+      "grad_norm": 1.7234156131744385,
+      "learning_rate": 2e-05,
+      "loss": 0.05661245,
+      "step": 12805
+    },
+    {
+      "epoch": 25.612,
+      "grad_norm": 1.7372621297836304,
+      "learning_rate": 2e-05,
+      "loss": 0.03441635,
+      "step": 12806
+    },
+    {
+      "epoch": 25.614,
+      "grad_norm": 1.6260429620742798,
+      "learning_rate": 2e-05,
+      "loss": 0.0488297,
+      "step": 12807
+    },
+    {
+      "epoch": 25.616,
+      "grad_norm": 1.0251811742782593,
+      "learning_rate": 2e-05,
+      "loss": 0.03384798,
+      "step": 12808
+    },
+    {
+      "epoch": 25.618,
+      "grad_norm": 1.0292855501174927,
+      "learning_rate": 2e-05,
+      "loss": 0.04441176,
+      "step": 12809
+    },
+    {
+      "epoch": 25.62,
+      "grad_norm": 0.9246246218681335,
+      "learning_rate": 2e-05,
+      "loss": 0.0311376,
+      "step": 12810
+    },
+    {
+      "epoch": 25.622,
+      "grad_norm": 1.1222032308578491,
+      "learning_rate": 2e-05,
+      "loss": 0.03349977,
+      "step": 12811
+    },
+    {
+      "epoch": 25.624,
+      "grad_norm": 1.044348955154419,
+      "learning_rate": 2e-05,
+      "loss": 0.03562807,
+      "step": 12812
+    },
+    {
+      "epoch": 25.626,
+      "grad_norm": 1.314139723777771,
+      "learning_rate": 2e-05,
+      "loss": 0.048839,
+      "step": 12813
+    },
+    {
+      "epoch": 25.628,
+      "grad_norm": 1.1304693222045898,
+      "learning_rate": 2e-05,
+      "loss": 0.05354474,
+      "step": 12814
+    },
+    {
+      "epoch": 25.63,
+      "grad_norm": 1.2466613054275513,
+      "learning_rate": 2e-05,
+      "loss": 0.0536761,
+      "step": 12815
+    },
+    {
+      "epoch": 25.632,
+      "grad_norm": 1.0759629011154175,
+      "learning_rate": 2e-05,
+      "loss": 0.04039822,
+      "step": 12816
+    },
+    {
+      "epoch": 25.634,
+      "grad_norm": 1.232729196548462,
+      "learning_rate": 2e-05,
+      "loss": 0.04778222,
+      "step": 12817
+    },
+    {
+      "epoch": 25.636,
+      "grad_norm": 1.3825780153274536,
+      "learning_rate": 2e-05,
+      "loss": 0.0383257,
+      "step": 12818
+    },
+    {
+      "epoch": 25.638,
+      "grad_norm": 1.105268120765686,
+      "learning_rate": 2e-05,
+      "loss": 0.03146917,
+      "step": 12819
+    },
+    {
+      "epoch": 25.64,
+      "grad_norm": 3.4076695442199707,
+      "learning_rate": 2e-05,
+      "loss": 0.04861237,
+      "step": 12820
+    },
+    {
+      "epoch": 25.642,
+      "grad_norm": 1.1343084573745728,
+      "learning_rate": 2e-05,
+      "loss": 0.0362022,
+      "step": 12821
+    },
+    {
+      "epoch": 25.644,
+      "grad_norm": 1.5663788318634033,
+      "learning_rate": 2e-05,
+      "loss": 0.04831579,
+      "step": 12822
+    },
+    {
+      "epoch": 25.646,
+      "grad_norm": 1.799911618232727,
+      "learning_rate": 2e-05,
+      "loss": 0.05474582,
+      "step": 12823
+    },
+    {
+      "epoch": 25.648,
+      "grad_norm": 1.5941003561019897,
+      "learning_rate": 2e-05,
+      "loss": 0.04637301,
+      "step": 12824
+    },
+    {
+      "epoch": 25.65,
+      "grad_norm": 1.7795392274856567,
+      "learning_rate": 2e-05,
+      "loss": 0.0298002,
+      "step": 12825
+    },
+    {
+      "epoch": 25.652,
+      "grad_norm": 1.5927716493606567,
+      "learning_rate": 2e-05,
+      "loss": 0.0472445,
+      "step": 12826
+    },
+    {
+      "epoch": 25.654,
+      "grad_norm": 1.3922079801559448,
+      "learning_rate": 2e-05,
+      "loss": 0.04148681,
+      "step": 12827
+    },
+    {
+      "epoch": 25.656,
+      "grad_norm": 1.3033424615859985,
+      "learning_rate": 2e-05,
+      "loss": 0.03751112,
+      "step": 12828
+    },
+    {
+      "epoch": 25.658,
+      "grad_norm": 1.3551427125930786,
+      "learning_rate": 2e-05,
+      "loss": 0.04426542,
+      "step": 12829
+    },
+    {
+      "epoch": 25.66,
+      "grad_norm": 4.914473056793213,
+      "learning_rate": 2e-05,
+      "loss": 0.05875074,
+      "step": 12830
+    },
+    {
+      "epoch": 25.662,
+      "grad_norm": 1.2864055633544922,
+      "learning_rate": 2e-05,
+      "loss": 0.05747083,
+      "step": 12831
+    },
+    {
+      "epoch": 25.664,
+      "grad_norm": 1.245685338973999,
+      "learning_rate": 2e-05,
+      "loss": 0.04856413,
+      "step": 12832
+    },
+    {
+      "epoch": 25.666,
+      "grad_norm": 1.2700167894363403,
+      "learning_rate": 2e-05,
+      "loss": 0.04627567,
+      "step": 12833
+    },
+    {
+      "epoch": 25.668,
+      "grad_norm": 2.3879594802856445,
+      "learning_rate": 2e-05,
+      "loss": 0.04573566,
+      "step": 12834
+    },
+    {
+      "epoch": 25.67,
+      "grad_norm": 1.0853418111801147,
+      "learning_rate": 2e-05,
+      "loss": 0.03547914,
+      "step": 12835
+    },
+    {
+      "epoch": 25.672,
+      "grad_norm": 2.4202558994293213,
+      "learning_rate": 2e-05,
+      "loss": 0.05616173,
+      "step": 12836
+    },
+    {
+      "epoch": 25.674,
+      "grad_norm": 1.213026523590088,
+      "learning_rate": 2e-05,
+      "loss": 0.04792723,
+      "step": 12837
+    },
+    {
+      "epoch": 25.676,
+      "grad_norm": 1.4499080181121826,
+      "learning_rate": 2e-05,
+      "loss": 0.04635759,
+      "step": 12838
+    },
+    {
+      "epoch": 25.678,
+      "grad_norm": 1.0613125562667847,
+      "learning_rate": 2e-05,
+      "loss": 0.04478652,
+      "step": 12839
+    },
+    {
+      "epoch": 25.68,
+      "grad_norm": 0.9953880906105042,
+      "learning_rate": 2e-05,
+      "loss": 0.02956217,
+      "step": 12840
+    },
+    {
+      "epoch": 25.682,
+      "grad_norm": 2.4529874324798584,
+      "learning_rate": 2e-05,
+      "loss": 0.04293549,
+      "step": 12841
+    },
+    {
+      "epoch": 25.684,
+      "grad_norm": 1.058303952217102,
+      "learning_rate": 2e-05,
+      "loss": 0.03956451,
+      "step": 12842
+    },
+    {
+      "epoch": 25.686,
+      "grad_norm": 1.2368803024291992,
+      "learning_rate": 2e-05,
+      "loss": 0.05040938,
+      "step": 12843
+    },
+    {
+      "epoch": 25.688,
+      "grad_norm": 1.3737293481826782,
+      "learning_rate": 2e-05,
+      "loss": 0.05072392,
+      "step": 12844
+    },
+    {
+      "epoch": 25.69,
+      "grad_norm": 1.0648189783096313,
+      "learning_rate": 2e-05,
+      "loss": 0.03040619,
+      "step": 12845
+    },
+    {
+      "epoch": 25.692,
+      "grad_norm": 1.3716273307800293,
+      "learning_rate": 2e-05,
+      "loss": 0.03951315,
+      "step": 12846
+    },
+    {
+      "epoch": 25.694,
+      "grad_norm": 1.4560929536819458,
+      "learning_rate": 2e-05,
+      "loss": 0.0416421,
+      "step": 12847
+    },
+    {
+      "epoch": 25.696,
+      "grad_norm": 1.9304611682891846,
+      "learning_rate": 2e-05,
+      "loss": 0.0425549,
+      "step": 12848
+    },
+    {
+      "epoch": 25.698,
+      "grad_norm": 0.9284382462501526,
+      "learning_rate": 2e-05,
+      "loss": 0.03185133,
+      "step": 12849
+    },
+    {
+      "epoch": 25.7,
+      "grad_norm": 1.3706117868423462,
+      "learning_rate": 2e-05,
+      "loss": 0.04668346,
+      "step": 12850
+    },
+    {
+      "epoch": 25.701999999999998,
+      "grad_norm": 1.7373970746994019,
+      "learning_rate": 2e-05,
+      "loss": 0.03499283,
+      "step": 12851
+    },
+    {
+      "epoch": 25.704,
+      "grad_norm": 2.239835262298584,
+      "learning_rate": 2e-05,
+      "loss": 0.0479632,
+      "step": 12852
+    },
+    {
+      "epoch": 25.706,
+      "grad_norm": 1.1869913339614868,
+      "learning_rate": 2e-05,
+      "loss": 0.04254553,
+      "step": 12853
+    },
+    {
+      "epoch": 25.708,
+      "grad_norm": 1.1868373155593872,
+      "learning_rate": 2e-05,
+      "loss": 0.03624144,
+      "step": 12854
+    },
+    {
+      "epoch": 25.71,
+      "grad_norm": 1.2735049724578857,
+      "learning_rate": 2e-05,
+      "loss": 0.04621403,
+      "step": 12855
+    },
+    {
+      "epoch": 25.712,
+      "grad_norm": 2.1693756580352783,
+      "learning_rate": 2e-05,
+      "loss": 0.05169474,
+      "step": 12856
+    },
+    {
+      "epoch": 25.714,
+      "grad_norm": 1.0680606365203857,
+      "learning_rate": 2e-05,
+      "loss": 0.03689266,
+      "step": 12857
+    },
+    {
+      "epoch": 25.716,
+      "grad_norm": 1.085115671157837,
+      "learning_rate": 2e-05,
+      "loss": 0.04505826,
+      "step": 12858
+    },
+    {
+      "epoch": 25.718,
+      "grad_norm": 1.4041411876678467,
+      "learning_rate": 2e-05,
+      "loss": 0.05596658,
+      "step": 12859
+    },
+    {
+      "epoch": 25.72,
+      "grad_norm": 1.3453253507614136,
+      "learning_rate": 2e-05,
+      "loss": 0.0477687,
+      "step": 12860
+    },
+    {
+      "epoch": 25.722,
+      "grad_norm": 1.0043227672576904,
+      "learning_rate": 2e-05,
+      "loss": 0.03752232,
+      "step": 12861
+    },
+    {
+      "epoch": 25.724,
+      "grad_norm": 1.8186486959457397,
+      "learning_rate": 2e-05,
+      "loss": 0.04446795,
+      "step": 12862
+    },
+    {
+      "epoch": 25.726,
+      "grad_norm": 1.1549947261810303,
+      "learning_rate": 2e-05,
+      "loss": 0.03548076,
+      "step": 12863
+    },
+    {
+      "epoch": 25.728,
+      "grad_norm": 1.8598767518997192,
+      "learning_rate": 2e-05,
+      "loss": 0.05279416,
+      "step": 12864
+    },
+    {
+      "epoch": 25.73,
+      "grad_norm": 1.2044051885604858,
+      "learning_rate": 2e-05,
+      "loss": 0.04757049,
+      "step": 12865
+    },
+    {
+      "epoch": 25.732,
+      "grad_norm": 1.5500402450561523,
+      "learning_rate": 2e-05,
+      "loss": 0.04289635,
+      "step": 12866
+    },
+    {
+      "epoch": 25.734,
+      "grad_norm": 1.505210041999817,
+      "learning_rate": 2e-05,
+      "loss": 0.03871378,
+      "step": 12867
+    },
+    {
+      "epoch": 25.736,
+      "grad_norm": 1.5146814584732056,
+      "learning_rate": 2e-05,
+      "loss": 0.06822012,
+      "step": 12868
+    },
+    {
+      "epoch": 25.738,
+      "grad_norm": 1.9045121669769287,
+      "learning_rate": 2e-05,
+      "loss": 0.04598961,
+      "step": 12869
+    },
+    {
+      "epoch": 25.74,
+      "grad_norm": 1.5291447639465332,
+      "learning_rate": 2e-05,
+      "loss": 0.04768593,
+      "step": 12870
+    },
+    {
+      "epoch": 25.742,
+      "grad_norm": 1.6933094263076782,
+      "learning_rate": 2e-05,
+      "loss": 0.03961466,
+      "step": 12871
+    },
+    {
+      "epoch": 25.744,
+      "grad_norm": 1.7180720567703247,
+      "learning_rate": 2e-05,
+      "loss": 0.04465307,
+      "step": 12872
+    },
+    {
+      "epoch": 25.746,
+      "grad_norm": 1.5562777519226074,
+      "learning_rate": 2e-05,
+      "loss": 0.04760274,
+      "step": 12873
+    },
+    {
+      "epoch": 25.748,
+      "grad_norm": 1.8690643310546875,
+      "learning_rate": 2e-05,
+      "loss": 0.0456716,
+      "step": 12874
+    },
+    {
+      "epoch": 25.75,
+      "grad_norm": 1.6642539501190186,
+      "learning_rate": 2e-05,
+      "loss": 0.0361044,
+      "step": 12875
+    },
+    {
+      "epoch": 25.752,
+      "grad_norm": 1.0987035036087036,
+      "learning_rate": 2e-05,
+      "loss": 0.0358355,
+      "step": 12876
+    },
+    {
+      "epoch": 25.754,
+      "grad_norm": 1.276738166809082,
+      "learning_rate": 2e-05,
+      "loss": 0.04325083,
+      "step": 12877
+    },
+    {
+      "epoch": 25.756,
+      "grad_norm": 1.4052026271820068,
+      "learning_rate": 2e-05,
+      "loss": 0.05564051,
+      "step": 12878
+    },
+    {
+      "epoch": 25.758,
+      "grad_norm": 1.1691828966140747,
+      "learning_rate": 2e-05,
+      "loss": 0.0309645,
+      "step": 12879
+    },
+    {
+      "epoch": 25.76,
+      "grad_norm": 1.5935523509979248,
+      "learning_rate": 2e-05,
+      "loss": 0.04389683,
+      "step": 12880
+    },
+    {
+      "epoch": 25.762,
+      "grad_norm": 1.4796446561813354,
+      "learning_rate": 2e-05,
+      "loss": 0.05674142,
+      "step": 12881
+    },
+    {
+      "epoch": 25.764,
+      "grad_norm": 1.473631501197815,
+      "learning_rate": 2e-05,
+      "loss": 0.05766422,
+      "step": 12882
+    },
+    {
+      "epoch": 25.766,
+      "grad_norm": 1.003163456916809,
+      "learning_rate": 2e-05,
+      "loss": 0.03590136,
+      "step": 12883
+    },
+    {
+      "epoch": 25.768,
+      "grad_norm": 1.1867930889129639,
+      "learning_rate": 2e-05,
+      "loss": 0.04083221,
+      "step": 12884
+    },
+    {
+      "epoch": 25.77,
+      "grad_norm": 1.342621922492981,
+      "learning_rate": 2e-05,
+      "loss": 0.04845904,
+      "step": 12885
+    },
+    {
+      "epoch": 25.772,
+      "grad_norm": 1.7419025897979736,
+      "learning_rate": 2e-05,
+      "loss": 0.04856285,
+      "step": 12886
+    },
+    {
+      "epoch": 25.774,
+      "grad_norm": 0.9666280150413513,
+      "learning_rate": 2e-05,
+      "loss": 0.03569352,
+      "step": 12887
+    },
+    {
+      "epoch": 25.776,
+      "grad_norm": 1.4296828508377075,
+      "learning_rate": 2e-05,
+      "loss": 0.03583002,
+      "step": 12888
+    },
+    {
+      "epoch": 25.778,
+      "grad_norm": 1.1190946102142334,
+      "learning_rate": 2e-05,
+      "loss": 0.02983772,
+      "step": 12889
+    },
+    {
+      "epoch": 25.78,
+      "grad_norm": 1.2515982389450073,
+      "learning_rate": 2e-05,
+      "loss": 0.04836629,
+      "step": 12890
+    },
+    {
+      "epoch": 25.782,
+      "grad_norm": 2.553123712539673,
+      "learning_rate": 2e-05,
+      "loss": 0.05472127,
+      "step": 12891
+    },
+    {
+      "epoch": 25.784,
+      "grad_norm": 1.407144546508789,
+      "learning_rate": 2e-05,
+      "loss": 0.04308041,
+      "step": 12892
+    },
+    {
+      "epoch": 25.786,
+      "grad_norm": 1.1658834218978882,
+      "learning_rate": 2e-05,
+      "loss": 0.03924808,
+      "step": 12893
+    },
+    {
+      "epoch": 25.788,
+      "grad_norm": 1.917145848274231,
+      "learning_rate": 2e-05,
+      "loss": 0.06007392,
+      "step": 12894
+    },
+    {
+      "epoch": 25.79,
+      "grad_norm": 2.5494346618652344,
+      "learning_rate": 2e-05,
+      "loss": 0.04623754,
+      "step": 12895
+    },
+    {
+      "epoch": 25.792,
+      "grad_norm": 1.3239128589630127,
+      "learning_rate": 2e-05,
+      "loss": 0.04573008,
+      "step": 12896
+    },
+    {
+      "epoch": 25.794,
+      "grad_norm": 1.2541325092315674,
+      "learning_rate": 2e-05,
+      "loss": 0.03188618,
+      "step": 12897
+    },
+    {
+      "epoch": 25.796,
+      "grad_norm": 1.2667912244796753,
+      "learning_rate": 2e-05,
+      "loss": 0.04757137,
+      "step": 12898
+    },
+    {
+      "epoch": 25.798000000000002,
+      "grad_norm": 1.1675081253051758,
+      "learning_rate": 2e-05,
+      "loss": 0.03582035,
+      "step": 12899
+    },
+    {
+      "epoch": 25.8,
+      "grad_norm": 1.2599587440490723,
+      "learning_rate": 2e-05,
+      "loss": 0.05033941,
+      "step": 12900
+    },
+    {
+      "epoch": 25.802,
+      "grad_norm": 1.4526861906051636,
+      "learning_rate": 2e-05,
+      "loss": 0.02551962,
+      "step": 12901
+    },
+    {
+      "epoch": 25.804,
+      "grad_norm": 2.03424334526062,
+      "learning_rate": 2e-05,
+      "loss": 0.04258246,
+      "step": 12902
+    },
+    {
+      "epoch": 25.806,
+      "grad_norm": 1.5461796522140503,
+      "learning_rate": 2e-05,
+      "loss": 0.05401737,
+      "step": 12903
+    },
+    {
+      "epoch": 25.808,
+      "grad_norm": 1.3652410507202148,
+      "learning_rate": 2e-05,
+      "loss": 0.04861759,
+      "step": 12904
+    },
+    {
+      "epoch": 25.81,
+      "grad_norm": 2.5918078422546387,
+      "learning_rate": 2e-05,
+      "loss": 0.04921463,
+      "step": 12905
+    },
+    {
+      "epoch": 25.812,
+      "grad_norm": 1.6058486700057983,
+      "learning_rate": 2e-05,
+      "loss": 0.03778245,
+      "step": 12906
+    },
+    {
+      "epoch": 25.814,
+      "grad_norm": 1.3766907453536987,
+      "learning_rate": 2e-05,
+      "loss": 0.03459726,
+      "step": 12907
+    },
+    {
+      "epoch": 25.816,
+      "grad_norm": 1.255175232887268,
+      "learning_rate": 2e-05,
+      "loss": 0.04270238,
+      "step": 12908
+    },
+    {
+      "epoch": 25.818,
+      "grad_norm": 1.1050204038619995,
+      "learning_rate": 2e-05,
+      "loss": 0.03457598,
+      "step": 12909
+    },
+    {
+      "epoch": 25.82,
+      "grad_norm": 1.7525054216384888,
+      "learning_rate": 2e-05,
+      "loss": 0.04051763,
+      "step": 12910
+    },
+    {
+      "epoch": 25.822,
+      "grad_norm": 3.4903576374053955,
+      "learning_rate": 2e-05,
+      "loss": 0.04965971,
+      "step": 12911
+    },
+    {
+      "epoch": 25.824,
+      "grad_norm": 1.1599464416503906,
+      "learning_rate": 2e-05,
+      "loss": 0.03711332,
+      "step": 12912
+    },
+    {
+      "epoch": 25.826,
+      "grad_norm": 3.031643867492676,
+      "learning_rate": 2e-05,
+      "loss": 0.05846217,
+      "step": 12913
+    },
+    {
+      "epoch": 25.828,
+      "grad_norm": 1.1548397541046143,
+      "learning_rate": 2e-05,
+      "loss": 0.03609941,
+      "step": 12914
+    },
+    {
+      "epoch": 25.83,
+      "grad_norm": 6.252331733703613,
+      "learning_rate": 2e-05,
+      "loss": 0.04764732,
+      "step": 12915
+    },
+    {
+      "epoch": 25.832,
+      "grad_norm": 1.6452500820159912,
+      "learning_rate": 2e-05,
+      "loss": 0.05718385,
+      "step": 12916
+    },
+    {
+      "epoch": 25.834,
+      "grad_norm": 1.1862694025039673,
+      "learning_rate": 2e-05,
+      "loss": 0.06088445,
+      "step": 12917
+    },
+    {
+      "epoch": 25.836,
+      "grad_norm": 1.5176570415496826,
+      "learning_rate": 2e-05,
+      "loss": 0.03762556,
+      "step": 12918
+    },
+    {
+      "epoch": 25.838,
+      "grad_norm": 1.6177486181259155,
+      "learning_rate": 2e-05,
+      "loss": 0.05010308,
+      "step": 12919
+    },
+    {
+      "epoch": 25.84,
+      "grad_norm": 3.6465678215026855,
+      "learning_rate": 2e-05,
+      "loss": 0.04568088,
+      "step": 12920
+    },
+    {
+      "epoch": 25.842,
+      "grad_norm": 1.225893497467041,
+      "learning_rate": 2e-05,
+      "loss": 0.04407287,
+      "step": 12921
+    },
+    {
+      "epoch": 25.844,
+      "grad_norm": 1.419485092163086,
+      "learning_rate": 2e-05,
+      "loss": 0.05671504,
+      "step": 12922
+    },
+    {
+      "epoch": 25.846,
+      "grad_norm": 0.9688834547996521,
+      "learning_rate": 2e-05,
+      "loss": 0.0258717,
+      "step": 12923
+    },
+    {
+      "epoch": 25.848,
+      "grad_norm": 1.4948381185531616,
+      "learning_rate": 2e-05,
+      "loss": 0.05070211,
+      "step": 12924
+    },
+    {
+      "epoch": 25.85,
+      "grad_norm": 1.48432195186615,
+      "learning_rate": 2e-05,
+      "loss": 0.03078519,
+      "step": 12925
+    },
+    {
+      "epoch": 25.852,
+      "grad_norm": 1.1909409761428833,
+      "learning_rate": 2e-05,
+      "loss": 0.04312345,
+      "step": 12926
+    },
+    {
+      "epoch": 25.854,
+      "grad_norm": 1.25083327293396,
+      "learning_rate": 2e-05,
+      "loss": 0.04205637,
+      "step": 12927
+    },
+    {
+      "epoch": 25.856,
+      "grad_norm": 1.1452685594558716,
+      "learning_rate": 2e-05,
+      "loss": 0.04423122,
+      "step": 12928
+    },
+    {
+      "epoch": 25.858,
+      "grad_norm": 1.53263258934021,
+      "learning_rate": 2e-05,
+      "loss": 0.04331882,
+      "step": 12929
+    },
+    {
+      "epoch": 25.86,
+      "grad_norm": 1.451874852180481,
+      "learning_rate": 2e-05,
+      "loss": 0.04954249,
+      "step": 12930
+    },
+    {
+      "epoch": 25.862,
+      "grad_norm": 1.3785700798034668,
+      "learning_rate": 2e-05,
+      "loss": 0.04673723,
+      "step": 12931
+    },
+    {
+      "epoch": 25.864,
+      "grad_norm": 1.375231146812439,
+      "learning_rate": 2e-05,
+      "loss": 0.0459612,
+      "step": 12932
+    },
+    {
+      "epoch": 25.866,
+      "grad_norm": 1.2162593603134155,
+      "learning_rate": 2e-05,
+      "loss": 0.04416041,
+      "step": 12933
+    },
+    {
+      "epoch": 25.868,
+      "grad_norm": 1.1460282802581787,
+      "learning_rate": 2e-05,
+      "loss": 0.0397279,
+      "step": 12934
+    },
+    {
+      "epoch": 25.87,
+      "grad_norm": 1.4855611324310303,
+      "learning_rate": 2e-05,
+      "loss": 0.04811959,
+      "step": 12935
+    },
+    {
+      "epoch": 25.872,
+      "grad_norm": 1.6014434099197388,
+      "learning_rate": 2e-05,
+      "loss": 0.05218223,
+      "step": 12936
+    },
+    {
+      "epoch": 25.874,
+      "grad_norm": 1.3930706977844238,
+      "learning_rate": 2e-05,
+      "loss": 0.05569389,
+      "step": 12937
+    },
+    {
+      "epoch": 25.876,
+      "grad_norm": 1.5376465320587158,
+      "learning_rate": 2e-05,
+      "loss": 0.05116809,
+      "step": 12938
+    },
+    {
+      "epoch": 25.878,
+      "grad_norm": 1.3046684265136719,
+      "learning_rate": 2e-05,
+      "loss": 0.05488844,
+      "step": 12939
+    },
+    {
+      "epoch": 25.88,
+      "grad_norm": 1.029398798942566,
+      "learning_rate": 2e-05,
+      "loss": 0.03596478,
+      "step": 12940
+    },
+    {
+      "epoch": 25.882,
+      "grad_norm": 1.2957504987716675,
+      "learning_rate": 2e-05,
+      "loss": 0.03197503,
+      "step": 12941
+    },
+    {
+      "epoch": 25.884,
+      "grad_norm": 1.1034080982208252,
+      "learning_rate": 2e-05,
+      "loss": 0.04575316,
+      "step": 12942
+    },
+    {
+      "epoch": 25.886,
+      "grad_norm": 1.4886722564697266,
+      "learning_rate": 2e-05,
+      "loss": 0.04241584,
+      "step": 12943
+    },
+    {
+      "epoch": 25.888,
+      "grad_norm": 1.763895869255066,
+      "learning_rate": 2e-05,
+      "loss": 0.04630455,
+      "step": 12944
+    },
+    {
+      "epoch": 25.89,
+      "grad_norm": 1.2453800439834595,
+      "learning_rate": 2e-05,
+      "loss": 0.04511211,
+      "step": 12945
+    },
+    {
+      "epoch": 25.892,
+      "grad_norm": 1.3455971479415894,
+      "learning_rate": 2e-05,
+      "loss": 0.03952543,
+      "step": 12946
+    },
+    {
+      "epoch": 25.894,
+      "grad_norm": 1.4724493026733398,
+      "learning_rate": 2e-05,
+      "loss": 0.06532818,
+      "step": 12947
+    },
+    {
+      "epoch": 25.896,
+      "grad_norm": 1.60633385181427,
+      "learning_rate": 2e-05,
+      "loss": 0.06741303,
+      "step": 12948
+    },
+    {
+      "epoch": 25.898,
+      "grad_norm": 1.3577247858047485,
+      "learning_rate": 2e-05,
+      "loss": 0.04642791,
+      "step": 12949
+    },
+    {
+      "epoch": 25.9,
+      "grad_norm": 0.9978728890419006,
+      "learning_rate": 2e-05,
+      "loss": 0.0259608,
+      "step": 12950
+    },
+    {
+      "epoch": 25.902,
+      "grad_norm": 1.5392905473709106,
+      "learning_rate": 2e-05,
+      "loss": 0.03973429,
+      "step": 12951
+    },
+    {
+      "epoch": 25.904,
+      "grad_norm": 1.0129889249801636,
+      "learning_rate": 2e-05,
+      "loss": 0.02280146,
+      "step": 12952
+    },
+    {
+      "epoch": 25.906,
+      "grad_norm": 2.133425712585449,
+      "learning_rate": 2e-05,
+      "loss": 0.04928362,
+      "step": 12953
+    },
+    {
+      "epoch": 25.908,
+      "grad_norm": 1.3658127784729004,
+      "learning_rate": 2e-05,
+      "loss": 0.0481901,
+      "step": 12954
+    },
+    {
+      "epoch": 25.91,
+      "grad_norm": 2.754051446914673,
+      "learning_rate": 2e-05,
+      "loss": 0.05194226,
+      "step": 12955
+    },
+    {
+      "epoch": 25.912,
+      "grad_norm": 1.568504810333252,
+      "learning_rate": 2e-05,
+      "loss": 0.04438355,
+      "step": 12956
+    },
+    {
+      "epoch": 25.914,
+      "grad_norm": 1.2474275827407837,
+      "learning_rate": 2e-05,
+      "loss": 0.0374986,
+      "step": 12957
+    },
+    {
+      "epoch": 25.916,
+      "grad_norm": 1.8854233026504517,
+      "learning_rate": 2e-05,
+      "loss": 0.05303825,
+      "step": 12958
+    },
+    {
+      "epoch": 25.918,
+      "grad_norm": 1.0603793859481812,
+      "learning_rate": 2e-05,
+      "loss": 0.04348066,
+      "step": 12959
+    },
+    {
+      "epoch": 25.92,
+      "grad_norm": 1.2751173973083496,
+      "learning_rate": 2e-05,
+      "loss": 0.04942772,
+      "step": 12960
+    },
+    {
+      "epoch": 25.922,
+      "grad_norm": 1.5532885789871216,
+      "learning_rate": 2e-05,
+      "loss": 0.04615701,
+      "step": 12961
+    },
+    {
+      "epoch": 25.924,
+      "grad_norm": 1.4249638319015503,
+      "learning_rate": 2e-05,
+      "loss": 0.04614986,
+      "step": 12962
+    },
+    {
+      "epoch": 25.926,
+      "grad_norm": 1.7213596105575562,
+      "learning_rate": 2e-05,
+      "loss": 0.04523218,
+      "step": 12963
+    },
+    {
+      "epoch": 25.928,
+      "grad_norm": 1.564770221710205,
+      "learning_rate": 2e-05,
+      "loss": 0.06634088,
+      "step": 12964
+    },
+    {
+      "epoch": 25.93,
+      "grad_norm": 2.3734843730926514,
+      "learning_rate": 2e-05,
+      "loss": 0.06825857,
+      "step": 12965
+    },
+    {
+      "epoch": 25.932,
+      "grad_norm": 0.9273267388343811,
+      "learning_rate": 2e-05,
+      "loss": 0.02456025,
+      "step": 12966
+    },
+    {
+      "epoch": 25.934,
+      "grad_norm": 2.232722759246826,
+      "learning_rate": 2e-05,
+      "loss": 0.05355635,
+      "step": 12967
+    },
+    {
+      "epoch": 25.936,
+      "grad_norm": 3.6321330070495605,
+      "learning_rate": 2e-05,
+      "loss": 0.05499596,
+      "step": 12968
+    },
+    {
+      "epoch": 25.938,
+      "grad_norm": 1.379671335220337,
+      "learning_rate": 2e-05,
+      "loss": 0.05246424,
+      "step": 12969
+    },
+    {
+      "epoch": 25.94,
+      "grad_norm": 1.2768114805221558,
+      "learning_rate": 2e-05,
+      "loss": 0.03318752,
+      "step": 12970
+    },
+    {
+      "epoch": 25.942,
+      "grad_norm": 1.4348204135894775,
+      "learning_rate": 2e-05,
+      "loss": 0.04715006,
+      "step": 12971
+    },
+    {
+      "epoch": 25.944,
+      "grad_norm": 1.4383244514465332,
+      "learning_rate": 2e-05,
+      "loss": 0.05146648,
+      "step": 12972
+    },
+    {
+      "epoch": 25.946,
+      "grad_norm": 2.579630136489868,
+      "learning_rate": 2e-05,
+      "loss": 0.05426867,
+      "step": 12973
+    },
+    {
+      "epoch": 25.948,
+      "grad_norm": 1.199095606803894,
+      "learning_rate": 2e-05,
+      "loss": 0.04479141,
+      "step": 12974
+    },
+    {
+      "epoch": 25.95,
+      "grad_norm": 1.3972845077514648,
+      "learning_rate": 2e-05,
+      "loss": 0.0324418,
+      "step": 12975
+    },
+    {
+      "epoch": 25.951999999999998,
+      "grad_norm": 1.0197676420211792,
+      "learning_rate": 2e-05,
+      "loss": 0.03971727,
+      "step": 12976
+    },
+    {
+      "epoch": 25.954,
+      "grad_norm": 1.1759026050567627,
+      "learning_rate": 2e-05,
+      "loss": 0.03519825,
+      "step": 12977
+    },
+    {
+      "epoch": 25.956,
+      "grad_norm": 1.27419912815094,
+      "learning_rate": 2e-05,
+      "loss": 0.04272161,
+      "step": 12978
+    },
+    {
+      "epoch": 25.958,
+      "grad_norm": 1.3605846166610718,
+      "learning_rate": 2e-05,
+      "loss": 0.05668283,
+      "step": 12979
+    },
+    {
+      "epoch": 25.96,
+      "grad_norm": 2.3534092903137207,
+      "learning_rate": 2e-05,
+      "loss": 0.04117073,
+      "step": 12980
+    },
+    {
+      "epoch": 25.962,
+      "grad_norm": 1.1537200212478638,
+      "learning_rate": 2e-05,
+      "loss": 0.06345319,
+      "step": 12981
+    },
+    {
+      "epoch": 25.964,
+      "grad_norm": 2.1213245391845703,
+      "learning_rate": 2e-05,
+      "loss": 0.04125203,
+      "step": 12982
+    },
+    {
+      "epoch": 25.966,
+      "grad_norm": 1.309527039527893,
+      "learning_rate": 2e-05,
+      "loss": 0.04417682,
+      "step": 12983
+    },
+    {
+      "epoch": 25.968,
+      "grad_norm": 1.206042766571045,
+      "learning_rate": 2e-05,
+      "loss": 0.04790517,
+      "step": 12984
+    },
+    {
+      "epoch": 25.97,
+      "grad_norm": 1.2381349802017212,
+      "learning_rate": 2e-05,
+      "loss": 0.03844728,
+      "step": 12985
+    },
+    {
+      "epoch": 25.972,
+      "grad_norm": 1.0183463096618652,
+      "learning_rate": 2e-05,
+      "loss": 0.04048822,
+      "step": 12986
+    },
+    {
+      "epoch": 25.974,
+      "grad_norm": 1.5077916383743286,
+      "learning_rate": 2e-05,
+      "loss": 0.03975806,
+      "step": 12987
+    },
+    {
+      "epoch": 25.976,
+      "grad_norm": 1.284839391708374,
+      "learning_rate": 2e-05,
+      "loss": 0.03966205,
+      "step": 12988
+    },
+    {
+      "epoch": 25.978,
+      "grad_norm": 2.6741936206817627,
+      "learning_rate": 2e-05,
+      "loss": 0.06823039,
+      "step": 12989
+    },
+    {
+      "epoch": 25.98,
+      "grad_norm": 1.2821533679962158,
+      "learning_rate": 2e-05,
+      "loss": 0.02755376,
+      "step": 12990
+    },
+    {
+      "epoch": 25.982,
+      "grad_norm": 1.1977559328079224,
+      "learning_rate": 2e-05,
+      "loss": 0.05067346,
+      "step": 12991
+    },
+    {
+      "epoch": 25.984,
+      "grad_norm": 1.267210602760315,
+      "learning_rate": 2e-05,
+      "loss": 0.05051599,
+      "step": 12992
+    },
+    {
+      "epoch": 25.986,
+      "grad_norm": 1.6138875484466553,
+      "learning_rate": 2e-05,
+      "loss": 0.05310912,
+      "step": 12993
+    },
+    {
+      "epoch": 25.988,
+      "grad_norm": 1.2645289897918701,
+      "learning_rate": 2e-05,
+      "loss": 0.05147942,
+      "step": 12994
+    },
+    {
+      "epoch": 25.99,
+      "grad_norm": 1.5932942628860474,
+      "learning_rate": 2e-05,
+      "loss": 0.02298946,
+      "step": 12995
+    },
+    {
+      "epoch": 25.992,
+      "grad_norm": 1.1337193250656128,
+      "learning_rate": 2e-05,
+      "loss": 0.02549054,
+      "step": 12996
+    },
+    {
+      "epoch": 25.994,
+      "grad_norm": 1.2463297843933105,
+      "learning_rate": 2e-05,
+      "loss": 0.04182723,
+      "step": 12997
+    },
+    {
+      "epoch": 25.996,
+      "grad_norm": 2.0521812438964844,
+      "learning_rate": 2e-05,
+      "loss": 0.04027394,
+      "step": 12998
+    },
+    {
+      "epoch": 25.998,
+      "grad_norm": 1.0944292545318604,
+      "learning_rate": 2e-05,
+      "loss": 0.05282188,
+      "step": 12999
+    },
+    {
+      "epoch": 26.0,
+      "grad_norm": 0.9855789542198181,
+      "learning_rate": 2e-05,
+      "loss": 0.03689377,
+      "step": 13000
+    },
+    {
+      "epoch": 26.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9780439121756487,
+        "Equal_1": 0.996,
+        "Equal_2": 0.9840319361277445,
+        "Equal_3": 0.9800399201596807,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9960079840319361,
+        "Parallel_1": 0.9879759519038076,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.992,
+        "Perpendicular_1": 0.99,
+        "Perpendicular_2": 0.982,
+        "Perpendicular_3": 0.8717434869739479,
+        "PointLiesOnCircle_1": 0.9996659986639947,
+        "PointLiesOnCircle_2": 0.9996666666666667,
+        "PointLiesOnCircle_3": 0.9936,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9700598802395209
+      },
+      "eval_runtime": 319.7653,
+      "eval_samples_per_second": 32.837,
+      "eval_steps_per_second": 0.657,
+      "step": 13000
+    },
+    {
+      "epoch": 26.002,
+      "grad_norm": 1.2228409051895142,
+      "learning_rate": 2e-05,
+      "loss": 0.03640506,
+      "step": 13001
+    },
+    {
+      "epoch": 26.004,
+      "grad_norm": 2.15970778465271,
+      "learning_rate": 2e-05,
+      "loss": 0.05572034,
+      "step": 13002
+    },
+    {
+      "epoch": 26.006,
+      "grad_norm": 1.3950461149215698,
+      "learning_rate": 2e-05,
+      "loss": 0.04152773,
+      "step": 13003
+    },
+    {
+      "epoch": 26.008,
+      "grad_norm": 1.6164307594299316,
+      "learning_rate": 2e-05,
+      "loss": 0.04839995,
+      "step": 13004
+    },
+    {
+      "epoch": 26.01,
+      "grad_norm": 1.7224260568618774,
+      "learning_rate": 2e-05,
+      "loss": 0.04432855,
+      "step": 13005
+    },
+    {
+      "epoch": 26.012,
+      "grad_norm": 1.3991963863372803,
+      "learning_rate": 2e-05,
+      "loss": 0.05045583,
+      "step": 13006
+    },
+    {
+      "epoch": 26.014,
+      "grad_norm": 1.8230713605880737,
+      "learning_rate": 2e-05,
+      "loss": 0.06015655,
+      "step": 13007
+    },
+    {
+      "epoch": 26.016,
+      "grad_norm": 1.7791680097579956,
+      "learning_rate": 2e-05,
+      "loss": 0.0451679,
+      "step": 13008
+    },
+    {
+      "epoch": 26.018,
+      "grad_norm": 1.7463334798812866,
+      "learning_rate": 2e-05,
+      "loss": 0.05355278,
+      "step": 13009
+    },
+    {
+      "epoch": 26.02,
+      "grad_norm": 1.9259138107299805,
+      "learning_rate": 2e-05,
+      "loss": 0.05147279,
+      "step": 13010
+    },
+    {
+      "epoch": 26.022,
+      "grad_norm": 1.498757243156433,
+      "learning_rate": 2e-05,
+      "loss": 0.05272644,
+      "step": 13011
+    },
+    {
+      "epoch": 26.024,
+      "grad_norm": 2.2757580280303955,
+      "learning_rate": 2e-05,
+      "loss": 0.0499465,
+      "step": 13012
+    },
+    {
+      "epoch": 26.026,
+      "grad_norm": 2.213714122772217,
+      "learning_rate": 2e-05,
+      "loss": 0.04541406,
+      "step": 13013
+    },
+    {
+      "epoch": 26.028,
+      "grad_norm": 1.893241286277771,
+      "learning_rate": 2e-05,
+      "loss": 0.05529698,
+      "step": 13014
+    },
+    {
+      "epoch": 26.03,
+      "grad_norm": 1.8718706369400024,
+      "learning_rate": 2e-05,
+      "loss": 0.04376329,
+      "step": 13015
+    },
+    {
+      "epoch": 26.032,
+      "grad_norm": 1.3967310190200806,
+      "learning_rate": 2e-05,
+      "loss": 0.03513469,
+      "step": 13016
+    },
+    {
+      "epoch": 26.034,
+      "grad_norm": 1.8421751260757446,
+      "learning_rate": 2e-05,
+      "loss": 0.04883641,
+      "step": 13017
+    },
+    {
+      "epoch": 26.036,
+      "grad_norm": 1.9495335817337036,
+      "learning_rate": 2e-05,
+      "loss": 0.03737961,
+      "step": 13018
+    },
+    {
+      "epoch": 26.038,
+      "grad_norm": 1.0064893960952759,
+      "learning_rate": 2e-05,
+      "loss": 0.04025814,
+      "step": 13019
+    },
+    {
+      "epoch": 26.04,
+      "grad_norm": 1.1454533338546753,
+      "learning_rate": 2e-05,
+      "loss": 0.05090432,
+      "step": 13020
+    },
+    {
+      "epoch": 26.042,
+      "grad_norm": 1.2419127225875854,
+      "learning_rate": 2e-05,
+      "loss": 0.04127611,
+      "step": 13021
+    },
+    {
+      "epoch": 26.044,
+      "grad_norm": 1.4765992164611816,
+      "learning_rate": 2e-05,
+      "loss": 0.053638,
+      "step": 13022
+    },
+    {
+      "epoch": 26.046,
+      "grad_norm": 1.3810056447982788,
+      "learning_rate": 2e-05,
+      "loss": 0.02833214,
+      "step": 13023
+    },
+    {
+      "epoch": 26.048,
+      "grad_norm": 1.1815146207809448,
+      "learning_rate": 2e-05,
+      "loss": 0.0438899,
+      "step": 13024
+    },
+    {
+      "epoch": 26.05,
+      "grad_norm": 1.4837486743927002,
+      "learning_rate": 2e-05,
+      "loss": 0.04868677,
+      "step": 13025
+    },
+    {
+      "epoch": 26.052,
+      "grad_norm": 1.6083544492721558,
+      "learning_rate": 2e-05,
+      "loss": 0.0421802,
+      "step": 13026
+    },
+    {
+      "epoch": 26.054,
+      "grad_norm": 1.2895910739898682,
+      "learning_rate": 2e-05,
+      "loss": 0.04041936,
+      "step": 13027
+    },
+    {
+      "epoch": 26.056,
+      "grad_norm": 1.0532166957855225,
+      "learning_rate": 2e-05,
+      "loss": 0.03752539,
+      "step": 13028
+    },
+    {
+      "epoch": 26.058,
+      "grad_norm": 1.7177013158798218,
+      "learning_rate": 2e-05,
+      "loss": 0.04090296,
+      "step": 13029
+    },
+    {
+      "epoch": 26.06,
+      "grad_norm": 1.4352203607559204,
+      "learning_rate": 2e-05,
+      "loss": 0.06325287,
+      "step": 13030
+    },
+    {
+      "epoch": 26.062,
+      "grad_norm": 1.2944906949996948,
+      "learning_rate": 2e-05,
+      "loss": 0.03913593,
+      "step": 13031
+    },
+    {
+      "epoch": 26.064,
+      "grad_norm": 1.1036475896835327,
+      "learning_rate": 2e-05,
+      "loss": 0.03017885,
+      "step": 13032
+    },
+    {
+      "epoch": 26.066,
+      "grad_norm": 1.4907349348068237,
+      "learning_rate": 2e-05,
+      "loss": 0.06610879,
+      "step": 13033
+    },
+    {
+      "epoch": 26.068,
+      "grad_norm": 2.620415210723877,
+      "learning_rate": 2e-05,
+      "loss": 0.06170364,
+      "step": 13034
+    },
+    {
+      "epoch": 26.07,
+      "grad_norm": 1.2906463146209717,
+      "learning_rate": 2e-05,
+      "loss": 0.05566068,
+      "step": 13035
+    },
+    {
+      "epoch": 26.072,
+      "grad_norm": 1.5933210849761963,
+      "learning_rate": 2e-05,
+      "loss": 0.06452239,
+      "step": 13036
+    },
+    {
+      "epoch": 26.074,
+      "grad_norm": 1.2673341035842896,
+      "learning_rate": 2e-05,
+      "loss": 0.04142977,
+      "step": 13037
+    },
+    {
+      "epoch": 26.076,
+      "grad_norm": 1.4276865720748901,
+      "learning_rate": 2e-05,
+      "loss": 0.04855426,
+      "step": 13038
+    },
+    {
+      "epoch": 26.078,
+      "grad_norm": 1.5951037406921387,
+      "learning_rate": 2e-05,
+      "loss": 0.04814789,
+      "step": 13039
+    },
+    {
+      "epoch": 26.08,
+      "grad_norm": 1.3216607570648193,
+      "learning_rate": 2e-05,
+      "loss": 0.04974344,
+      "step": 13040
+    },
+    {
+      "epoch": 26.082,
+      "grad_norm": 1.4452524185180664,
+      "learning_rate": 2e-05,
+      "loss": 0.04898542,
+      "step": 13041
+    },
+    {
+      "epoch": 26.084,
+      "grad_norm": 1.5695279836654663,
+      "learning_rate": 2e-05,
+      "loss": 0.05378542,
+      "step": 13042
+    },
+    {
+      "epoch": 26.086,
+      "grad_norm": 2.6487820148468018,
+      "learning_rate": 2e-05,
+      "loss": 0.05485958,
+      "step": 13043
+    },
+    {
+      "epoch": 26.088,
+      "grad_norm": 1.1026345491409302,
+      "learning_rate": 2e-05,
+      "loss": 0.03617822,
+      "step": 13044
+    },
+    {
+      "epoch": 26.09,
+      "grad_norm": 2.1571977138519287,
+      "learning_rate": 2e-05,
+      "loss": 0.05426963,
+      "step": 13045
+    },
+    {
+      "epoch": 26.092,
+      "grad_norm": 1.3975282907485962,
+      "learning_rate": 2e-05,
+      "loss": 0.05105602,
+      "step": 13046
+    },
+    {
+      "epoch": 26.094,
+      "grad_norm": 1.2141252756118774,
+      "learning_rate": 2e-05,
+      "loss": 0.05173802,
+      "step": 13047
+    },
+    {
+      "epoch": 26.096,
+      "grad_norm": 2.8250155448913574,
+      "learning_rate": 2e-05,
+      "loss": 0.06396338,
+      "step": 13048
+    },
+    {
+      "epoch": 26.098,
+      "grad_norm": 1.162996530532837,
+      "learning_rate": 2e-05,
+      "loss": 0.05161465,
+      "step": 13049
+    },
+    {
+      "epoch": 26.1,
+      "grad_norm": 1.0605719089508057,
+      "learning_rate": 2e-05,
+      "loss": 0.03604612,
+      "step": 13050
+    },
+    {
+      "epoch": 26.102,
+      "grad_norm": 1.1841809749603271,
+      "learning_rate": 2e-05,
+      "loss": 0.04377447,
+      "step": 13051
+    },
+    {
+      "epoch": 26.104,
+      "grad_norm": 1.2295196056365967,
+      "learning_rate": 2e-05,
+      "loss": 0.05042927,
+      "step": 13052
+    },
+    {
+      "epoch": 26.106,
+      "grad_norm": 1.04031503200531,
+      "learning_rate": 2e-05,
+      "loss": 0.02950481,
+      "step": 13053
+    },
+    {
+      "epoch": 26.108,
+      "grad_norm": 1.689014196395874,
+      "learning_rate": 2e-05,
+      "loss": 0.05845881,
+      "step": 13054
+    },
+    {
+      "epoch": 26.11,
+      "grad_norm": 1.123939871788025,
+      "learning_rate": 2e-05,
+      "loss": 0.04595611,
+      "step": 13055
+    },
+    {
+      "epoch": 26.112,
+      "grad_norm": 2.1183228492736816,
+      "learning_rate": 2e-05,
+      "loss": 0.03145133,
+      "step": 13056
+    },
+    {
+      "epoch": 26.114,
+      "grad_norm": 1.138588786125183,
+      "learning_rate": 2e-05,
+      "loss": 0.04393515,
+      "step": 13057
+    },
+    {
+      "epoch": 26.116,
+      "grad_norm": 1.1965045928955078,
+      "learning_rate": 2e-05,
+      "loss": 0.05045221,
+      "step": 13058
+    },
+    {
+      "epoch": 26.118,
+      "grad_norm": 1.3790438175201416,
+      "learning_rate": 2e-05,
+      "loss": 0.05804875,
+      "step": 13059
+    },
+    {
+      "epoch": 26.12,
+      "grad_norm": 1.0586308240890503,
+      "learning_rate": 2e-05,
+      "loss": 0.03835727,
+      "step": 13060
+    },
+    {
+      "epoch": 26.122,
+      "grad_norm": 0.9596161842346191,
+      "learning_rate": 2e-05,
+      "loss": 0.03627136,
+      "step": 13061
+    },
+    {
+      "epoch": 26.124,
+      "grad_norm": 1.7609083652496338,
+      "learning_rate": 2e-05,
+      "loss": 0.04719493,
+      "step": 13062
+    },
+    {
+      "epoch": 26.126,
+      "grad_norm": 1.9145840406417847,
+      "learning_rate": 2e-05,
+      "loss": 0.05495848,
+      "step": 13063
+    },
+    {
+      "epoch": 26.128,
+      "grad_norm": 3.6397500038146973,
+      "learning_rate": 2e-05,
+      "loss": 0.06928319,
+      "step": 13064
+    },
+    {
+      "epoch": 26.13,
+      "grad_norm": 1.7477335929870605,
+      "learning_rate": 2e-05,
+      "loss": 0.06259966,
+      "step": 13065
+    },
+    {
+      "epoch": 26.132,
+      "grad_norm": 1.4324450492858887,
+      "learning_rate": 2e-05,
+      "loss": 0.05550072,
+      "step": 13066
+    },
+    {
+      "epoch": 26.134,
+      "grad_norm": 1.6972732543945312,
+      "learning_rate": 2e-05,
+      "loss": 0.03786982,
+      "step": 13067
+    },
+    {
+      "epoch": 26.136,
+      "grad_norm": 2.0623652935028076,
+      "learning_rate": 2e-05,
+      "loss": 0.0455045,
+      "step": 13068
+    },
+    {
+      "epoch": 26.138,
+      "grad_norm": 1.056091547012329,
+      "learning_rate": 2e-05,
+      "loss": 0.03710656,
+      "step": 13069
+    },
+    {
+      "epoch": 26.14,
+      "grad_norm": 1.2790498733520508,
+      "learning_rate": 2e-05,
+      "loss": 0.0524893,
+      "step": 13070
+    },
+    {
+      "epoch": 26.142,
+      "grad_norm": 1.587206244468689,
+      "learning_rate": 2e-05,
+      "loss": 0.04856799,
+      "step": 13071
+    },
+    {
+      "epoch": 26.144,
+      "grad_norm": 1.0538597106933594,
+      "learning_rate": 2e-05,
+      "loss": 0.03798814,
+      "step": 13072
+    },
+    {
+      "epoch": 26.146,
+      "grad_norm": 1.031635046005249,
+      "learning_rate": 2e-05,
+      "loss": 0.03815435,
+      "step": 13073
+    },
+    {
+      "epoch": 26.148,
+      "grad_norm": 1.4059712886810303,
+      "learning_rate": 2e-05,
+      "loss": 0.0483502,
+      "step": 13074
+    },
+    {
+      "epoch": 26.15,
+      "grad_norm": 0.9544400572776794,
+      "learning_rate": 2e-05,
+      "loss": 0.02810543,
+      "step": 13075
+    },
+    {
+      "epoch": 26.152,
+      "grad_norm": 1.027392029762268,
+      "learning_rate": 2e-05,
+      "loss": 0.03287507,
+      "step": 13076
+    },
+    {
+      "epoch": 26.154,
+      "grad_norm": 1.084064245223999,
+      "learning_rate": 2e-05,
+      "loss": 0.03592873,
+      "step": 13077
+    },
+    {
+      "epoch": 26.156,
+      "grad_norm": 1.2711600065231323,
+      "learning_rate": 2e-05,
+      "loss": 0.04065077,
+      "step": 13078
+    },
+    {
+      "epoch": 26.158,
+      "grad_norm": 1.0948227643966675,
+      "learning_rate": 2e-05,
+      "loss": 0.03384892,
+      "step": 13079
+    },
+    {
+      "epoch": 26.16,
+      "grad_norm": 1.3854420185089111,
+      "learning_rate": 2e-05,
+      "loss": 0.04718272,
+      "step": 13080
+    },
+    {
+      "epoch": 26.162,
+      "grad_norm": 1.2978075742721558,
+      "learning_rate": 2e-05,
+      "loss": 0.0398887,
+      "step": 13081
+    },
+    {
+      "epoch": 26.164,
+      "grad_norm": 1.2315889596939087,
+      "learning_rate": 2e-05,
+      "loss": 0.03722582,
+      "step": 13082
+    },
+    {
+      "epoch": 26.166,
+      "grad_norm": 1.9231492280960083,
+      "learning_rate": 2e-05,
+      "loss": 0.05172444,
+      "step": 13083
+    },
+    {
+      "epoch": 26.168,
+      "grad_norm": 1.1194469928741455,
+      "learning_rate": 2e-05,
+      "loss": 0.03749369,
+      "step": 13084
+    },
+    {
+      "epoch": 26.17,
+      "grad_norm": 1.6649547815322876,
+      "learning_rate": 2e-05,
+      "loss": 0.05333608,
+      "step": 13085
+    },
+    {
+      "epoch": 26.172,
+      "grad_norm": 2.011101245880127,
+      "learning_rate": 2e-05,
+      "loss": 0.04135909,
+      "step": 13086
+    },
+    {
+      "epoch": 26.174,
+      "grad_norm": 1.6736245155334473,
+      "learning_rate": 2e-05,
+      "loss": 0.043147,
+      "step": 13087
+    },
+    {
+      "epoch": 26.176,
+      "grad_norm": 1.3968009948730469,
+      "learning_rate": 2e-05,
+      "loss": 0.04125686,
+      "step": 13088
+    },
+    {
+      "epoch": 26.178,
+      "grad_norm": 1.05874764919281,
+      "learning_rate": 2e-05,
+      "loss": 0.03640444,
+      "step": 13089
+    },
+    {
+      "epoch": 26.18,
+      "grad_norm": 1.14805006980896,
+      "learning_rate": 2e-05,
+      "loss": 0.0389007,
+      "step": 13090
+    },
+    {
+      "epoch": 26.182,
+      "grad_norm": 1.106278419494629,
+      "learning_rate": 2e-05,
+      "loss": 0.04055985,
+      "step": 13091
+    },
+    {
+      "epoch": 26.184,
+      "grad_norm": 2.2514469623565674,
+      "learning_rate": 2e-05,
+      "loss": 0.04158842,
+      "step": 13092
+    },
+    {
+      "epoch": 26.186,
+      "grad_norm": 1.33916437625885,
+      "learning_rate": 2e-05,
+      "loss": 0.04522097,
+      "step": 13093
+    },
+    {
+      "epoch": 26.188,
+      "grad_norm": 2.1648190021514893,
+      "learning_rate": 2e-05,
+      "loss": 0.04654589,
+      "step": 13094
+    },
+    {
+      "epoch": 26.19,
+      "grad_norm": 1.1959937810897827,
+      "learning_rate": 2e-05,
+      "loss": 0.04704513,
+      "step": 13095
+    },
+    {
+      "epoch": 26.192,
+      "grad_norm": 1.733241081237793,
+      "learning_rate": 2e-05,
+      "loss": 0.05815672,
+      "step": 13096
+    },
+    {
+      "epoch": 26.194,
+      "grad_norm": 0.8504005670547485,
+      "learning_rate": 2e-05,
+      "loss": 0.02225319,
+      "step": 13097
+    },
+    {
+      "epoch": 26.196,
+      "grad_norm": 2.0521740913391113,
+      "learning_rate": 2e-05,
+      "loss": 0.05633967,
+      "step": 13098
+    },
+    {
+      "epoch": 26.198,
+      "grad_norm": 1.087705373764038,
+      "learning_rate": 2e-05,
+      "loss": 0.04992017,
+      "step": 13099
+    },
+    {
+      "epoch": 26.2,
+      "grad_norm": 1.121282696723938,
+      "learning_rate": 2e-05,
+      "loss": 0.03436579,
+      "step": 13100
+    },
+    {
+      "epoch": 26.202,
+      "grad_norm": 1.6202627420425415,
+      "learning_rate": 2e-05,
+      "loss": 0.04655264,
+      "step": 13101
+    },
+    {
+      "epoch": 26.204,
+      "grad_norm": 1.7769203186035156,
+      "learning_rate": 2e-05,
+      "loss": 0.05040466,
+      "step": 13102
+    },
+    {
+      "epoch": 26.206,
+      "grad_norm": 1.3449256420135498,
+      "learning_rate": 2e-05,
+      "loss": 0.0527843,
+      "step": 13103
+    },
+    {
+      "epoch": 26.208,
+      "grad_norm": 2.069218397140503,
+      "learning_rate": 2e-05,
+      "loss": 0.05726243,
+      "step": 13104
+    },
+    {
+      "epoch": 26.21,
+      "grad_norm": 1.4746137857437134,
+      "learning_rate": 2e-05,
+      "loss": 0.06106601,
+      "step": 13105
+    },
+    {
+      "epoch": 26.212,
+      "grad_norm": 1.4619065523147583,
+      "learning_rate": 2e-05,
+      "loss": 0.03955628,
+      "step": 13106
+    },
+    {
+      "epoch": 26.214,
+      "grad_norm": 0.9846628308296204,
+      "learning_rate": 2e-05,
+      "loss": 0.03640333,
+      "step": 13107
+    },
+    {
+      "epoch": 26.216,
+      "grad_norm": 1.4153640270233154,
+      "learning_rate": 2e-05,
+      "loss": 0.04586072,
+      "step": 13108
+    },
+    {
+      "epoch": 26.218,
+      "grad_norm": 1.6593786478042603,
+      "learning_rate": 2e-05,
+      "loss": 0.04747347,
+      "step": 13109
+    },
+    {
+      "epoch": 26.22,
+      "grad_norm": 1.1394518613815308,
+      "learning_rate": 2e-05,
+      "loss": 0.04001589,
+      "step": 13110
+    },
+    {
+      "epoch": 26.222,
+      "grad_norm": 1.1682400703430176,
+      "learning_rate": 2e-05,
+      "loss": 0.05425475,
+      "step": 13111
+    },
+    {
+      "epoch": 26.224,
+      "grad_norm": 1.2645248174667358,
+      "learning_rate": 2e-05,
+      "loss": 0.03757692,
+      "step": 13112
+    },
+    {
+      "epoch": 26.226,
+      "grad_norm": 1.0900676250457764,
+      "learning_rate": 2e-05,
+      "loss": 0.03077891,
+      "step": 13113
+    },
+    {
+      "epoch": 26.228,
+      "grad_norm": 1.570024013519287,
+      "learning_rate": 2e-05,
+      "loss": 0.04349901,
+      "step": 13114
+    },
+    {
+      "epoch": 26.23,
+      "grad_norm": 1.2463085651397705,
+      "learning_rate": 2e-05,
+      "loss": 0.05144939,
+      "step": 13115
+    },
+    {
+      "epoch": 26.232,
+      "grad_norm": 2.3603127002716064,
+      "learning_rate": 2e-05,
+      "loss": 0.05303069,
+      "step": 13116
+    },
+    {
+      "epoch": 26.234,
+      "grad_norm": 1.4877861738204956,
+      "learning_rate": 2e-05,
+      "loss": 0.04663004,
+      "step": 13117
+    },
+    {
+      "epoch": 26.236,
+      "grad_norm": 2.002016544342041,
+      "learning_rate": 2e-05,
+      "loss": 0.02264161,
+      "step": 13118
+    },
+    {
+      "epoch": 26.238,
+      "grad_norm": 1.2714250087738037,
+      "learning_rate": 2e-05,
+      "loss": 0.05852698,
+      "step": 13119
+    },
+    {
+      "epoch": 26.24,
+      "grad_norm": 1.6923631429672241,
+      "learning_rate": 2e-05,
+      "loss": 0.03670363,
+      "step": 13120
+    },
+    {
+      "epoch": 26.242,
+      "grad_norm": 2.9363479614257812,
+      "learning_rate": 2e-05,
+      "loss": 0.05860969,
+      "step": 13121
+    },
+    {
+      "epoch": 26.244,
+      "grad_norm": 0.9841324090957642,
+      "learning_rate": 2e-05,
+      "loss": 0.03121854,
+      "step": 13122
+    },
+    {
+      "epoch": 26.246,
+      "grad_norm": 1.421994686126709,
+      "learning_rate": 2e-05,
+      "loss": 0.04863666,
+      "step": 13123
+    },
+    {
+      "epoch": 26.248,
+      "grad_norm": 1.1231731176376343,
+      "learning_rate": 2e-05,
+      "loss": 0.04469264,
+      "step": 13124
+    },
+    {
+      "epoch": 26.25,
+      "grad_norm": 1.8403480052947998,
+      "learning_rate": 2e-05,
+      "loss": 0.0488335,
+      "step": 13125
+    },
+    {
+      "epoch": 26.252,
+      "grad_norm": 2.583953619003296,
+      "learning_rate": 2e-05,
+      "loss": 0.04319778,
+      "step": 13126
+    },
+    {
+      "epoch": 26.254,
+      "grad_norm": 9.854668617248535,
+      "learning_rate": 2e-05,
+      "loss": 0.08909784,
+      "step": 13127
+    },
+    {
+      "epoch": 26.256,
+      "grad_norm": 52.158538818359375,
+      "learning_rate": 2e-05,
+      "loss": 0.10964724,
+      "step": 13128
+    },
+    {
+      "epoch": 26.258,
+      "grad_norm": 1.0165956020355225,
+      "learning_rate": 2e-05,
+      "loss": 0.0289845,
+      "step": 13129
+    },
+    {
+      "epoch": 26.26,
+      "grad_norm": 18.144968032836914,
+      "learning_rate": 2e-05,
+      "loss": 0.25045502,
+      "step": 13130
+    },
+    {
+      "epoch": 26.262,
+      "grad_norm": 4.379713535308838,
+      "learning_rate": 2e-05,
+      "loss": 0.03416047,
+      "step": 13131
+    },
+    {
+      "epoch": 26.264,
+      "grad_norm": 25.38080596923828,
+      "learning_rate": 2e-05,
+      "loss": 0.12389896,
+      "step": 13132
+    },
+    {
+      "epoch": 26.266,
+      "grad_norm": 1.384544014930725,
+      "learning_rate": 2e-05,
+      "loss": 0.04459827,
+      "step": 13133
+    },
+    {
+      "epoch": 26.268,
+      "grad_norm": 1.378836750984192,
+      "learning_rate": 2e-05,
+      "loss": 0.04477947,
+      "step": 13134
+    },
+    {
+      "epoch": 26.27,
+      "grad_norm": 2.1596052646636963,
+      "learning_rate": 2e-05,
+      "loss": 0.05260035,
+      "step": 13135
+    },
+    {
+      "epoch": 26.272,
+      "grad_norm": 1.0437805652618408,
+      "learning_rate": 2e-05,
+      "loss": 0.03832313,
+      "step": 13136
+    },
+    {
+      "epoch": 26.274,
+      "grad_norm": 1.2373123168945312,
+      "learning_rate": 2e-05,
+      "loss": 0.04758352,
+      "step": 13137
+    },
+    {
+      "epoch": 26.276,
+      "grad_norm": 1.18471097946167,
+      "learning_rate": 2e-05,
+      "loss": 0.0513614,
+      "step": 13138
+    },
+    {
+      "epoch": 26.278,
+      "grad_norm": 1.364261507987976,
+      "learning_rate": 2e-05,
+      "loss": 0.03552637,
+      "step": 13139
+    },
+    {
+      "epoch": 26.28,
+      "grad_norm": 1.291282296180725,
+      "learning_rate": 2e-05,
+      "loss": 0.04767456,
+      "step": 13140
+    },
+    {
+      "epoch": 26.282,
+      "grad_norm": 2.0756161212921143,
+      "learning_rate": 2e-05,
+      "loss": 0.05183939,
+      "step": 13141
+    },
+    {
+      "epoch": 26.284,
+      "grad_norm": 0.7976016402244568,
+      "learning_rate": 2e-05,
+      "loss": 0.02524626,
+      "step": 13142
+    },
+    {
+      "epoch": 26.286,
+      "grad_norm": 1.0466339588165283,
+      "learning_rate": 2e-05,
+      "loss": 0.05251211,
+      "step": 13143
+    },
+    {
+      "epoch": 26.288,
+      "grad_norm": 1.6141102313995361,
+      "learning_rate": 2e-05,
+      "loss": 0.02955076,
+      "step": 13144
+    },
+    {
+      "epoch": 26.29,
+      "grad_norm": 1.5093724727630615,
+      "learning_rate": 2e-05,
+      "loss": 0.05642843,
+      "step": 13145
+    },
+    {
+      "epoch": 26.292,
+      "grad_norm": 1.431347131729126,
+      "learning_rate": 2e-05,
+      "loss": 0.04590548,
+      "step": 13146
+    },
+    {
+      "epoch": 26.294,
+      "grad_norm": 1.239392876625061,
+      "learning_rate": 2e-05,
+      "loss": 0.04617111,
+      "step": 13147
+    },
+    {
+      "epoch": 26.296,
+      "grad_norm": 1.1694923639297485,
+      "learning_rate": 2e-05,
+      "loss": 0.03655789,
+      "step": 13148
+    },
+    {
+      "epoch": 26.298,
+      "grad_norm": 1.7410540580749512,
+      "learning_rate": 2e-05,
+      "loss": 0.05551188,
+      "step": 13149
+    },
+    {
+      "epoch": 26.3,
+      "grad_norm": 1.2234365940093994,
+      "learning_rate": 2e-05,
+      "loss": 0.03679156,
+      "step": 13150
+    },
+    {
+      "epoch": 26.302,
+      "grad_norm": 0.9119114875793457,
+      "learning_rate": 2e-05,
+      "loss": 0.02980712,
+      "step": 13151
+    },
+    {
+      "epoch": 26.304,
+      "grad_norm": 2.018742561340332,
+      "learning_rate": 2e-05,
+      "loss": 0.05444905,
+      "step": 13152
+    },
+    {
+      "epoch": 26.306,
+      "grad_norm": 1.7591094970703125,
+      "learning_rate": 2e-05,
+      "loss": 0.06758714,
+      "step": 13153
+    },
+    {
+      "epoch": 26.308,
+      "grad_norm": 1.300772786140442,
+      "learning_rate": 2e-05,
+      "loss": 0.0449839,
+      "step": 13154
+    },
+    {
+      "epoch": 26.31,
+      "grad_norm": 1.3861204385757446,
+      "learning_rate": 2e-05,
+      "loss": 0.04811238,
+      "step": 13155
+    },
+    {
+      "epoch": 26.312,
+      "grad_norm": 1.2385214567184448,
+      "learning_rate": 2e-05,
+      "loss": 0.03959964,
+      "step": 13156
+    },
+    {
+      "epoch": 26.314,
+      "grad_norm": 1.3340177536010742,
+      "learning_rate": 2e-05,
+      "loss": 0.04750549,
+      "step": 13157
+    },
+    {
+      "epoch": 26.316,
+      "grad_norm": 2.2569944858551025,
+      "learning_rate": 2e-05,
+      "loss": 0.05605095,
+      "step": 13158
+    },
+    {
+      "epoch": 26.318,
+      "grad_norm": 2.1105170249938965,
+      "learning_rate": 2e-05,
+      "loss": 0.05581967,
+      "step": 13159
+    },
+    {
+      "epoch": 26.32,
+      "grad_norm": 1.4911856651306152,
+      "learning_rate": 2e-05,
+      "loss": 0.06944424,
+      "step": 13160
+    },
+    {
+      "epoch": 26.322,
+      "grad_norm": 0.8688299655914307,
+      "learning_rate": 2e-05,
+      "loss": 0.01928959,
+      "step": 13161
+    },
+    {
+      "epoch": 26.324,
+      "grad_norm": 1.162561297416687,
+      "learning_rate": 2e-05,
+      "loss": 0.03705761,
+      "step": 13162
+    },
+    {
+      "epoch": 26.326,
+      "grad_norm": 1.117928385734558,
+      "learning_rate": 2e-05,
+      "loss": 0.04116834,
+      "step": 13163
+    },
+    {
+      "epoch": 26.328,
+      "grad_norm": 3.4923346042633057,
+      "learning_rate": 2e-05,
+      "loss": 0.04038919,
+      "step": 13164
+    },
+    {
+      "epoch": 26.33,
+      "grad_norm": 1.4581316709518433,
+      "learning_rate": 2e-05,
+      "loss": 0.03603051,
+      "step": 13165
+    },
+    {
+      "epoch": 26.332,
+      "grad_norm": 1.2061631679534912,
+      "learning_rate": 2e-05,
+      "loss": 0.03987415,
+      "step": 13166
+    },
+    {
+      "epoch": 26.334,
+      "grad_norm": 2.4524641036987305,
+      "learning_rate": 2e-05,
+      "loss": 0.06676086,
+      "step": 13167
+    },
+    {
+      "epoch": 26.336,
+      "grad_norm": 1.0566414594650269,
+      "learning_rate": 2e-05,
+      "loss": 0.04725492,
+      "step": 13168
+    },
+    {
+      "epoch": 26.338,
+      "grad_norm": 0.9778428077697754,
+      "learning_rate": 2e-05,
+      "loss": 0.03469347,
+      "step": 13169
+    },
+    {
+      "epoch": 26.34,
+      "grad_norm": 1.4649147987365723,
+      "learning_rate": 2e-05,
+      "loss": 0.04087276,
+      "step": 13170
+    },
+    {
+      "epoch": 26.342,
+      "grad_norm": 1.4261488914489746,
+      "learning_rate": 2e-05,
+      "loss": 0.04543655,
+      "step": 13171
+    },
+    {
+      "epoch": 26.344,
+      "grad_norm": 1.2056576013565063,
+      "learning_rate": 2e-05,
+      "loss": 0.04065976,
+      "step": 13172
+    },
+    {
+      "epoch": 26.346,
+      "grad_norm": 1.3527207374572754,
+      "learning_rate": 2e-05,
+      "loss": 0.05951834,
+      "step": 13173
+    },
+    {
+      "epoch": 26.348,
+      "grad_norm": 1.4392766952514648,
+      "learning_rate": 2e-05,
+      "loss": 0.04477938,
+      "step": 13174
+    },
+    {
+      "epoch": 26.35,
+      "grad_norm": 1.0581060647964478,
+      "learning_rate": 2e-05,
+      "loss": 0.04285652,
+      "step": 13175
+    },
+    {
+      "epoch": 26.352,
+      "grad_norm": 1.3287581205368042,
+      "learning_rate": 2e-05,
+      "loss": 0.05369517,
+      "step": 13176
+    },
+    {
+      "epoch": 26.354,
+      "grad_norm": 1.4318236112594604,
+      "learning_rate": 2e-05,
+      "loss": 0.03889168,
+      "step": 13177
+    },
+    {
+      "epoch": 26.356,
+      "grad_norm": 4.168168544769287,
+      "learning_rate": 2e-05,
+      "loss": 0.04014147,
+      "step": 13178
+    },
+    {
+      "epoch": 26.358,
+      "grad_norm": 2.5664639472961426,
+      "learning_rate": 2e-05,
+      "loss": 0.0509958,
+      "step": 13179
+    },
+    {
+      "epoch": 26.36,
+      "grad_norm": 1.6791688203811646,
+      "learning_rate": 2e-05,
+      "loss": 0.0571983,
+      "step": 13180
+    },
+    {
+      "epoch": 26.362,
+      "grad_norm": 1.597809076309204,
+      "learning_rate": 2e-05,
+      "loss": 0.04761202,
+      "step": 13181
+    },
+    {
+      "epoch": 26.364,
+      "grad_norm": 1.0649417638778687,
+      "learning_rate": 2e-05,
+      "loss": 0.03972602,
+      "step": 13182
+    },
+    {
+      "epoch": 26.366,
+      "grad_norm": 0.9360977411270142,
+      "learning_rate": 2e-05,
+      "loss": 0.03613855,
+      "step": 13183
+    },
+    {
+      "epoch": 26.368,
+      "grad_norm": 0.9736573696136475,
+      "learning_rate": 2e-05,
+      "loss": 0.03216821,
+      "step": 13184
+    },
+    {
+      "epoch": 26.37,
+      "grad_norm": 1.6165627241134644,
+      "learning_rate": 2e-05,
+      "loss": 0.05862963,
+      "step": 13185
+    },
+    {
+      "epoch": 26.372,
+      "grad_norm": 1.3186026811599731,
+      "learning_rate": 2e-05,
+      "loss": 0.04542302,
+      "step": 13186
+    },
+    {
+      "epoch": 26.374,
+      "grad_norm": 1.6366772651672363,
+      "learning_rate": 2e-05,
+      "loss": 0.04529631,
+      "step": 13187
+    },
+    {
+      "epoch": 26.376,
+      "grad_norm": 1.0712040662765503,
+      "learning_rate": 2e-05,
+      "loss": 0.03381637,
+      "step": 13188
+    },
+    {
+      "epoch": 26.378,
+      "grad_norm": 1.1308493614196777,
+      "learning_rate": 2e-05,
+      "loss": 0.04229905,
+      "step": 13189
+    },
+    {
+      "epoch": 26.38,
+      "grad_norm": 1.3662320375442505,
+      "learning_rate": 2e-05,
+      "loss": 0.05177809,
+      "step": 13190
+    },
+    {
+      "epoch": 26.382,
+      "grad_norm": 1.9889822006225586,
+      "learning_rate": 2e-05,
+      "loss": 0.07416862,
+      "step": 13191
+    },
+    {
+      "epoch": 26.384,
+      "grad_norm": 1.464272379875183,
+      "learning_rate": 2e-05,
+      "loss": 0.04299095,
+      "step": 13192
+    },
+    {
+      "epoch": 26.386,
+      "grad_norm": 1.1341749429702759,
+      "learning_rate": 2e-05,
+      "loss": 0.03406891,
+      "step": 13193
+    },
+    {
+      "epoch": 26.388,
+      "grad_norm": 1.3196316957473755,
+      "learning_rate": 2e-05,
+      "loss": 0.05596439,
+      "step": 13194
+    },
+    {
+      "epoch": 26.39,
+      "grad_norm": 1.1319140195846558,
+      "learning_rate": 2e-05,
+      "loss": 0.03612085,
+      "step": 13195
+    },
+    {
+      "epoch": 26.392,
+      "grad_norm": 0.9723718762397766,
+      "learning_rate": 2e-05,
+      "loss": 0.03945912,
+      "step": 13196
+    },
+    {
+      "epoch": 26.394,
+      "grad_norm": 0.8996081352233887,
+      "learning_rate": 2e-05,
+      "loss": 0.03478167,
+      "step": 13197
+    },
+    {
+      "epoch": 26.396,
+      "grad_norm": 1.8404150009155273,
+      "learning_rate": 2e-05,
+      "loss": 0.04371741,
+      "step": 13198
+    },
+    {
+      "epoch": 26.398,
+      "grad_norm": 2.319556713104248,
+      "learning_rate": 2e-05,
+      "loss": 0.05231041,
+      "step": 13199
+    },
+    {
+      "epoch": 26.4,
+      "grad_norm": 2.335930824279785,
+      "learning_rate": 2e-05,
+      "loss": 0.0324939,
+      "step": 13200
+    },
+    {
+      "epoch": 26.402,
+      "grad_norm": 1.6622105836868286,
+      "learning_rate": 2e-05,
+      "loss": 0.02796814,
+      "step": 13201
+    },
+    {
+      "epoch": 26.404,
+      "grad_norm": 2.356870651245117,
+      "learning_rate": 2e-05,
+      "loss": 0.04440472,
+      "step": 13202
+    },
+    {
+      "epoch": 26.406,
+      "grad_norm": 1.4844690561294556,
+      "learning_rate": 2e-05,
+      "loss": 0.05918201,
+      "step": 13203
+    },
+    {
+      "epoch": 26.408,
+      "grad_norm": 1.3682526350021362,
+      "learning_rate": 2e-05,
+      "loss": 0.064179,
+      "step": 13204
+    },
+    {
+      "epoch": 26.41,
+      "grad_norm": 1.0957056283950806,
+      "learning_rate": 2e-05,
+      "loss": 0.046661,
+      "step": 13205
+    },
+    {
+      "epoch": 26.412,
+      "grad_norm": 1.2934621572494507,
+      "learning_rate": 2e-05,
+      "loss": 0.04822002,
+      "step": 13206
+    },
+    {
+      "epoch": 26.414,
+      "grad_norm": 1.6465322971343994,
+      "learning_rate": 2e-05,
+      "loss": 0.05369683,
+      "step": 13207
+    },
+    {
+      "epoch": 26.416,
+      "grad_norm": 1.0616811513900757,
+      "learning_rate": 2e-05,
+      "loss": 0.04072348,
+      "step": 13208
+    },
+    {
+      "epoch": 26.418,
+      "grad_norm": 4.23490047454834,
+      "learning_rate": 2e-05,
+      "loss": 0.05251167,
+      "step": 13209
+    },
+    {
+      "epoch": 26.42,
+      "grad_norm": 1.170454978942871,
+      "learning_rate": 2e-05,
+      "loss": 0.03898567,
+      "step": 13210
+    },
+    {
+      "epoch": 26.422,
+      "grad_norm": 1.7174979448318481,
+      "learning_rate": 2e-05,
+      "loss": 0.04406247,
+      "step": 13211
+    },
+    {
+      "epoch": 26.424,
+      "grad_norm": 1.1600885391235352,
+      "learning_rate": 2e-05,
+      "loss": 0.05435538,
+      "step": 13212
+    },
+    {
+      "epoch": 26.426,
+      "grad_norm": 2.2355101108551025,
+      "learning_rate": 2e-05,
+      "loss": 0.05025771,
+      "step": 13213
+    },
+    {
+      "epoch": 26.428,
+      "grad_norm": 1.322678804397583,
+      "learning_rate": 2e-05,
+      "loss": 0.04307056,
+      "step": 13214
+    },
+    {
+      "epoch": 26.43,
+      "grad_norm": 1.323901891708374,
+      "learning_rate": 2e-05,
+      "loss": 0.04015271,
+      "step": 13215
+    },
+    {
+      "epoch": 26.432,
+      "grad_norm": 1.8236982822418213,
+      "learning_rate": 2e-05,
+      "loss": 0.04085752,
+      "step": 13216
+    },
+    {
+      "epoch": 26.434,
+      "grad_norm": 1.3549261093139648,
+      "learning_rate": 2e-05,
+      "loss": 0.04854722,
+      "step": 13217
+    },
+    {
+      "epoch": 26.436,
+      "grad_norm": 1.7481671571731567,
+      "learning_rate": 2e-05,
+      "loss": 0.04112818,
+      "step": 13218
+    },
+    {
+      "epoch": 26.438,
+      "grad_norm": 1.3064723014831543,
+      "learning_rate": 2e-05,
+      "loss": 0.05516483,
+      "step": 13219
+    },
+    {
+      "epoch": 26.44,
+      "grad_norm": 3.367008686065674,
+      "learning_rate": 2e-05,
+      "loss": 0.05630918,
+      "step": 13220
+    },
+    {
+      "epoch": 26.442,
+      "grad_norm": 1.8773056268692017,
+      "learning_rate": 2e-05,
+      "loss": 0.06338409,
+      "step": 13221
+    },
+    {
+      "epoch": 26.444,
+      "grad_norm": 1.168461561203003,
+      "learning_rate": 2e-05,
+      "loss": 0.04579566,
+      "step": 13222
+    },
+    {
+      "epoch": 26.446,
+      "grad_norm": 1.4216398000717163,
+      "learning_rate": 2e-05,
+      "loss": 0.03986656,
+      "step": 13223
+    },
+    {
+      "epoch": 26.448,
+      "grad_norm": 0.9925347566604614,
+      "learning_rate": 2e-05,
+      "loss": 0.03974314,
+      "step": 13224
+    },
+    {
+      "epoch": 26.45,
+      "grad_norm": 1.569463849067688,
+      "learning_rate": 2e-05,
+      "loss": 0.0388345,
+      "step": 13225
+    },
+    {
+      "epoch": 26.452,
+      "grad_norm": 1.6125260591506958,
+      "learning_rate": 2e-05,
+      "loss": 0.0357227,
+      "step": 13226
+    },
+    {
+      "epoch": 26.454,
+      "grad_norm": 1.2239407300949097,
+      "learning_rate": 2e-05,
+      "loss": 0.04946052,
+      "step": 13227
+    },
+    {
+      "epoch": 26.456,
+      "grad_norm": 0.8565706014633179,
+      "learning_rate": 2e-05,
+      "loss": 0.03200175,
+      "step": 13228
+    },
+    {
+      "epoch": 26.458,
+      "grad_norm": 1.0081415176391602,
+      "learning_rate": 2e-05,
+      "loss": 0.0341515,
+      "step": 13229
+    },
+    {
+      "epoch": 26.46,
+      "grad_norm": 1.0593496561050415,
+      "learning_rate": 2e-05,
+      "loss": 0.03936413,
+      "step": 13230
+    },
+    {
+      "epoch": 26.462,
+      "grad_norm": 1.1382449865341187,
+      "learning_rate": 2e-05,
+      "loss": 0.02872925,
+      "step": 13231
+    },
+    {
+      "epoch": 26.464,
+      "grad_norm": 1.4055805206298828,
+      "learning_rate": 2e-05,
+      "loss": 0.03323968,
+      "step": 13232
+    },
+    {
+      "epoch": 26.466,
+      "grad_norm": 1.23487389087677,
+      "learning_rate": 2e-05,
+      "loss": 0.05444343,
+      "step": 13233
+    },
+    {
+      "epoch": 26.468,
+      "grad_norm": 1.100627064704895,
+      "learning_rate": 2e-05,
+      "loss": 0.04639729,
+      "step": 13234
+    },
+    {
+      "epoch": 26.47,
+      "grad_norm": 1.6120901107788086,
+      "learning_rate": 2e-05,
+      "loss": 0.05141643,
+      "step": 13235
+    },
+    {
+      "epoch": 26.472,
+      "grad_norm": 1.5615583658218384,
+      "learning_rate": 2e-05,
+      "loss": 0.04910383,
+      "step": 13236
+    },
+    {
+      "epoch": 26.474,
+      "grad_norm": 1.7524781227111816,
+      "learning_rate": 2e-05,
+      "loss": 0.0517026,
+      "step": 13237
+    },
+    {
+      "epoch": 26.476,
+      "grad_norm": 1.4189718961715698,
+      "learning_rate": 2e-05,
+      "loss": 0.03197357,
+      "step": 13238
+    },
+    {
+      "epoch": 26.478,
+      "grad_norm": 1.2002609968185425,
+      "learning_rate": 2e-05,
+      "loss": 0.04137202,
+      "step": 13239
+    },
+    {
+      "epoch": 26.48,
+      "grad_norm": 1.5820086002349854,
+      "learning_rate": 2e-05,
+      "loss": 0.05161813,
+      "step": 13240
+    },
+    {
+      "epoch": 26.482,
+      "grad_norm": 2.00372052192688,
+      "learning_rate": 2e-05,
+      "loss": 0.05241256,
+      "step": 13241
+    },
+    {
+      "epoch": 26.484,
+      "grad_norm": 1.582104206085205,
+      "learning_rate": 2e-05,
+      "loss": 0.05075792,
+      "step": 13242
+    },
+    {
+      "epoch": 26.486,
+      "grad_norm": 1.3096644878387451,
+      "learning_rate": 2e-05,
+      "loss": 0.04678864,
+      "step": 13243
+    },
+    {
+      "epoch": 26.488,
+      "grad_norm": 1.2743966579437256,
+      "learning_rate": 2e-05,
+      "loss": 0.05511482,
+      "step": 13244
+    },
+    {
+      "epoch": 26.49,
+      "grad_norm": 0.9095728397369385,
+      "learning_rate": 2e-05,
+      "loss": 0.0362159,
+      "step": 13245
+    },
+    {
+      "epoch": 26.492,
+      "grad_norm": 2.057610511779785,
+      "learning_rate": 2e-05,
+      "loss": 0.04670215,
+      "step": 13246
+    },
+    {
+      "epoch": 26.494,
+      "grad_norm": 1.7796850204467773,
+      "learning_rate": 2e-05,
+      "loss": 0.0543985,
+      "step": 13247
+    },
+    {
+      "epoch": 26.496,
+      "grad_norm": 2.0312533378601074,
+      "learning_rate": 2e-05,
+      "loss": 0.03435779,
+      "step": 13248
+    },
+    {
+      "epoch": 26.498,
+      "grad_norm": 0.8784269690513611,
+      "learning_rate": 2e-05,
+      "loss": 0.02505475,
+      "step": 13249
+    },
+    {
+      "epoch": 26.5,
+      "grad_norm": 1.3337388038635254,
+      "learning_rate": 2e-05,
+      "loss": 0.03425723,
+      "step": 13250
+    },
+    {
+      "epoch": 26.502,
+      "grad_norm": 1.1322087049484253,
+      "learning_rate": 2e-05,
+      "loss": 0.0380166,
+      "step": 13251
+    },
+    {
+      "epoch": 26.504,
+      "grad_norm": 1.4790630340576172,
+      "learning_rate": 2e-05,
+      "loss": 0.05530652,
+      "step": 13252
+    },
+    {
+      "epoch": 26.506,
+      "grad_norm": 1.203283429145813,
+      "learning_rate": 2e-05,
+      "loss": 0.04369481,
+      "step": 13253
+    },
+    {
+      "epoch": 26.508,
+      "grad_norm": 0.9502495527267456,
+      "learning_rate": 2e-05,
+      "loss": 0.03597327,
+      "step": 13254
+    },
+    {
+      "epoch": 26.51,
+      "grad_norm": 1.1296265125274658,
+      "learning_rate": 2e-05,
+      "loss": 0.03838349,
+      "step": 13255
+    },
+    {
+      "epoch": 26.512,
+      "grad_norm": 1.4536200761795044,
+      "learning_rate": 2e-05,
+      "loss": 0.06651304,
+      "step": 13256
+    },
+    {
+      "epoch": 26.514,
+      "grad_norm": 1.022311806678772,
+      "learning_rate": 2e-05,
+      "loss": 0.03070169,
+      "step": 13257
+    },
+    {
+      "epoch": 26.516,
+      "grad_norm": 1.7230098247528076,
+      "learning_rate": 2e-05,
+      "loss": 0.05207474,
+      "step": 13258
+    },
+    {
+      "epoch": 26.518,
+      "grad_norm": 1.0804290771484375,
+      "learning_rate": 2e-05,
+      "loss": 0.03778263,
+      "step": 13259
+    },
+    {
+      "epoch": 26.52,
+      "grad_norm": 0.8944292068481445,
+      "learning_rate": 2e-05,
+      "loss": 0.02704319,
+      "step": 13260
+    },
+    {
+      "epoch": 26.522,
+      "grad_norm": 1.2151002883911133,
+      "learning_rate": 2e-05,
+      "loss": 0.03234539,
+      "step": 13261
+    },
+    {
+      "epoch": 26.524,
+      "grad_norm": 1.4448111057281494,
+      "learning_rate": 2e-05,
+      "loss": 0.04379795,
+      "step": 13262
+    },
+    {
+      "epoch": 26.526,
+      "grad_norm": 3.976080894470215,
+      "learning_rate": 2e-05,
+      "loss": 0.04685978,
+      "step": 13263
+    },
+    {
+      "epoch": 26.528,
+      "grad_norm": 1.7629016637802124,
+      "learning_rate": 2e-05,
+      "loss": 0.0644666,
+      "step": 13264
+    },
+    {
+      "epoch": 26.53,
+      "grad_norm": 1.4677865505218506,
+      "learning_rate": 2e-05,
+      "loss": 0.0506825,
+      "step": 13265
+    },
+    {
+      "epoch": 26.532,
+      "grad_norm": 0.9562329649925232,
+      "learning_rate": 2e-05,
+      "loss": 0.03811152,
+      "step": 13266
+    },
+    {
+      "epoch": 26.534,
+      "grad_norm": 2.1469244956970215,
+      "learning_rate": 2e-05,
+      "loss": 0.04386814,
+      "step": 13267
+    },
+    {
+      "epoch": 26.536,
+      "grad_norm": 1.172168254852295,
+      "learning_rate": 2e-05,
+      "loss": 0.03443911,
+      "step": 13268
+    },
+    {
+      "epoch": 26.538,
+      "grad_norm": 1.5751445293426514,
+      "learning_rate": 2e-05,
+      "loss": 0.05732818,
+      "step": 13269
+    },
+    {
+      "epoch": 26.54,
+      "grad_norm": 0.9219292402267456,
+      "learning_rate": 2e-05,
+      "loss": 0.02910371,
+      "step": 13270
+    },
+    {
+      "epoch": 26.542,
+      "grad_norm": 1.7685173749923706,
+      "learning_rate": 2e-05,
+      "loss": 0.05187953,
+      "step": 13271
+    },
+    {
+      "epoch": 26.544,
+      "grad_norm": 1.689509630203247,
+      "learning_rate": 2e-05,
+      "loss": 0.05508012,
+      "step": 13272
+    },
+    {
+      "epoch": 26.546,
+      "grad_norm": 1.1748220920562744,
+      "learning_rate": 2e-05,
+      "loss": 0.03818913,
+      "step": 13273
+    },
+    {
+      "epoch": 26.548000000000002,
+      "grad_norm": 2.35014009475708,
+      "learning_rate": 2e-05,
+      "loss": 0.05988345,
+      "step": 13274
+    },
+    {
+      "epoch": 26.55,
+      "grad_norm": 0.9859991669654846,
+      "learning_rate": 2e-05,
+      "loss": 0.02704637,
+      "step": 13275
+    },
+    {
+      "epoch": 26.552,
+      "grad_norm": 1.1411195993423462,
+      "learning_rate": 2e-05,
+      "loss": 0.03679326,
+      "step": 13276
+    },
+    {
+      "epoch": 26.554,
+      "grad_norm": 2.2478442192077637,
+      "learning_rate": 2e-05,
+      "loss": 0.05884918,
+      "step": 13277
+    },
+    {
+      "epoch": 26.556,
+      "grad_norm": 1.2798776626586914,
+      "learning_rate": 2e-05,
+      "loss": 0.03827868,
+      "step": 13278
+    },
+    {
+      "epoch": 26.558,
+      "grad_norm": 1.3902195692062378,
+      "learning_rate": 2e-05,
+      "loss": 0.04380446,
+      "step": 13279
+    },
+    {
+      "epoch": 26.56,
+      "grad_norm": 1.3358991146087646,
+      "learning_rate": 2e-05,
+      "loss": 0.04080527,
+      "step": 13280
+    },
+    {
+      "epoch": 26.562,
+      "grad_norm": 1.769484043121338,
+      "learning_rate": 2e-05,
+      "loss": 0.06003635,
+      "step": 13281
+    },
+    {
+      "epoch": 26.564,
+      "grad_norm": 1.0746612548828125,
+      "learning_rate": 2e-05,
+      "loss": 0.02794596,
+      "step": 13282
+    },
+    {
+      "epoch": 26.566,
+      "grad_norm": 1.3699100017547607,
+      "learning_rate": 2e-05,
+      "loss": 0.03206124,
+      "step": 13283
+    },
+    {
+      "epoch": 26.568,
+      "grad_norm": 1.0902633666992188,
+      "learning_rate": 2e-05,
+      "loss": 0.0326319,
+      "step": 13284
+    },
+    {
+      "epoch": 26.57,
+      "grad_norm": 1.6095848083496094,
+      "learning_rate": 2e-05,
+      "loss": 0.05489502,
+      "step": 13285
+    },
+    {
+      "epoch": 26.572,
+      "grad_norm": 1.2423216104507446,
+      "learning_rate": 2e-05,
+      "loss": 0.03849361,
+      "step": 13286
+    },
+    {
+      "epoch": 26.574,
+      "grad_norm": 1.2105095386505127,
+      "learning_rate": 2e-05,
+      "loss": 0.0360062,
+      "step": 13287
+    },
+    {
+      "epoch": 26.576,
+      "grad_norm": 1.267501950263977,
+      "learning_rate": 2e-05,
+      "loss": 0.04801185,
+      "step": 13288
+    },
+    {
+      "epoch": 26.578,
+      "grad_norm": 2.0049567222595215,
+      "learning_rate": 2e-05,
+      "loss": 0.03750632,
+      "step": 13289
+    },
+    {
+      "epoch": 26.58,
+      "grad_norm": 2.5399057865142822,
+      "learning_rate": 2e-05,
+      "loss": 0.0422065,
+      "step": 13290
+    },
+    {
+      "epoch": 26.582,
+      "grad_norm": 1.4443379640579224,
+      "learning_rate": 2e-05,
+      "loss": 0.03480319,
+      "step": 13291
+    },
+    {
+      "epoch": 26.584,
+      "grad_norm": 1.135719656944275,
+      "learning_rate": 2e-05,
+      "loss": 0.03749589,
+      "step": 13292
+    },
+    {
+      "epoch": 26.586,
+      "grad_norm": 1.231929063796997,
+      "learning_rate": 2e-05,
+      "loss": 0.04628847,
+      "step": 13293
+    },
+    {
+      "epoch": 26.588,
+      "grad_norm": 1.4225656986236572,
+      "learning_rate": 2e-05,
+      "loss": 0.05674511,
+      "step": 13294
+    },
+    {
+      "epoch": 26.59,
+      "grad_norm": 1.0678924322128296,
+      "learning_rate": 2e-05,
+      "loss": 0.03472822,
+      "step": 13295
+    },
+    {
+      "epoch": 26.592,
+      "grad_norm": 1.3472518920898438,
+      "learning_rate": 2e-05,
+      "loss": 0.04637545,
+      "step": 13296
+    },
+    {
+      "epoch": 26.594,
+      "grad_norm": 1.0952340364456177,
+      "learning_rate": 2e-05,
+      "loss": 0.0380681,
+      "step": 13297
+    },
+    {
+      "epoch": 26.596,
+      "grad_norm": 2.9363811016082764,
+      "learning_rate": 2e-05,
+      "loss": 0.05129644,
+      "step": 13298
+    },
+    {
+      "epoch": 26.598,
+      "grad_norm": 1.1145724058151245,
+      "learning_rate": 2e-05,
+      "loss": 0.02599895,
+      "step": 13299
+    },
+    {
+      "epoch": 26.6,
+      "grad_norm": 0.9582138061523438,
+      "learning_rate": 2e-05,
+      "loss": 0.03655213,
+      "step": 13300
+    },
+    {
+      "epoch": 26.602,
+      "grad_norm": 0.9235168099403381,
+      "learning_rate": 2e-05,
+      "loss": 0.03182671,
+      "step": 13301
+    },
+    {
+      "epoch": 26.604,
+      "grad_norm": 1.1051247119903564,
+      "learning_rate": 2e-05,
+      "loss": 0.03536615,
+      "step": 13302
+    },
+    {
+      "epoch": 26.606,
+      "grad_norm": 1.4593842029571533,
+      "learning_rate": 2e-05,
+      "loss": 0.06312197,
+      "step": 13303
+    },
+    {
+      "epoch": 26.608,
+      "grad_norm": 1.04954993724823,
+      "learning_rate": 2e-05,
+      "loss": 0.04040704,
+      "step": 13304
+    },
+    {
+      "epoch": 26.61,
+      "grad_norm": 1.3381813764572144,
+      "learning_rate": 2e-05,
+      "loss": 0.04579471,
+      "step": 13305
+    },
+    {
+      "epoch": 26.612,
+      "grad_norm": 1.8087036609649658,
+      "learning_rate": 2e-05,
+      "loss": 0.05931551,
+      "step": 13306
+    },
+    {
+      "epoch": 26.614,
+      "grad_norm": 1.8034483194351196,
+      "learning_rate": 2e-05,
+      "loss": 0.05207924,
+      "step": 13307
+    },
+    {
+      "epoch": 26.616,
+      "grad_norm": 1.1185029745101929,
+      "learning_rate": 2e-05,
+      "loss": 0.03802615,
+      "step": 13308
+    },
+    {
+      "epoch": 26.618,
+      "grad_norm": 2.3116369247436523,
+      "learning_rate": 2e-05,
+      "loss": 0.05495331,
+      "step": 13309
+    },
+    {
+      "epoch": 26.62,
+      "grad_norm": 1.7399510145187378,
+      "learning_rate": 2e-05,
+      "loss": 0.04346514,
+      "step": 13310
+    },
+    {
+      "epoch": 26.622,
+      "grad_norm": 1.2430211305618286,
+      "learning_rate": 2e-05,
+      "loss": 0.03976776,
+      "step": 13311
+    },
+    {
+      "epoch": 26.624,
+      "grad_norm": 1.2265691757202148,
+      "learning_rate": 2e-05,
+      "loss": 0.03571905,
+      "step": 13312
+    },
+    {
+      "epoch": 26.626,
+      "grad_norm": 3.572852611541748,
+      "learning_rate": 2e-05,
+      "loss": 0.06186254,
+      "step": 13313
+    },
+    {
+      "epoch": 26.628,
+      "grad_norm": 1.096924066543579,
+      "learning_rate": 2e-05,
+      "loss": 0.03906501,
+      "step": 13314
+    },
+    {
+      "epoch": 26.63,
+      "grad_norm": 2.9369330406188965,
+      "learning_rate": 2e-05,
+      "loss": 0.05661511,
+      "step": 13315
+    },
+    {
+      "epoch": 26.632,
+      "grad_norm": 1.467205286026001,
+      "learning_rate": 2e-05,
+      "loss": 0.02951013,
+      "step": 13316
+    },
+    {
+      "epoch": 26.634,
+      "grad_norm": 1.185747742652893,
+      "learning_rate": 2e-05,
+      "loss": 0.03714722,
+      "step": 13317
+    },
+    {
+      "epoch": 26.636,
+      "grad_norm": 2.264068126678467,
+      "learning_rate": 2e-05,
+      "loss": 0.0546186,
+      "step": 13318
+    },
+    {
+      "epoch": 26.638,
+      "grad_norm": 1.3490537405014038,
+      "learning_rate": 2e-05,
+      "loss": 0.03904513,
+      "step": 13319
+    },
+    {
+      "epoch": 26.64,
+      "grad_norm": 2.2553672790527344,
+      "learning_rate": 2e-05,
+      "loss": 0.04899627,
+      "step": 13320
+    },
+    {
+      "epoch": 26.642,
+      "grad_norm": 1.3507367372512817,
+      "learning_rate": 2e-05,
+      "loss": 0.04580309,
+      "step": 13321
+    },
+    {
+      "epoch": 26.644,
+      "grad_norm": 1.3584458827972412,
+      "learning_rate": 2e-05,
+      "loss": 0.04628047,
+      "step": 13322
+    },
+    {
+      "epoch": 26.646,
+      "grad_norm": 1.6204533576965332,
+      "learning_rate": 2e-05,
+      "loss": 0.05554104,
+      "step": 13323
+    },
+    {
+      "epoch": 26.648,
+      "grad_norm": 2.109013319015503,
+      "learning_rate": 2e-05,
+      "loss": 0.05139774,
+      "step": 13324
+    },
+    {
+      "epoch": 26.65,
+      "grad_norm": 1.177474021911621,
+      "learning_rate": 2e-05,
+      "loss": 0.04520677,
+      "step": 13325
+    },
+    {
+      "epoch": 26.652,
+      "grad_norm": 1.265397071838379,
+      "learning_rate": 2e-05,
+      "loss": 0.0448864,
+      "step": 13326
+    },
+    {
+      "epoch": 26.654,
+      "grad_norm": 0.9993850588798523,
+      "learning_rate": 2e-05,
+      "loss": 0.04312583,
+      "step": 13327
+    },
+    {
+      "epoch": 26.656,
+      "grad_norm": 2.175631523132324,
+      "learning_rate": 2e-05,
+      "loss": 0.04794764,
+      "step": 13328
+    },
+    {
+      "epoch": 26.658,
+      "grad_norm": 1.602054476737976,
+      "learning_rate": 2e-05,
+      "loss": 0.06128681,
+      "step": 13329
+    },
+    {
+      "epoch": 26.66,
+      "grad_norm": 1.4241081476211548,
+      "learning_rate": 2e-05,
+      "loss": 0.04915302,
+      "step": 13330
+    },
+    {
+      "epoch": 26.662,
+      "grad_norm": 1.6712620258331299,
+      "learning_rate": 2e-05,
+      "loss": 0.0512425,
+      "step": 13331
+    },
+    {
+      "epoch": 26.664,
+      "grad_norm": 2.4270451068878174,
+      "learning_rate": 2e-05,
+      "loss": 0.07092887,
+      "step": 13332
+    },
+    {
+      "epoch": 26.666,
+      "grad_norm": 1.007644772529602,
+      "learning_rate": 2e-05,
+      "loss": 0.03094992,
+      "step": 13333
+    },
+    {
+      "epoch": 26.668,
+      "grad_norm": 1.4786583185195923,
+      "learning_rate": 2e-05,
+      "loss": 0.05066321,
+      "step": 13334
+    },
+    {
+      "epoch": 26.67,
+      "grad_norm": 1.4783406257629395,
+      "learning_rate": 2e-05,
+      "loss": 0.04284883,
+      "step": 13335
+    },
+    {
+      "epoch": 26.672,
+      "grad_norm": 2.875722885131836,
+      "learning_rate": 2e-05,
+      "loss": 0.05276472,
+      "step": 13336
+    },
+    {
+      "epoch": 26.674,
+      "grad_norm": 1.0497190952301025,
+      "learning_rate": 2e-05,
+      "loss": 0.03627689,
+      "step": 13337
+    },
+    {
+      "epoch": 26.676,
+      "grad_norm": 1.1127156019210815,
+      "learning_rate": 2e-05,
+      "loss": 0.03644489,
+      "step": 13338
+    },
+    {
+      "epoch": 26.678,
+      "grad_norm": 1.5005398988723755,
+      "learning_rate": 2e-05,
+      "loss": 0.0399859,
+      "step": 13339
+    },
+    {
+      "epoch": 26.68,
+      "grad_norm": 1.4886964559555054,
+      "learning_rate": 2e-05,
+      "loss": 0.05113001,
+      "step": 13340
+    },
+    {
+      "epoch": 26.682,
+      "grad_norm": 1.1941289901733398,
+      "learning_rate": 2e-05,
+      "loss": 0.04594672,
+      "step": 13341
+    },
+    {
+      "epoch": 26.684,
+      "grad_norm": 1.4600493907928467,
+      "learning_rate": 2e-05,
+      "loss": 0.05219441,
+      "step": 13342
+    },
+    {
+      "epoch": 26.686,
+      "grad_norm": 1.2766631841659546,
+      "learning_rate": 2e-05,
+      "loss": 0.05881347,
+      "step": 13343
+    },
+    {
+      "epoch": 26.688,
+      "grad_norm": 1.2152045965194702,
+      "learning_rate": 2e-05,
+      "loss": 0.02941299,
+      "step": 13344
+    },
+    {
+      "epoch": 26.69,
+      "grad_norm": 1.2594579458236694,
+      "learning_rate": 2e-05,
+      "loss": 0.04704573,
+      "step": 13345
+    },
+    {
+      "epoch": 26.692,
+      "grad_norm": 1.4651315212249756,
+      "learning_rate": 2e-05,
+      "loss": 0.04764215,
+      "step": 13346
+    },
+    {
+      "epoch": 26.694,
+      "grad_norm": 1.5274326801300049,
+      "learning_rate": 2e-05,
+      "loss": 0.06374133,
+      "step": 13347
+    },
+    {
+      "epoch": 26.696,
+      "grad_norm": 1.9418085813522339,
+      "learning_rate": 2e-05,
+      "loss": 0.06519874,
+      "step": 13348
+    },
+    {
+      "epoch": 26.698,
+      "grad_norm": 1.6129246950149536,
+      "learning_rate": 2e-05,
+      "loss": 0.03720499,
+      "step": 13349
+    },
+    {
+      "epoch": 26.7,
+      "grad_norm": 0.7866122126579285,
+      "learning_rate": 2e-05,
+      "loss": 0.02468162,
+      "step": 13350
+    },
+    {
+      "epoch": 26.701999999999998,
+      "grad_norm": 1.3163105249404907,
+      "learning_rate": 2e-05,
+      "loss": 0.05013556,
+      "step": 13351
+    },
+    {
+      "epoch": 26.704,
+      "grad_norm": 1.4515641927719116,
+      "learning_rate": 2e-05,
+      "loss": 0.04763116,
+      "step": 13352
+    },
+    {
+      "epoch": 26.706,
+      "grad_norm": 1.3559627532958984,
+      "learning_rate": 2e-05,
+      "loss": 0.04021575,
+      "step": 13353
+    },
+    {
+      "epoch": 26.708,
+      "grad_norm": 2.714416742324829,
+      "learning_rate": 2e-05,
+      "loss": 0.05345739,
+      "step": 13354
+    },
+    {
+      "epoch": 26.71,
+      "grad_norm": 2.2046430110931396,
+      "learning_rate": 2e-05,
+      "loss": 0.04232349,
+      "step": 13355
+    },
+    {
+      "epoch": 26.712,
+      "grad_norm": 1.0346970558166504,
+      "learning_rate": 2e-05,
+      "loss": 0.03661943,
+      "step": 13356
+    },
+    {
+      "epoch": 26.714,
+      "grad_norm": 1.037308692932129,
+      "learning_rate": 2e-05,
+      "loss": 0.0360332,
+      "step": 13357
+    },
+    {
+      "epoch": 26.716,
+      "grad_norm": 1.399788498878479,
+      "learning_rate": 2e-05,
+      "loss": 0.04450074,
+      "step": 13358
+    },
+    {
+      "epoch": 26.718,
+      "grad_norm": 1.1871050596237183,
+      "learning_rate": 2e-05,
+      "loss": 0.04489849,
+      "step": 13359
+    },
+    {
+      "epoch": 26.72,
+      "grad_norm": 1.4322409629821777,
+      "learning_rate": 2e-05,
+      "loss": 0.06084087,
+      "step": 13360
+    },
+    {
+      "epoch": 26.722,
+      "grad_norm": 1.2092519998550415,
+      "learning_rate": 2e-05,
+      "loss": 0.04185452,
+      "step": 13361
+    },
+    {
+      "epoch": 26.724,
+      "grad_norm": 1.5179123878479004,
+      "learning_rate": 2e-05,
+      "loss": 0.05089298,
+      "step": 13362
+    },
+    {
+      "epoch": 26.726,
+      "grad_norm": 1.2479588985443115,
+      "learning_rate": 2e-05,
+      "loss": 0.04334253,
+      "step": 13363
+    },
+    {
+      "epoch": 26.728,
+      "grad_norm": 1.2945444583892822,
+      "learning_rate": 2e-05,
+      "loss": 0.03954721,
+      "step": 13364
+    },
+    {
+      "epoch": 26.73,
+      "grad_norm": 1.0343196392059326,
+      "learning_rate": 2e-05,
+      "loss": 0.03208121,
+      "step": 13365
+    },
+    {
+      "epoch": 26.732,
+      "grad_norm": 1.5581246614456177,
+      "learning_rate": 2e-05,
+      "loss": 0.05093454,
+      "step": 13366
+    },
+    {
+      "epoch": 26.734,
+      "grad_norm": 1.2444391250610352,
+      "learning_rate": 2e-05,
+      "loss": 0.04247945,
+      "step": 13367
+    },
+    {
+      "epoch": 26.736,
+      "grad_norm": 0.9851118326187134,
+      "learning_rate": 2e-05,
+      "loss": 0.03192877,
+      "step": 13368
+    },
+    {
+      "epoch": 26.738,
+      "grad_norm": 1.6827157735824585,
+      "learning_rate": 2e-05,
+      "loss": 0.04289787,
+      "step": 13369
+    },
+    {
+      "epoch": 26.74,
+      "grad_norm": 1.6345899105072021,
+      "learning_rate": 2e-05,
+      "loss": 0.05687879,
+      "step": 13370
+    },
+    {
+      "epoch": 26.742,
+      "grad_norm": 1.2550289630889893,
+      "learning_rate": 2e-05,
+      "loss": 0.04343875,
+      "step": 13371
+    },
+    {
+      "epoch": 26.744,
+      "grad_norm": 1.3297144174575806,
+      "learning_rate": 2e-05,
+      "loss": 0.03608532,
+      "step": 13372
+    },
+    {
+      "epoch": 26.746,
+      "grad_norm": 1.9059828519821167,
+      "learning_rate": 2e-05,
+      "loss": 0.06531657,
+      "step": 13373
+    },
+    {
+      "epoch": 26.748,
+      "grad_norm": 1.4299732446670532,
+      "learning_rate": 2e-05,
+      "loss": 0.05242453,
+      "step": 13374
+    },
+    {
+      "epoch": 26.75,
+      "grad_norm": 1.1574034690856934,
+      "learning_rate": 2e-05,
+      "loss": 0.0402981,
+      "step": 13375
+    },
+    {
+      "epoch": 26.752,
+      "grad_norm": 2.168186902999878,
+      "learning_rate": 2e-05,
+      "loss": 0.052399,
+      "step": 13376
+    },
+    {
+      "epoch": 26.754,
+      "grad_norm": 1.0010058879852295,
+      "learning_rate": 2e-05,
+      "loss": 0.03435899,
+      "step": 13377
+    },
+    {
+      "epoch": 26.756,
+      "grad_norm": 1.2347971200942993,
+      "learning_rate": 2e-05,
+      "loss": 0.04020776,
+      "step": 13378
+    },
+    {
+      "epoch": 26.758,
+      "grad_norm": 1.1826847791671753,
+      "learning_rate": 2e-05,
+      "loss": 0.03945916,
+      "step": 13379
+    },
+    {
+      "epoch": 26.76,
+      "grad_norm": 1.6033371686935425,
+      "learning_rate": 2e-05,
+      "loss": 0.0463616,
+      "step": 13380
+    },
+    {
+      "epoch": 26.762,
+      "grad_norm": 1.1534563302993774,
+      "learning_rate": 2e-05,
+      "loss": 0.04242963,
+      "step": 13381
+    },
+    {
+      "epoch": 26.764,
+      "grad_norm": 1.3048988580703735,
+      "learning_rate": 2e-05,
+      "loss": 0.05592864,
+      "step": 13382
+    },
+    {
+      "epoch": 26.766,
+      "grad_norm": 1.2142225503921509,
+      "learning_rate": 2e-05,
+      "loss": 0.04550473,
+      "step": 13383
+    },
+    {
+      "epoch": 26.768,
+      "grad_norm": 1.946568489074707,
+      "learning_rate": 2e-05,
+      "loss": 0.06550965,
+      "step": 13384
+    },
+    {
+      "epoch": 26.77,
+      "grad_norm": 1.2773605585098267,
+      "learning_rate": 2e-05,
+      "loss": 0.06554952,
+      "step": 13385
+    },
+    {
+      "epoch": 26.772,
+      "grad_norm": 1.2231202125549316,
+      "learning_rate": 2e-05,
+      "loss": 0.03776004,
+      "step": 13386
+    },
+    {
+      "epoch": 26.774,
+      "grad_norm": 1.1051039695739746,
+      "learning_rate": 2e-05,
+      "loss": 0.02784972,
+      "step": 13387
+    },
+    {
+      "epoch": 26.776,
+      "grad_norm": 1.1587811708450317,
+      "learning_rate": 2e-05,
+      "loss": 0.03560489,
+      "step": 13388
+    },
+    {
+      "epoch": 26.778,
+      "grad_norm": 1.0926035642623901,
+      "learning_rate": 2e-05,
+      "loss": 0.03061144,
+      "step": 13389
+    },
+    {
+      "epoch": 26.78,
+      "grad_norm": 0.9534873962402344,
+      "learning_rate": 2e-05,
+      "loss": 0.03757609,
+      "step": 13390
+    },
+    {
+      "epoch": 26.782,
+      "grad_norm": 1.3637419939041138,
+      "learning_rate": 2e-05,
+      "loss": 0.04305605,
+      "step": 13391
+    },
+    {
+      "epoch": 26.784,
+      "grad_norm": 1.3764069080352783,
+      "learning_rate": 2e-05,
+      "loss": 0.0500427,
+      "step": 13392
+    },
+    {
+      "epoch": 26.786,
+      "grad_norm": 1.036009430885315,
+      "learning_rate": 2e-05,
+      "loss": 0.04027718,
+      "step": 13393
+    },
+    {
+      "epoch": 26.788,
+      "grad_norm": 1.7560526132583618,
+      "learning_rate": 2e-05,
+      "loss": 0.06162121,
+      "step": 13394
+    },
+    {
+      "epoch": 26.79,
+      "grad_norm": 1.822432279586792,
+      "learning_rate": 2e-05,
+      "loss": 0.04765713,
+      "step": 13395
+    },
+    {
+      "epoch": 26.792,
+      "grad_norm": 1.488826036453247,
+      "learning_rate": 2e-05,
+      "loss": 0.03322754,
+      "step": 13396
+    },
+    {
+      "epoch": 26.794,
+      "grad_norm": 1.4484844207763672,
+      "learning_rate": 2e-05,
+      "loss": 0.04716588,
+      "step": 13397
+    },
+    {
+      "epoch": 26.796,
+      "grad_norm": 1.1175645589828491,
+      "learning_rate": 2e-05,
+      "loss": 0.03686436,
+      "step": 13398
+    },
+    {
+      "epoch": 26.798000000000002,
+      "grad_norm": 1.6677278280258179,
+      "learning_rate": 2e-05,
+      "loss": 0.03573879,
+      "step": 13399
+    },
+    {
+      "epoch": 26.8,
+      "grad_norm": 1.8486576080322266,
+      "learning_rate": 2e-05,
+      "loss": 0.05908503,
+      "step": 13400
+    },
+    {
+      "epoch": 26.802,
+      "grad_norm": 2.1419644355773926,
+      "learning_rate": 2e-05,
+      "loss": 0.04133147,
+      "step": 13401
+    },
+    {
+      "epoch": 26.804,
+      "grad_norm": 1.5237059593200684,
+      "learning_rate": 2e-05,
+      "loss": 0.04093695,
+      "step": 13402
+    },
+    {
+      "epoch": 26.806,
+      "grad_norm": 0.9631497263908386,
+      "learning_rate": 2e-05,
+      "loss": 0.03032106,
+      "step": 13403
+    },
+    {
+      "epoch": 26.808,
+      "grad_norm": 1.6903769969940186,
+      "learning_rate": 2e-05,
+      "loss": 0.03851802,
+      "step": 13404
+    },
+    {
+      "epoch": 26.81,
+      "grad_norm": 2.2639379501342773,
+      "learning_rate": 2e-05,
+      "loss": 0.0339236,
+      "step": 13405
+    },
+    {
+      "epoch": 26.812,
+      "grad_norm": 1.8234732151031494,
+      "learning_rate": 2e-05,
+      "loss": 0.04727671,
+      "step": 13406
+    },
+    {
+      "epoch": 26.814,
+      "grad_norm": 1.174176573753357,
+      "learning_rate": 2e-05,
+      "loss": 0.04155207,
+      "step": 13407
+    },
+    {
+      "epoch": 26.816,
+      "grad_norm": 1.388486385345459,
+      "learning_rate": 2e-05,
+      "loss": 0.03067862,
+      "step": 13408
+    },
+    {
+      "epoch": 26.818,
+      "grad_norm": 1.2704886198043823,
+      "learning_rate": 2e-05,
+      "loss": 0.02589917,
+      "step": 13409
+    },
+    {
+      "epoch": 26.82,
+      "grad_norm": 1.151698350906372,
+      "learning_rate": 2e-05,
+      "loss": 0.05481917,
+      "step": 13410
+    },
+    {
+      "epoch": 26.822,
+      "grad_norm": 1.2374764680862427,
+      "learning_rate": 2e-05,
+      "loss": 0.04253447,
+      "step": 13411
+    },
+    {
+      "epoch": 26.824,
+      "grad_norm": 1.3624358177185059,
+      "learning_rate": 2e-05,
+      "loss": 0.04338268,
+      "step": 13412
+    },
+    {
+      "epoch": 26.826,
+      "grad_norm": 1.1399511098861694,
+      "learning_rate": 2e-05,
+      "loss": 0.04491112,
+      "step": 13413
+    },
+    {
+      "epoch": 26.828,
+      "grad_norm": 1.8094592094421387,
+      "learning_rate": 2e-05,
+      "loss": 0.04950622,
+      "step": 13414
+    },
+    {
+      "epoch": 26.83,
+      "grad_norm": 1.5777703523635864,
+      "learning_rate": 2e-05,
+      "loss": 0.04463634,
+      "step": 13415
+    },
+    {
+      "epoch": 26.832,
+      "grad_norm": 1.1766401529312134,
+      "learning_rate": 2e-05,
+      "loss": 0.04658622,
+      "step": 13416
+    },
+    {
+      "epoch": 26.834,
+      "grad_norm": 1.3368674516677856,
+      "learning_rate": 2e-05,
+      "loss": 0.04229524,
+      "step": 13417
+    },
+    {
+      "epoch": 26.836,
+      "grad_norm": 1.3476362228393555,
+      "learning_rate": 2e-05,
+      "loss": 0.03923023,
+      "step": 13418
+    },
+    {
+      "epoch": 26.838,
+      "grad_norm": 1.4750795364379883,
+      "learning_rate": 2e-05,
+      "loss": 0.06064068,
+      "step": 13419
+    },
+    {
+      "epoch": 26.84,
+      "grad_norm": 1.1680598258972168,
+      "learning_rate": 2e-05,
+      "loss": 0.04480653,
+      "step": 13420
+    },
+    {
+      "epoch": 26.842,
+      "grad_norm": 2.2910873889923096,
+      "learning_rate": 2e-05,
+      "loss": 0.06019218,
+      "step": 13421
+    },
+    {
+      "epoch": 26.844,
+      "grad_norm": 1.3044073581695557,
+      "learning_rate": 2e-05,
+      "loss": 0.06455293,
+      "step": 13422
+    },
+    {
+      "epoch": 26.846,
+      "grad_norm": 1.560138463973999,
+      "learning_rate": 2e-05,
+      "loss": 0.06546596,
+      "step": 13423
+    },
+    {
+      "epoch": 26.848,
+      "grad_norm": 1.6893231868743896,
+      "learning_rate": 2e-05,
+      "loss": 0.0381937,
+      "step": 13424
+    },
+    {
+      "epoch": 26.85,
+      "grad_norm": 1.344321846961975,
+      "learning_rate": 2e-05,
+      "loss": 0.05524303,
+      "step": 13425
+    },
+    {
+      "epoch": 26.852,
+      "grad_norm": 1.0075470209121704,
+      "learning_rate": 2e-05,
+      "loss": 0.03442932,
+      "step": 13426
+    },
+    {
+      "epoch": 26.854,
+      "grad_norm": 1.832734227180481,
+      "learning_rate": 2e-05,
+      "loss": 0.06405198,
+      "step": 13427
+    },
+    {
+      "epoch": 26.856,
+      "grad_norm": 0.9961326122283936,
+      "learning_rate": 2e-05,
+      "loss": 0.02738804,
+      "step": 13428
+    },
+    {
+      "epoch": 26.858,
+      "grad_norm": 0.9834625124931335,
+      "learning_rate": 2e-05,
+      "loss": 0.03841186,
+      "step": 13429
+    },
+    {
+      "epoch": 26.86,
+      "grad_norm": 1.1662216186523438,
+      "learning_rate": 2e-05,
+      "loss": 0.04747712,
+      "step": 13430
+    },
+    {
+      "epoch": 26.862,
+      "grad_norm": 1.8224124908447266,
+      "learning_rate": 2e-05,
+      "loss": 0.038079,
+      "step": 13431
+    },
+    {
+      "epoch": 26.864,
+      "grad_norm": 1.2939532995224,
+      "learning_rate": 2e-05,
+      "loss": 0.04183429,
+      "step": 13432
+    },
+    {
+      "epoch": 26.866,
+      "grad_norm": 1.1002147197723389,
+      "learning_rate": 2e-05,
+      "loss": 0.03624891,
+      "step": 13433
+    },
+    {
+      "epoch": 26.868,
+      "grad_norm": 1.4251444339752197,
+      "learning_rate": 2e-05,
+      "loss": 0.05983438,
+      "step": 13434
+    },
+    {
+      "epoch": 26.87,
+      "grad_norm": 1.7216922044754028,
+      "learning_rate": 2e-05,
+      "loss": 0.09802464,
+      "step": 13435
+    },
+    {
+      "epoch": 26.872,
+      "grad_norm": 1.3873647451400757,
+      "learning_rate": 2e-05,
+      "loss": 0.04079521,
+      "step": 13436
+    },
+    {
+      "epoch": 26.874,
+      "grad_norm": 1.0737560987472534,
+      "learning_rate": 2e-05,
+      "loss": 0.04044875,
+      "step": 13437
+    },
+    {
+      "epoch": 26.876,
+      "grad_norm": 1.148348331451416,
+      "learning_rate": 2e-05,
+      "loss": 0.04982531,
+      "step": 13438
+    },
+    {
+      "epoch": 26.878,
+      "grad_norm": 1.1479979753494263,
+      "learning_rate": 2e-05,
+      "loss": 0.05051372,
+      "step": 13439
+    },
+    {
+      "epoch": 26.88,
+      "grad_norm": 1.727099061012268,
+      "learning_rate": 2e-05,
+      "loss": 0.04605647,
+      "step": 13440
+    },
+    {
+      "epoch": 26.882,
+      "grad_norm": 1.8774471282958984,
+      "learning_rate": 2e-05,
+      "loss": 0.04778759,
+      "step": 13441
+    },
+    {
+      "epoch": 26.884,
+      "grad_norm": 1.4502348899841309,
+      "learning_rate": 2e-05,
+      "loss": 0.03749793,
+      "step": 13442
+    },
+    {
+      "epoch": 26.886,
+      "grad_norm": 2.354149103164673,
+      "learning_rate": 2e-05,
+      "loss": 0.05970661,
+      "step": 13443
+    },
+    {
+      "epoch": 26.888,
+      "grad_norm": 1.1341004371643066,
+      "learning_rate": 2e-05,
+      "loss": 0.04527812,
+      "step": 13444
+    },
+    {
+      "epoch": 26.89,
+      "grad_norm": 1.6915284395217896,
+      "learning_rate": 2e-05,
+      "loss": 0.04350384,
+      "step": 13445
+    },
+    {
+      "epoch": 26.892,
+      "grad_norm": 1.5579638481140137,
+      "learning_rate": 2e-05,
+      "loss": 0.0307783,
+      "step": 13446
+    },
+    {
+      "epoch": 26.894,
+      "grad_norm": 1.3309636116027832,
+      "learning_rate": 2e-05,
+      "loss": 0.0582713,
+      "step": 13447
+    },
+    {
+      "epoch": 26.896,
+      "grad_norm": 1.2728379964828491,
+      "learning_rate": 2e-05,
+      "loss": 0.04255494,
+      "step": 13448
+    },
+    {
+      "epoch": 26.898,
+      "grad_norm": 1.1662172079086304,
+      "learning_rate": 2e-05,
+      "loss": 0.03874854,
+      "step": 13449
+    },
+    {
+      "epoch": 26.9,
+      "grad_norm": 2.72029447555542,
+      "learning_rate": 2e-05,
+      "loss": 0.0405199,
+      "step": 13450
+    },
+    {
+      "epoch": 26.902,
+      "grad_norm": 1.25313401222229,
+      "learning_rate": 2e-05,
+      "loss": 0.04713894,
+      "step": 13451
+    },
+    {
+      "epoch": 26.904,
+      "grad_norm": 1.2333053350448608,
+      "learning_rate": 2e-05,
+      "loss": 0.03969055,
+      "step": 13452
+    },
+    {
+      "epoch": 26.906,
+      "grad_norm": 1.149762511253357,
+      "learning_rate": 2e-05,
+      "loss": 0.03785865,
+      "step": 13453
+    },
+    {
+      "epoch": 26.908,
+      "grad_norm": 0.8945139646530151,
+      "learning_rate": 2e-05,
+      "loss": 0.03088951,
+      "step": 13454
+    },
+    {
+      "epoch": 26.91,
+      "grad_norm": 1.6503828763961792,
+      "learning_rate": 2e-05,
+      "loss": 0.03936726,
+      "step": 13455
+    },
+    {
+      "epoch": 26.912,
+      "grad_norm": 1.1440142393112183,
+      "learning_rate": 2e-05,
+      "loss": 0.04367042,
+      "step": 13456
+    },
+    {
+      "epoch": 26.914,
+      "grad_norm": 1.2169008255004883,
+      "learning_rate": 2e-05,
+      "loss": 0.04358334,
+      "step": 13457
+    },
+    {
+      "epoch": 26.916,
+      "grad_norm": 0.9487123489379883,
+      "learning_rate": 2e-05,
+      "loss": 0.02613301,
+      "step": 13458
+    },
+    {
+      "epoch": 26.918,
+      "grad_norm": 1.0501741170883179,
+      "learning_rate": 2e-05,
+      "loss": 0.03885425,
+      "step": 13459
+    },
+    {
+      "epoch": 26.92,
+      "grad_norm": 1.3245856761932373,
+      "learning_rate": 2e-05,
+      "loss": 0.03801511,
+      "step": 13460
+    },
+    {
+      "epoch": 26.922,
+      "grad_norm": 1.1823177337646484,
+      "learning_rate": 2e-05,
+      "loss": 0.04091566,
+      "step": 13461
+    },
+    {
+      "epoch": 26.924,
+      "grad_norm": 1.5307384729385376,
+      "learning_rate": 2e-05,
+      "loss": 0.04803543,
+      "step": 13462
+    },
+    {
+      "epoch": 26.926,
+      "grad_norm": 1.2377997636795044,
+      "learning_rate": 2e-05,
+      "loss": 0.04242143,
+      "step": 13463
+    },
+    {
+      "epoch": 26.928,
+      "grad_norm": 1.0577019453048706,
+      "learning_rate": 2e-05,
+      "loss": 0.04687814,
+      "step": 13464
+    },
+    {
+      "epoch": 26.93,
+      "grad_norm": 1.1497957706451416,
+      "learning_rate": 2e-05,
+      "loss": 0.03771542,
+      "step": 13465
+    },
+    {
+      "epoch": 26.932,
+      "grad_norm": 1.1796302795410156,
+      "learning_rate": 2e-05,
+      "loss": 0.04946107,
+      "step": 13466
+    },
+    {
+      "epoch": 26.934,
+      "grad_norm": 1.092475175857544,
+      "learning_rate": 2e-05,
+      "loss": 0.03278045,
+      "step": 13467
+    },
+    {
+      "epoch": 26.936,
+      "grad_norm": 1.85981285572052,
+      "learning_rate": 2e-05,
+      "loss": 0.04644259,
+      "step": 13468
+    },
+    {
+      "epoch": 26.938,
+      "grad_norm": 1.8343428373336792,
+      "learning_rate": 2e-05,
+      "loss": 0.05847539,
+      "step": 13469
+    },
+    {
+      "epoch": 26.94,
+      "grad_norm": 2.2233643531799316,
+      "learning_rate": 2e-05,
+      "loss": 0.05392168,
+      "step": 13470
+    },
+    {
+      "epoch": 26.942,
+      "grad_norm": 2.9804506301879883,
+      "learning_rate": 2e-05,
+      "loss": 0.05396551,
+      "step": 13471
+    },
+    {
+      "epoch": 26.944,
+      "grad_norm": 0.8914803862571716,
+      "learning_rate": 2e-05,
+      "loss": 0.02593549,
+      "step": 13472
+    },
+    {
+      "epoch": 26.946,
+      "grad_norm": 1.0326448678970337,
+      "learning_rate": 2e-05,
+      "loss": 0.03847109,
+      "step": 13473
+    },
+    {
+      "epoch": 26.948,
+      "grad_norm": 1.935031533241272,
+      "learning_rate": 2e-05,
+      "loss": 0.0416937,
+      "step": 13474
+    },
+    {
+      "epoch": 26.95,
+      "grad_norm": 1.1312302350997925,
+      "learning_rate": 2e-05,
+      "loss": 0.04824629,
+      "step": 13475
+    },
+    {
+      "epoch": 26.951999999999998,
+      "grad_norm": 1.6324800252914429,
+      "learning_rate": 2e-05,
+      "loss": 0.04897019,
+      "step": 13476
+    },
+    {
+      "epoch": 26.954,
+      "grad_norm": 1.1452162265777588,
+      "learning_rate": 2e-05,
+      "loss": 0.03952339,
+      "step": 13477
+    },
+    {
+      "epoch": 26.956,
+      "grad_norm": 1.284656047821045,
+      "learning_rate": 2e-05,
+      "loss": 0.03469507,
+      "step": 13478
+    },
+    {
+      "epoch": 26.958,
+      "grad_norm": 1.433944582939148,
+      "learning_rate": 2e-05,
+      "loss": 0.03940383,
+      "step": 13479
+    },
+    {
+      "epoch": 26.96,
+      "grad_norm": 0.9404771327972412,
+      "learning_rate": 2e-05,
+      "loss": 0.03143159,
+      "step": 13480
+    },
+    {
+      "epoch": 26.962,
+      "grad_norm": 1.0665420293807983,
+      "learning_rate": 2e-05,
+      "loss": 0.03889358,
+      "step": 13481
+    },
+    {
+      "epoch": 26.964,
+      "grad_norm": 1.0098894834518433,
+      "learning_rate": 2e-05,
+      "loss": 0.03096025,
+      "step": 13482
+    },
+    {
+      "epoch": 26.966,
+      "grad_norm": 1.2511407136917114,
+      "learning_rate": 2e-05,
+      "loss": 0.05308896,
+      "step": 13483
+    },
+    {
+      "epoch": 26.968,
+      "grad_norm": 1.9953383207321167,
+      "learning_rate": 2e-05,
+      "loss": 0.03955588,
+      "step": 13484
+    },
+    {
+      "epoch": 26.97,
+      "grad_norm": 1.377296805381775,
+      "learning_rate": 2e-05,
+      "loss": 0.03692595,
+      "step": 13485
+    },
+    {
+      "epoch": 26.972,
+      "grad_norm": 1.0224653482437134,
+      "learning_rate": 2e-05,
+      "loss": 0.03503982,
+      "step": 13486
+    },
+    {
+      "epoch": 26.974,
+      "grad_norm": 1.0944488048553467,
+      "learning_rate": 2e-05,
+      "loss": 0.03338355,
+      "step": 13487
+    },
+    {
+      "epoch": 26.976,
+      "grad_norm": 1.0018296241760254,
+      "learning_rate": 2e-05,
+      "loss": 0.02892867,
+      "step": 13488
+    },
+    {
+      "epoch": 26.978,
+      "grad_norm": 1.019363522529602,
+      "learning_rate": 2e-05,
+      "loss": 0.03823832,
+      "step": 13489
+    },
+    {
+      "epoch": 26.98,
+      "grad_norm": 1.2241965532302856,
+      "learning_rate": 2e-05,
+      "loss": 0.04636896,
+      "step": 13490
+    },
+    {
+      "epoch": 26.982,
+      "grad_norm": 1.5487970113754272,
+      "learning_rate": 2e-05,
+      "loss": 0.05474398,
+      "step": 13491
+    },
+    {
+      "epoch": 26.984,
+      "grad_norm": 0.9554542899131775,
+      "learning_rate": 2e-05,
+      "loss": 0.03364764,
+      "step": 13492
+    },
+    {
+      "epoch": 26.986,
+      "grad_norm": 2.8217360973358154,
+      "learning_rate": 2e-05,
+      "loss": 0.03697766,
+      "step": 13493
+    },
+    {
+      "epoch": 26.988,
+      "grad_norm": 1.6074271202087402,
+      "learning_rate": 2e-05,
+      "loss": 0.04182863,
+      "step": 13494
+    },
+    {
+      "epoch": 26.99,
+      "grad_norm": 1.3544893264770508,
+      "learning_rate": 2e-05,
+      "loss": 0.04771724,
+      "step": 13495
+    },
+    {
+      "epoch": 26.992,
+      "grad_norm": 0.9438838362693787,
+      "learning_rate": 2e-05,
+      "loss": 0.03837147,
+      "step": 13496
+    },
+    {
+      "epoch": 26.994,
+      "grad_norm": 1.067697525024414,
+      "learning_rate": 2e-05,
+      "loss": 0.03082527,
+      "step": 13497
+    },
+    {
+      "epoch": 26.996,
+      "grad_norm": 3.1041669845581055,
+      "learning_rate": 2e-05,
+      "loss": 0.05515031,
+      "step": 13498
+    },
+    {
+      "epoch": 26.998,
+      "grad_norm": 1.1457984447479248,
+      "learning_rate": 2e-05,
+      "loss": 0.04233011,
+      "step": 13499
+    },
+    {
+      "epoch": 27.0,
+      "grad_norm": 1.2711799144744873,
+      "learning_rate": 2e-05,
+      "loss": 0.04493605,
+      "step": 13500
+    },
+    {
+      "epoch": 27.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.998,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9740518962075848,
+        "Equal_1": 0.998,
+        "Equal_2": 0.9860279441117764,
+        "Equal_3": 0.9920159680638723,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 0.998003992015968,
+        "LineComparison_3": 0.998003992015968,
+        "Parallel_1": 0.9939879759519038,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.992,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.99,
+        "Perpendicular_3": 0.8977955911823647,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 0.9996666666666667,
+        "PointLiesOnCircle_3": 0.986,
+        "PointLiesOnLine_1": 0.9959919839679359,
+        "PointLiesOnLine_2": 0.9939879759519038,
+        "PointLiesOnLine_3": 0.9880239520958084
+      },
+      "eval_runtime": 319.7304,
+      "eval_samples_per_second": 32.84,
+      "eval_steps_per_second": 0.657,
+      "step": 13500
+    },
+    {
+      "epoch": 27.002,
+      "grad_norm": 1.1527739763259888,
+      "learning_rate": 2e-05,
+      "loss": 0.04508916,
+      "step": 13501
+    },
+    {
+      "epoch": 27.004,
+      "grad_norm": 1.562675952911377,
+      "learning_rate": 2e-05,
+      "loss": 0.04754071,
+      "step": 13502
+    },
+    {
+      "epoch": 27.006,
+      "grad_norm": 1.1957426071166992,
+      "learning_rate": 2e-05,
+      "loss": 0.03879406,
+      "step": 13503
+    },
+    {
+      "epoch": 27.008,
+      "grad_norm": 1.1542894840240479,
+      "learning_rate": 2e-05,
+      "loss": 0.0369472,
+      "step": 13504
+    },
+    {
+      "epoch": 27.01,
+      "grad_norm": 0.9654834866523743,
+      "learning_rate": 2e-05,
+      "loss": 0.03338242,
+      "step": 13505
+    },
+    {
+      "epoch": 27.012,
+      "grad_norm": 0.94996577501297,
+      "learning_rate": 2e-05,
+      "loss": 0.03290822,
+      "step": 13506
+    },
+    {
+      "epoch": 27.014,
+      "grad_norm": 1.7157975435256958,
+      "learning_rate": 2e-05,
+      "loss": 0.0513361,
+      "step": 13507
+    },
+    {
+      "epoch": 27.016,
+      "grad_norm": 1.248037576675415,
+      "learning_rate": 2e-05,
+      "loss": 0.03775504,
+      "step": 13508
+    },
+    {
+      "epoch": 27.018,
+      "grad_norm": 0.9875592589378357,
+      "learning_rate": 2e-05,
+      "loss": 0.03280458,
+      "step": 13509
+    },
+    {
+      "epoch": 27.02,
+      "grad_norm": 1.2043684720993042,
+      "learning_rate": 2e-05,
+      "loss": 0.03919297,
+      "step": 13510
+    },
+    {
+      "epoch": 27.022,
+      "grad_norm": 1.1797631978988647,
+      "learning_rate": 2e-05,
+      "loss": 0.03209981,
+      "step": 13511
+    },
+    {
+      "epoch": 27.024,
+      "grad_norm": 0.9786536693572998,
+      "learning_rate": 2e-05,
+      "loss": 0.03343519,
+      "step": 13512
+    },
+    {
+      "epoch": 27.026,
+      "grad_norm": 1.6166154146194458,
+      "learning_rate": 2e-05,
+      "loss": 0.04526688,
+      "step": 13513
+    },
+    {
+      "epoch": 27.028,
+      "grad_norm": 1.9152218103408813,
+      "learning_rate": 2e-05,
+      "loss": 0.04358869,
+      "step": 13514
+    },
+    {
+      "epoch": 27.03,
+      "grad_norm": 1.1101301908493042,
+      "learning_rate": 2e-05,
+      "loss": 0.02884656,
+      "step": 13515
+    },
+    {
+      "epoch": 27.032,
+      "grad_norm": 1.529909610748291,
+      "learning_rate": 2e-05,
+      "loss": 0.03502784,
+      "step": 13516
+    },
+    {
+      "epoch": 27.034,
+      "grad_norm": 0.97792649269104,
+      "learning_rate": 2e-05,
+      "loss": 0.02850207,
+      "step": 13517
+    },
+    {
+      "epoch": 27.036,
+      "grad_norm": 2.824786901473999,
+      "learning_rate": 2e-05,
+      "loss": 0.04891523,
+      "step": 13518
+    },
+    {
+      "epoch": 27.038,
+      "grad_norm": 1.4314439296722412,
+      "learning_rate": 2e-05,
+      "loss": 0.03873865,
+      "step": 13519
+    },
+    {
+      "epoch": 27.04,
+      "grad_norm": 1.3454880714416504,
+      "learning_rate": 2e-05,
+      "loss": 0.050864,
+      "step": 13520
+    },
+    {
+      "epoch": 27.042,
+      "grad_norm": 1.2566922903060913,
+      "learning_rate": 2e-05,
+      "loss": 0.05498883,
+      "step": 13521
+    },
+    {
+      "epoch": 27.044,
+      "grad_norm": 1.0700196027755737,
+      "learning_rate": 2e-05,
+      "loss": 0.03465045,
+      "step": 13522
+    },
+    {
+      "epoch": 27.046,
+      "grad_norm": 1.479763388633728,
+      "learning_rate": 2e-05,
+      "loss": 0.04723434,
+      "step": 13523
+    },
+    {
+      "epoch": 27.048,
+      "grad_norm": 1.4551644325256348,
+      "learning_rate": 2e-05,
+      "loss": 0.04056372,
+      "step": 13524
+    },
+    {
+      "epoch": 27.05,
+      "grad_norm": 1.3529044389724731,
+      "learning_rate": 2e-05,
+      "loss": 0.06563016,
+      "step": 13525
+    },
+    {
+      "epoch": 27.052,
+      "grad_norm": 1.122623085975647,
+      "learning_rate": 2e-05,
+      "loss": 0.04986753,
+      "step": 13526
+    },
+    {
+      "epoch": 27.054,
+      "grad_norm": 1.1177496910095215,
+      "learning_rate": 2e-05,
+      "loss": 0.04144656,
+      "step": 13527
+    },
+    {
+      "epoch": 27.056,
+      "grad_norm": 1.2102779150009155,
+      "learning_rate": 2e-05,
+      "loss": 0.04852476,
+      "step": 13528
+    },
+    {
+      "epoch": 27.058,
+      "grad_norm": 1.0115301609039307,
+      "learning_rate": 2e-05,
+      "loss": 0.03474714,
+      "step": 13529
+    },
+    {
+      "epoch": 27.06,
+      "grad_norm": 0.9496335983276367,
+      "learning_rate": 2e-05,
+      "loss": 0.03019882,
+      "step": 13530
+    },
+    {
+      "epoch": 27.062,
+      "grad_norm": 1.04624342918396,
+      "learning_rate": 2e-05,
+      "loss": 0.03820858,
+      "step": 13531
+    },
+    {
+      "epoch": 27.064,
+      "grad_norm": 1.3676940202713013,
+      "learning_rate": 2e-05,
+      "loss": 0.05121147,
+      "step": 13532
+    },
+    {
+      "epoch": 27.066,
+      "grad_norm": 1.2893519401550293,
+      "learning_rate": 2e-05,
+      "loss": 0.04120003,
+      "step": 13533
+    },
+    {
+      "epoch": 27.068,
+      "grad_norm": 3.120879650115967,
+      "learning_rate": 2e-05,
+      "loss": 0.05591538,
+      "step": 13534
+    },
+    {
+      "epoch": 27.07,
+      "grad_norm": 1.1986747980117798,
+      "learning_rate": 2e-05,
+      "loss": 0.04751917,
+      "step": 13535
+    },
+    {
+      "epoch": 27.072,
+      "grad_norm": 1.3522542715072632,
+      "learning_rate": 2e-05,
+      "loss": 0.04812311,
+      "step": 13536
+    },
+    {
+      "epoch": 27.074,
+      "grad_norm": 0.9945651292800903,
+      "learning_rate": 2e-05,
+      "loss": 0.04262705,
+      "step": 13537
+    },
+    {
+      "epoch": 27.076,
+      "grad_norm": 1.059237003326416,
+      "learning_rate": 2e-05,
+      "loss": 0.04054239,
+      "step": 13538
+    },
+    {
+      "epoch": 27.078,
+      "grad_norm": 2.535274028778076,
+      "learning_rate": 2e-05,
+      "loss": 0.0487668,
+      "step": 13539
+    },
+    {
+      "epoch": 27.08,
+      "grad_norm": 1.2254159450531006,
+      "learning_rate": 2e-05,
+      "loss": 0.04526926,
+      "step": 13540
+    },
+    {
+      "epoch": 27.082,
+      "grad_norm": 1.0619149208068848,
+      "learning_rate": 2e-05,
+      "loss": 0.03487138,
+      "step": 13541
+    },
+    {
+      "epoch": 27.084,
+      "grad_norm": 1.4296939373016357,
+      "learning_rate": 2e-05,
+      "loss": 0.05496912,
+      "step": 13542
+    },
+    {
+      "epoch": 27.086,
+      "grad_norm": 1.0750129222869873,
+      "learning_rate": 2e-05,
+      "loss": 0.04620029,
+      "step": 13543
+    },
+    {
+      "epoch": 27.088,
+      "grad_norm": 1.0724533796310425,
+      "learning_rate": 2e-05,
+      "loss": 0.0384186,
+      "step": 13544
+    },
+    {
+      "epoch": 27.09,
+      "grad_norm": 1.2280149459838867,
+      "learning_rate": 2e-05,
+      "loss": 0.04641897,
+      "step": 13545
+    },
+    {
+      "epoch": 27.092,
+      "grad_norm": 1.78925359249115,
+      "learning_rate": 2e-05,
+      "loss": 0.0408336,
+      "step": 13546
+    },
+    {
+      "epoch": 27.094,
+      "grad_norm": 1.6435742378234863,
+      "learning_rate": 2e-05,
+      "loss": 0.04276124,
+      "step": 13547
+    },
+    {
+      "epoch": 27.096,
+      "grad_norm": 1.4462822675704956,
+      "learning_rate": 2e-05,
+      "loss": 0.05522976,
+      "step": 13548
+    },
+    {
+      "epoch": 27.098,
+      "grad_norm": 1.1338396072387695,
+      "learning_rate": 2e-05,
+      "loss": 0.03534421,
+      "step": 13549
+    },
+    {
+      "epoch": 27.1,
+      "grad_norm": 1.2627896070480347,
+      "learning_rate": 2e-05,
+      "loss": 0.04895552,
+      "step": 13550
+    },
+    {
+      "epoch": 27.102,
+      "grad_norm": 1.2807172536849976,
+      "learning_rate": 2e-05,
+      "loss": 0.03973666,
+      "step": 13551
+    },
+    {
+      "epoch": 27.104,
+      "grad_norm": 1.1761025190353394,
+      "learning_rate": 2e-05,
+      "loss": 0.04104013,
+      "step": 13552
+    },
+    {
+      "epoch": 27.106,
+      "grad_norm": 2.8090035915374756,
+      "learning_rate": 2e-05,
+      "loss": 0.05685973,
+      "step": 13553
+    },
+    {
+      "epoch": 27.108,
+      "grad_norm": 1.8607357740402222,
+      "learning_rate": 2e-05,
+      "loss": 0.0470711,
+      "step": 13554
+    },
+    {
+      "epoch": 27.11,
+      "grad_norm": 1.0800814628601074,
+      "learning_rate": 2e-05,
+      "loss": 0.03861955,
+      "step": 13555
+    },
+    {
+      "epoch": 27.112,
+      "grad_norm": 2.145892858505249,
+      "learning_rate": 2e-05,
+      "loss": 0.05473135,
+      "step": 13556
+    },
+    {
+      "epoch": 27.114,
+      "grad_norm": 0.7809432148933411,
+      "learning_rate": 2e-05,
+      "loss": 0.02051185,
+      "step": 13557
+    },
+    {
+      "epoch": 27.116,
+      "grad_norm": 1.3810927867889404,
+      "learning_rate": 2e-05,
+      "loss": 0.05468324,
+      "step": 13558
+    },
+    {
+      "epoch": 27.118,
+      "grad_norm": 1.2400935888290405,
+      "learning_rate": 2e-05,
+      "loss": 0.03758835,
+      "step": 13559
+    },
+    {
+      "epoch": 27.12,
+      "grad_norm": 2.447911262512207,
+      "learning_rate": 2e-05,
+      "loss": 0.07153429,
+      "step": 13560
+    },
+    {
+      "epoch": 27.122,
+      "grad_norm": 1.063595175743103,
+      "learning_rate": 2e-05,
+      "loss": 0.04638623,
+      "step": 13561
+    },
+    {
+      "epoch": 27.124,
+      "grad_norm": 1.1095471382141113,
+      "learning_rate": 2e-05,
+      "loss": 0.03623333,
+      "step": 13562
+    },
+    {
+      "epoch": 27.126,
+      "grad_norm": 2.209012031555176,
+      "learning_rate": 2e-05,
+      "loss": 0.03872228,
+      "step": 13563
+    },
+    {
+      "epoch": 27.128,
+      "grad_norm": 1.9592890739440918,
+      "learning_rate": 2e-05,
+      "loss": 0.03570852,
+      "step": 13564
+    },
+    {
+      "epoch": 27.13,
+      "grad_norm": 1.3131749629974365,
+      "learning_rate": 2e-05,
+      "loss": 0.04663508,
+      "step": 13565
+    },
+    {
+      "epoch": 27.132,
+      "grad_norm": 1.461511492729187,
+      "learning_rate": 2e-05,
+      "loss": 0.04620817,
+      "step": 13566
+    },
+    {
+      "epoch": 27.134,
+      "grad_norm": 2.396193504333496,
+      "learning_rate": 2e-05,
+      "loss": 0.05443357,
+      "step": 13567
+    },
+    {
+      "epoch": 27.136,
+      "grad_norm": 1.4050869941711426,
+      "learning_rate": 2e-05,
+      "loss": 0.06443222,
+      "step": 13568
+    },
+    {
+      "epoch": 27.138,
+      "grad_norm": 2.008094072341919,
+      "learning_rate": 2e-05,
+      "loss": 0.05765811,
+      "step": 13569
+    },
+    {
+      "epoch": 27.14,
+      "grad_norm": 2.7625555992126465,
+      "learning_rate": 2e-05,
+      "loss": 0.07556622,
+      "step": 13570
+    },
+    {
+      "epoch": 27.142,
+      "grad_norm": 0.9222900867462158,
+      "learning_rate": 2e-05,
+      "loss": 0.02512233,
+      "step": 13571
+    },
+    {
+      "epoch": 27.144,
+      "grad_norm": 1.1549010276794434,
+      "learning_rate": 2e-05,
+      "loss": 0.03712163,
+      "step": 13572
+    },
+    {
+      "epoch": 27.146,
+      "grad_norm": 2.7713279724121094,
+      "learning_rate": 2e-05,
+      "loss": 0.04769222,
+      "step": 13573
+    },
+    {
+      "epoch": 27.148,
+      "grad_norm": 1.5160826444625854,
+      "learning_rate": 2e-05,
+      "loss": 0.04316017,
+      "step": 13574
+    },
+    {
+      "epoch": 27.15,
+      "grad_norm": 1.5735561847686768,
+      "learning_rate": 2e-05,
+      "loss": 0.05453978,
+      "step": 13575
+    },
+    {
+      "epoch": 27.152,
+      "grad_norm": 1.5878405570983887,
+      "learning_rate": 2e-05,
+      "loss": 0.0406678,
+      "step": 13576
+    },
+    {
+      "epoch": 27.154,
+      "grad_norm": 1.9090240001678467,
+      "learning_rate": 2e-05,
+      "loss": 0.04792298,
+      "step": 13577
+    },
+    {
+      "epoch": 27.156,
+      "grad_norm": 1.24784255027771,
+      "learning_rate": 2e-05,
+      "loss": 0.04576039,
+      "step": 13578
+    },
+    {
+      "epoch": 27.158,
+      "grad_norm": 1.1896952390670776,
+      "learning_rate": 2e-05,
+      "loss": 0.03995226,
+      "step": 13579
+    },
+    {
+      "epoch": 27.16,
+      "grad_norm": 1.1981700658798218,
+      "learning_rate": 2e-05,
+      "loss": 0.04131437,
+      "step": 13580
+    },
+    {
+      "epoch": 27.162,
+      "grad_norm": 1.583319067955017,
+      "learning_rate": 2e-05,
+      "loss": 0.03563969,
+      "step": 13581
+    },
+    {
+      "epoch": 27.164,
+      "grad_norm": 1.8085306882858276,
+      "learning_rate": 2e-05,
+      "loss": 0.04181949,
+      "step": 13582
+    },
+    {
+      "epoch": 27.166,
+      "grad_norm": 1.1413896083831787,
+      "learning_rate": 2e-05,
+      "loss": 0.04053574,
+      "step": 13583
+    },
+    {
+      "epoch": 27.168,
+      "grad_norm": 1.0028431415557861,
+      "learning_rate": 2e-05,
+      "loss": 0.04165848,
+      "step": 13584
+    },
+    {
+      "epoch": 27.17,
+      "grad_norm": 1.02912175655365,
+      "learning_rate": 2e-05,
+      "loss": 0.03777147,
+      "step": 13585
+    },
+    {
+      "epoch": 27.172,
+      "grad_norm": 1.0012383460998535,
+      "learning_rate": 2e-05,
+      "loss": 0.03452305,
+      "step": 13586
+    },
+    {
+      "epoch": 27.174,
+      "grad_norm": 1.588924765586853,
+      "learning_rate": 2e-05,
+      "loss": 0.06215196,
+      "step": 13587
+    },
+    {
+      "epoch": 27.176,
+      "grad_norm": 1.2847086191177368,
+      "learning_rate": 2e-05,
+      "loss": 0.0568499,
+      "step": 13588
+    },
+    {
+      "epoch": 27.178,
+      "grad_norm": 0.8857622146606445,
+      "learning_rate": 2e-05,
+      "loss": 0.03210337,
+      "step": 13589
+    },
+    {
+      "epoch": 27.18,
+      "grad_norm": 1.2372381687164307,
+      "learning_rate": 2e-05,
+      "loss": 0.03922543,
+      "step": 13590
+    },
+    {
+      "epoch": 27.182,
+      "grad_norm": 1.127703070640564,
+      "learning_rate": 2e-05,
+      "loss": 0.04630746,
+      "step": 13591
+    },
+    {
+      "epoch": 27.184,
+      "grad_norm": 1.2580335140228271,
+      "learning_rate": 2e-05,
+      "loss": 0.0422937,
+      "step": 13592
+    },
+    {
+      "epoch": 27.186,
+      "grad_norm": 1.6288437843322754,
+      "learning_rate": 2e-05,
+      "loss": 0.04852737,
+      "step": 13593
+    },
+    {
+      "epoch": 27.188,
+      "grad_norm": 1.931805968284607,
+      "learning_rate": 2e-05,
+      "loss": 0.05083385,
+      "step": 13594
+    },
+    {
+      "epoch": 27.19,
+      "grad_norm": 1.3048754930496216,
+      "learning_rate": 2e-05,
+      "loss": 0.05102141,
+      "step": 13595
+    },
+    {
+      "epoch": 27.192,
+      "grad_norm": 1.8813409805297852,
+      "learning_rate": 2e-05,
+      "loss": 0.06094917,
+      "step": 13596
+    },
+    {
+      "epoch": 27.194,
+      "grad_norm": 1.7660962343215942,
+      "learning_rate": 2e-05,
+      "loss": 0.04968596,
+      "step": 13597
+    },
+    {
+      "epoch": 27.196,
+      "grad_norm": 1.2359485626220703,
+      "learning_rate": 2e-05,
+      "loss": 0.05053048,
+      "step": 13598
+    },
+    {
+      "epoch": 27.198,
+      "grad_norm": 1.4421732425689697,
+      "learning_rate": 2e-05,
+      "loss": 0.03922078,
+      "step": 13599
+    },
+    {
+      "epoch": 27.2,
+      "grad_norm": 1.9196407794952393,
+      "learning_rate": 2e-05,
+      "loss": 0.04884792,
+      "step": 13600
+    },
+    {
+      "epoch": 27.202,
+      "grad_norm": 1.4350956678390503,
+      "learning_rate": 2e-05,
+      "loss": 0.05208459,
+      "step": 13601
+    },
+    {
+      "epoch": 27.204,
+      "grad_norm": 1.878467321395874,
+      "learning_rate": 2e-05,
+      "loss": 0.06585848,
+      "step": 13602
+    },
+    {
+      "epoch": 27.206,
+      "grad_norm": 2.1743922233581543,
+      "learning_rate": 2e-05,
+      "loss": 0.0544947,
+      "step": 13603
+    },
+    {
+      "epoch": 27.208,
+      "grad_norm": 1.2076090574264526,
+      "learning_rate": 2e-05,
+      "loss": 0.03394687,
+      "step": 13604
+    },
+    {
+      "epoch": 27.21,
+      "grad_norm": 2.1118581295013428,
+      "learning_rate": 2e-05,
+      "loss": 0.06034087,
+      "step": 13605
+    },
+    {
+      "epoch": 27.212,
+      "grad_norm": 0.812614381313324,
+      "learning_rate": 2e-05,
+      "loss": 0.03094057,
+      "step": 13606
+    },
+    {
+      "epoch": 27.214,
+      "grad_norm": 1.090105652809143,
+      "learning_rate": 2e-05,
+      "loss": 0.05721441,
+      "step": 13607
+    },
+    {
+      "epoch": 27.216,
+      "grad_norm": 1.6390467882156372,
+      "learning_rate": 2e-05,
+      "loss": 0.04660071,
+      "step": 13608
+    },
+    {
+      "epoch": 27.218,
+      "grad_norm": 2.038808584213257,
+      "learning_rate": 2e-05,
+      "loss": 0.05760818,
+      "step": 13609
+    },
+    {
+      "epoch": 27.22,
+      "grad_norm": 1.293735146522522,
+      "learning_rate": 2e-05,
+      "loss": 0.03360115,
+      "step": 13610
+    },
+    {
+      "epoch": 27.222,
+      "grad_norm": 1.192157506942749,
+      "learning_rate": 2e-05,
+      "loss": 0.03546806,
+      "step": 13611
+    },
+    {
+      "epoch": 27.224,
+      "grad_norm": 1.3775697946548462,
+      "learning_rate": 2e-05,
+      "loss": 0.0483359,
+      "step": 13612
+    },
+    {
+      "epoch": 27.226,
+      "grad_norm": 1.3804492950439453,
+      "learning_rate": 2e-05,
+      "loss": 0.05687822,
+      "step": 13613
+    },
+    {
+      "epoch": 27.228,
+      "grad_norm": 1.8396795988082886,
+      "learning_rate": 2e-05,
+      "loss": 0.04156945,
+      "step": 13614
+    },
+    {
+      "epoch": 27.23,
+      "grad_norm": 1.3716535568237305,
+      "learning_rate": 2e-05,
+      "loss": 0.05421146,
+      "step": 13615
+    },
+    {
+      "epoch": 27.232,
+      "grad_norm": 1.2649599313735962,
+      "learning_rate": 2e-05,
+      "loss": 0.04328211,
+      "step": 13616
+    },
+    {
+      "epoch": 27.234,
+      "grad_norm": 1.3094875812530518,
+      "learning_rate": 2e-05,
+      "loss": 0.04095453,
+      "step": 13617
+    },
+    {
+      "epoch": 27.236,
+      "grad_norm": 1.3521350622177124,
+      "learning_rate": 2e-05,
+      "loss": 0.0435282,
+      "step": 13618
+    },
+    {
+      "epoch": 27.238,
+      "grad_norm": 0.901831865310669,
+      "learning_rate": 2e-05,
+      "loss": 0.02451799,
+      "step": 13619
+    },
+    {
+      "epoch": 27.24,
+      "grad_norm": 1.344462275505066,
+      "learning_rate": 2e-05,
+      "loss": 0.06785841,
+      "step": 13620
+    },
+    {
+      "epoch": 27.242,
+      "grad_norm": 1.497972846031189,
+      "learning_rate": 2e-05,
+      "loss": 0.04425344,
+      "step": 13621
+    },
+    {
+      "epoch": 27.244,
+      "grad_norm": 0.9255624413490295,
+      "learning_rate": 2e-05,
+      "loss": 0.02206445,
+      "step": 13622
+    },
+    {
+      "epoch": 27.246,
+      "grad_norm": 0.988325834274292,
+      "learning_rate": 2e-05,
+      "loss": 0.02141338,
+      "step": 13623
+    },
+    {
+      "epoch": 27.248,
+      "grad_norm": 1.9612500667572021,
+      "learning_rate": 2e-05,
+      "loss": 0.03873473,
+      "step": 13624
+    },
+    {
+      "epoch": 27.25,
+      "grad_norm": 2.3645355701446533,
+      "learning_rate": 2e-05,
+      "loss": 0.0586706,
+      "step": 13625
+    },
+    {
+      "epoch": 27.252,
+      "grad_norm": 1.6442853212356567,
+      "learning_rate": 2e-05,
+      "loss": 0.04403327,
+      "step": 13626
+    },
+    {
+      "epoch": 27.254,
+      "grad_norm": 1.2499282360076904,
+      "learning_rate": 2e-05,
+      "loss": 0.05442981,
+      "step": 13627
+    },
+    {
+      "epoch": 27.256,
+      "grad_norm": 1.122299313545227,
+      "learning_rate": 2e-05,
+      "loss": 0.04064956,
+      "step": 13628
+    },
+    {
+      "epoch": 27.258,
+      "grad_norm": 1.537174105644226,
+      "learning_rate": 2e-05,
+      "loss": 0.04302373,
+      "step": 13629
+    },
+    {
+      "epoch": 27.26,
+      "grad_norm": 1.0611103773117065,
+      "learning_rate": 2e-05,
+      "loss": 0.03115848,
+      "step": 13630
+    },
+    {
+      "epoch": 27.262,
+      "grad_norm": 1.2895267009735107,
+      "learning_rate": 2e-05,
+      "loss": 0.03978854,
+      "step": 13631
+    },
+    {
+      "epoch": 27.264,
+      "grad_norm": 1.8378465175628662,
+      "learning_rate": 2e-05,
+      "loss": 0.06035455,
+      "step": 13632
+    },
+    {
+      "epoch": 27.266,
+      "grad_norm": 0.9278236031532288,
+      "learning_rate": 2e-05,
+      "loss": 0.0275078,
+      "step": 13633
+    },
+    {
+      "epoch": 27.268,
+      "grad_norm": 1.2331749200820923,
+      "learning_rate": 2e-05,
+      "loss": 0.03504308,
+      "step": 13634
+    },
+    {
+      "epoch": 27.27,
+      "grad_norm": 1.629676103591919,
+      "learning_rate": 2e-05,
+      "loss": 0.0618648,
+      "step": 13635
+    },
+    {
+      "epoch": 27.272,
+      "grad_norm": 1.405290961265564,
+      "learning_rate": 2e-05,
+      "loss": 0.0505509,
+      "step": 13636
+    },
+    {
+      "epoch": 27.274,
+      "grad_norm": 1.1765658855438232,
+      "learning_rate": 2e-05,
+      "loss": 0.05041191,
+      "step": 13637
+    },
+    {
+      "epoch": 27.276,
+      "grad_norm": 1.735039472579956,
+      "learning_rate": 2e-05,
+      "loss": 0.02883138,
+      "step": 13638
+    },
+    {
+      "epoch": 27.278,
+      "grad_norm": 0.9482174515724182,
+      "learning_rate": 2e-05,
+      "loss": 0.02577251,
+      "step": 13639
+    },
+    {
+      "epoch": 27.28,
+      "grad_norm": 1.1373050212860107,
+      "learning_rate": 2e-05,
+      "loss": 0.04171196,
+      "step": 13640
+    },
+    {
+      "epoch": 27.282,
+      "grad_norm": 2.016035795211792,
+      "learning_rate": 2e-05,
+      "loss": 0.0531563,
+      "step": 13641
+    },
+    {
+      "epoch": 27.284,
+      "grad_norm": 1.1706429719924927,
+      "learning_rate": 2e-05,
+      "loss": 0.04091615,
+      "step": 13642
+    },
+    {
+      "epoch": 27.286,
+      "grad_norm": 0.9347245693206787,
+      "learning_rate": 2e-05,
+      "loss": 0.03005055,
+      "step": 13643
+    },
+    {
+      "epoch": 27.288,
+      "grad_norm": 1.119093418121338,
+      "learning_rate": 2e-05,
+      "loss": 0.05457556,
+      "step": 13644
+    },
+    {
+      "epoch": 27.29,
+      "grad_norm": 1.3505531549453735,
+      "learning_rate": 2e-05,
+      "loss": 0.03677881,
+      "step": 13645
+    },
+    {
+      "epoch": 27.292,
+      "grad_norm": 1.270559549331665,
+      "learning_rate": 2e-05,
+      "loss": 0.03553203,
+      "step": 13646
+    },
+    {
+      "epoch": 27.294,
+      "grad_norm": 1.016276240348816,
+      "learning_rate": 2e-05,
+      "loss": 0.02860792,
+      "step": 13647
+    },
+    {
+      "epoch": 27.296,
+      "grad_norm": 2.6578431129455566,
+      "learning_rate": 2e-05,
+      "loss": 0.0371431,
+      "step": 13648
+    },
+    {
+      "epoch": 27.298,
+      "grad_norm": 1.0674734115600586,
+      "learning_rate": 2e-05,
+      "loss": 0.03743986,
+      "step": 13649
+    },
+    {
+      "epoch": 27.3,
+      "grad_norm": 0.987488865852356,
+      "learning_rate": 2e-05,
+      "loss": 0.02790395,
+      "step": 13650
+    },
+    {
+      "epoch": 27.302,
+      "grad_norm": 1.2899256944656372,
+      "learning_rate": 2e-05,
+      "loss": 0.05075889,
+      "step": 13651
+    },
+    {
+      "epoch": 27.304,
+      "grad_norm": 1.9835212230682373,
+      "learning_rate": 2e-05,
+      "loss": 0.05801808,
+      "step": 13652
+    },
+    {
+      "epoch": 27.306,
+      "grad_norm": 1.7673009634017944,
+      "learning_rate": 2e-05,
+      "loss": 0.05642558,
+      "step": 13653
+    },
+    {
+      "epoch": 27.308,
+      "grad_norm": 1.229543924331665,
+      "learning_rate": 2e-05,
+      "loss": 0.04363228,
+      "step": 13654
+    },
+    {
+      "epoch": 27.31,
+      "grad_norm": 1.8748433589935303,
+      "learning_rate": 2e-05,
+      "loss": 0.04515584,
+      "step": 13655
+    },
+    {
+      "epoch": 27.312,
+      "grad_norm": 1.0442911386489868,
+      "learning_rate": 2e-05,
+      "loss": 0.03930074,
+      "step": 13656
+    },
+    {
+      "epoch": 27.314,
+      "grad_norm": 1.443469524383545,
+      "learning_rate": 2e-05,
+      "loss": 0.04800137,
+      "step": 13657
+    },
+    {
+      "epoch": 27.316,
+      "grad_norm": 1.3523346185684204,
+      "learning_rate": 2e-05,
+      "loss": 0.04009256,
+      "step": 13658
+    },
+    {
+      "epoch": 27.318,
+      "grad_norm": 1.8447299003601074,
+      "learning_rate": 2e-05,
+      "loss": 0.0515473,
+      "step": 13659
+    },
+    {
+      "epoch": 27.32,
+      "grad_norm": 1.3202327489852905,
+      "learning_rate": 2e-05,
+      "loss": 0.05036566,
+      "step": 13660
+    },
+    {
+      "epoch": 27.322,
+      "grad_norm": 1.1027467250823975,
+      "learning_rate": 2e-05,
+      "loss": 0.04486959,
+      "step": 13661
+    },
+    {
+      "epoch": 27.324,
+      "grad_norm": 1.2013741731643677,
+      "learning_rate": 2e-05,
+      "loss": 0.04390001,
+      "step": 13662
+    },
+    {
+      "epoch": 27.326,
+      "grad_norm": 1.7813587188720703,
+      "learning_rate": 2e-05,
+      "loss": 0.03983445,
+      "step": 13663
+    },
+    {
+      "epoch": 27.328,
+      "grad_norm": 2.0535483360290527,
+      "learning_rate": 2e-05,
+      "loss": 0.0440939,
+      "step": 13664
+    },
+    {
+      "epoch": 27.33,
+      "grad_norm": 1.0619609355926514,
+      "learning_rate": 2e-05,
+      "loss": 0.03448323,
+      "step": 13665
+    },
+    {
+      "epoch": 27.332,
+      "grad_norm": 1.0306423902511597,
+      "learning_rate": 2e-05,
+      "loss": 0.03753592,
+      "step": 13666
+    },
+    {
+      "epoch": 27.334,
+      "grad_norm": 1.2383522987365723,
+      "learning_rate": 2e-05,
+      "loss": 0.03938786,
+      "step": 13667
+    },
+    {
+      "epoch": 27.336,
+      "grad_norm": 1.9096163511276245,
+      "learning_rate": 2e-05,
+      "loss": 0.07150873,
+      "step": 13668
+    },
+    {
+      "epoch": 27.338,
+      "grad_norm": 1.201596975326538,
+      "learning_rate": 2e-05,
+      "loss": 0.03716744,
+      "step": 13669
+    },
+    {
+      "epoch": 27.34,
+      "grad_norm": 1.5398945808410645,
+      "learning_rate": 2e-05,
+      "loss": 0.0461993,
+      "step": 13670
+    },
+    {
+      "epoch": 27.342,
+      "grad_norm": 1.073887586593628,
+      "learning_rate": 2e-05,
+      "loss": 0.0333119,
+      "step": 13671
+    },
+    {
+      "epoch": 27.344,
+      "grad_norm": 1.5942431688308716,
+      "learning_rate": 2e-05,
+      "loss": 0.0529358,
+      "step": 13672
+    },
+    {
+      "epoch": 27.346,
+      "grad_norm": 1.545664668083191,
+      "learning_rate": 2e-05,
+      "loss": 0.03542736,
+      "step": 13673
+    },
+    {
+      "epoch": 27.348,
+      "grad_norm": 1.5381520986557007,
+      "learning_rate": 2e-05,
+      "loss": 0.03843318,
+      "step": 13674
+    },
+    {
+      "epoch": 27.35,
+      "grad_norm": 1.3689863681793213,
+      "learning_rate": 2e-05,
+      "loss": 0.02982055,
+      "step": 13675
+    },
+    {
+      "epoch": 27.352,
+      "grad_norm": 1.2728350162506104,
+      "learning_rate": 2e-05,
+      "loss": 0.05316496,
+      "step": 13676
+    },
+    {
+      "epoch": 27.354,
+      "grad_norm": 0.9394140839576721,
+      "learning_rate": 2e-05,
+      "loss": 0.02646692,
+      "step": 13677
+    },
+    {
+      "epoch": 27.356,
+      "grad_norm": 0.9711006879806519,
+      "learning_rate": 2e-05,
+      "loss": 0.02917571,
+      "step": 13678
+    },
+    {
+      "epoch": 27.358,
+      "grad_norm": 1.5109331607818604,
+      "learning_rate": 2e-05,
+      "loss": 0.03243531,
+      "step": 13679
+    },
+    {
+      "epoch": 27.36,
+      "grad_norm": 1.113895297050476,
+      "learning_rate": 2e-05,
+      "loss": 0.03930811,
+      "step": 13680
+    },
+    {
+      "epoch": 27.362,
+      "grad_norm": 1.5176002979278564,
+      "learning_rate": 2e-05,
+      "loss": 0.06130125,
+      "step": 13681
+    },
+    {
+      "epoch": 27.364,
+      "grad_norm": 0.8235756754875183,
+      "learning_rate": 2e-05,
+      "loss": 0.02695323,
+      "step": 13682
+    },
+    {
+      "epoch": 27.366,
+      "grad_norm": 1.0173813104629517,
+      "learning_rate": 2e-05,
+      "loss": 0.03705809,
+      "step": 13683
+    },
+    {
+      "epoch": 27.368,
+      "grad_norm": 1.7335045337677002,
+      "learning_rate": 2e-05,
+      "loss": 0.04375722,
+      "step": 13684
+    },
+    {
+      "epoch": 27.37,
+      "grad_norm": 1.392207145690918,
+      "learning_rate": 2e-05,
+      "loss": 0.03899303,
+      "step": 13685
+    },
+    {
+      "epoch": 27.372,
+      "grad_norm": 1.5254584550857544,
+      "learning_rate": 2e-05,
+      "loss": 0.05612728,
+      "step": 13686
+    },
+    {
+      "epoch": 27.374,
+      "grad_norm": 1.4681589603424072,
+      "learning_rate": 2e-05,
+      "loss": 0.05056222,
+      "step": 13687
+    },
+    {
+      "epoch": 27.376,
+      "grad_norm": 1.21063232421875,
+      "learning_rate": 2e-05,
+      "loss": 0.03048548,
+      "step": 13688
+    },
+    {
+      "epoch": 27.378,
+      "grad_norm": 1.43213951587677,
+      "learning_rate": 2e-05,
+      "loss": 0.06458762,
+      "step": 13689
+    },
+    {
+      "epoch": 27.38,
+      "grad_norm": 2.742194652557373,
+      "learning_rate": 2e-05,
+      "loss": 0.05234894,
+      "step": 13690
+    },
+    {
+      "epoch": 27.382,
+      "grad_norm": 2.0506792068481445,
+      "learning_rate": 2e-05,
+      "loss": 0.05463105,
+      "step": 13691
+    },
+    {
+      "epoch": 27.384,
+      "grad_norm": 1.5875942707061768,
+      "learning_rate": 2e-05,
+      "loss": 0.03925296,
+      "step": 13692
+    },
+    {
+      "epoch": 27.386,
+      "grad_norm": 1.5243009328842163,
+      "learning_rate": 2e-05,
+      "loss": 0.05017775,
+      "step": 13693
+    },
+    {
+      "epoch": 27.388,
+      "grad_norm": 1.4063950777053833,
+      "learning_rate": 2e-05,
+      "loss": 0.05697461,
+      "step": 13694
+    },
+    {
+      "epoch": 27.39,
+      "grad_norm": 2.2699193954467773,
+      "learning_rate": 2e-05,
+      "loss": 0.03670984,
+      "step": 13695
+    },
+    {
+      "epoch": 27.392,
+      "grad_norm": 2.1173882484436035,
+      "learning_rate": 2e-05,
+      "loss": 0.05186939,
+      "step": 13696
+    },
+    {
+      "epoch": 27.394,
+      "grad_norm": 1.2842679023742676,
+      "learning_rate": 2e-05,
+      "loss": 0.06256726,
+      "step": 13697
+    },
+    {
+      "epoch": 27.396,
+      "grad_norm": 2.2587108612060547,
+      "learning_rate": 2e-05,
+      "loss": 0.06481649,
+      "step": 13698
+    },
+    {
+      "epoch": 27.398,
+      "grad_norm": 1.8947147130966187,
+      "learning_rate": 2e-05,
+      "loss": 0.04356439,
+      "step": 13699
+    },
+    {
+      "epoch": 27.4,
+      "grad_norm": 0.9812667369842529,
+      "learning_rate": 2e-05,
+      "loss": 0.02878316,
+      "step": 13700
+    },
+    {
+      "epoch": 27.402,
+      "grad_norm": 1.1468307971954346,
+      "learning_rate": 2e-05,
+      "loss": 0.04130252,
+      "step": 13701
+    },
+    {
+      "epoch": 27.404,
+      "grad_norm": 1.0459030866622925,
+      "learning_rate": 2e-05,
+      "loss": 0.03767227,
+      "step": 13702
+    },
+    {
+      "epoch": 27.406,
+      "grad_norm": 1.448286771774292,
+      "learning_rate": 2e-05,
+      "loss": 0.04471764,
+      "step": 13703
+    },
+    {
+      "epoch": 27.408,
+      "grad_norm": 1.1528844833374023,
+      "learning_rate": 2e-05,
+      "loss": 0.02849596,
+      "step": 13704
+    },
+    {
+      "epoch": 27.41,
+      "grad_norm": 1.6160606145858765,
+      "learning_rate": 2e-05,
+      "loss": 0.02551612,
+      "step": 13705
+    },
+    {
+      "epoch": 27.412,
+      "grad_norm": 2.2915830612182617,
+      "learning_rate": 2e-05,
+      "loss": 0.05899898,
+      "step": 13706
+    },
+    {
+      "epoch": 27.414,
+      "grad_norm": 1.510726809501648,
+      "learning_rate": 2e-05,
+      "loss": 0.05045645,
+      "step": 13707
+    },
+    {
+      "epoch": 27.416,
+      "grad_norm": 3.0780930519104004,
+      "learning_rate": 2e-05,
+      "loss": 0.06752353,
+      "step": 13708
+    },
+    {
+      "epoch": 27.418,
+      "grad_norm": 1.264102816581726,
+      "learning_rate": 2e-05,
+      "loss": 0.04427083,
+      "step": 13709
+    },
+    {
+      "epoch": 27.42,
+      "grad_norm": 1.073390007019043,
+      "learning_rate": 2e-05,
+      "loss": 0.04122963,
+      "step": 13710
+    },
+    {
+      "epoch": 27.422,
+      "grad_norm": 1.1260555982589722,
+      "learning_rate": 2e-05,
+      "loss": 0.04339466,
+      "step": 13711
+    },
+    {
+      "epoch": 27.424,
+      "grad_norm": 1.236524224281311,
+      "learning_rate": 2e-05,
+      "loss": 0.03498268,
+      "step": 13712
+    },
+    {
+      "epoch": 27.426,
+      "grad_norm": 1.594119906425476,
+      "learning_rate": 2e-05,
+      "loss": 0.04375029,
+      "step": 13713
+    },
+    {
+      "epoch": 27.428,
+      "grad_norm": 1.2735551595687866,
+      "learning_rate": 2e-05,
+      "loss": 0.03774681,
+      "step": 13714
+    },
+    {
+      "epoch": 27.43,
+      "grad_norm": 1.059618592262268,
+      "learning_rate": 2e-05,
+      "loss": 0.03130772,
+      "step": 13715
+    },
+    {
+      "epoch": 27.432,
+      "grad_norm": 1.5274853706359863,
+      "learning_rate": 2e-05,
+      "loss": 0.05236381,
+      "step": 13716
+    },
+    {
+      "epoch": 27.434,
+      "grad_norm": 0.9794763326644897,
+      "learning_rate": 2e-05,
+      "loss": 0.03129626,
+      "step": 13717
+    },
+    {
+      "epoch": 27.436,
+      "grad_norm": 1.2296133041381836,
+      "learning_rate": 2e-05,
+      "loss": 0.04068862,
+      "step": 13718
+    },
+    {
+      "epoch": 27.438,
+      "grad_norm": 1.5292551517486572,
+      "learning_rate": 2e-05,
+      "loss": 0.04712692,
+      "step": 13719
+    },
+    {
+      "epoch": 27.44,
+      "grad_norm": 1.157141089439392,
+      "learning_rate": 2e-05,
+      "loss": 0.05303669,
+      "step": 13720
+    },
+    {
+      "epoch": 27.442,
+      "grad_norm": 1.2794381380081177,
+      "learning_rate": 2e-05,
+      "loss": 0.05657168,
+      "step": 13721
+    },
+    {
+      "epoch": 27.444,
+      "grad_norm": 1.1883527040481567,
+      "learning_rate": 2e-05,
+      "loss": 0.05889457,
+      "step": 13722
+    },
+    {
+      "epoch": 27.446,
+      "grad_norm": 1.260737419128418,
+      "learning_rate": 2e-05,
+      "loss": 0.04855951,
+      "step": 13723
+    },
+    {
+      "epoch": 27.448,
+      "grad_norm": 1.1919201612472534,
+      "learning_rate": 2e-05,
+      "loss": 0.02818074,
+      "step": 13724
+    },
+    {
+      "epoch": 27.45,
+      "grad_norm": 1.3503472805023193,
+      "learning_rate": 2e-05,
+      "loss": 0.05925683,
+      "step": 13725
+    },
+    {
+      "epoch": 27.452,
+      "grad_norm": 1.102662205696106,
+      "learning_rate": 2e-05,
+      "loss": 0.03601834,
+      "step": 13726
+    },
+    {
+      "epoch": 27.454,
+      "grad_norm": 0.9547258019447327,
+      "learning_rate": 2e-05,
+      "loss": 0.0299718,
+      "step": 13727
+    },
+    {
+      "epoch": 27.456,
+      "grad_norm": 1.2688747644424438,
+      "learning_rate": 2e-05,
+      "loss": 0.04194263,
+      "step": 13728
+    },
+    {
+      "epoch": 27.458,
+      "grad_norm": 2.936462640762329,
+      "learning_rate": 2e-05,
+      "loss": 0.0673967,
+      "step": 13729
+    },
+    {
+      "epoch": 27.46,
+      "grad_norm": 1.7205671072006226,
+      "learning_rate": 2e-05,
+      "loss": 0.06203447,
+      "step": 13730
+    },
+    {
+      "epoch": 27.462,
+      "grad_norm": 1.4902185201644897,
+      "learning_rate": 2e-05,
+      "loss": 0.04702317,
+      "step": 13731
+    },
+    {
+      "epoch": 27.464,
+      "grad_norm": 1.2047202587127686,
+      "learning_rate": 2e-05,
+      "loss": 0.04467279,
+      "step": 13732
+    },
+    {
+      "epoch": 27.466,
+      "grad_norm": 1.2293957471847534,
+      "learning_rate": 2e-05,
+      "loss": 0.03134837,
+      "step": 13733
+    },
+    {
+      "epoch": 27.468,
+      "grad_norm": 1.4293403625488281,
+      "learning_rate": 2e-05,
+      "loss": 0.03927672,
+      "step": 13734
+    },
+    {
+      "epoch": 27.47,
+      "grad_norm": 1.2355231046676636,
+      "learning_rate": 2e-05,
+      "loss": 0.03786084,
+      "step": 13735
+    },
+    {
+      "epoch": 27.472,
+      "grad_norm": 1.4935061931610107,
+      "learning_rate": 2e-05,
+      "loss": 0.05582435,
+      "step": 13736
+    },
+    {
+      "epoch": 27.474,
+      "grad_norm": 1.0883551836013794,
+      "learning_rate": 2e-05,
+      "loss": 0.05210774,
+      "step": 13737
+    },
+    {
+      "epoch": 27.476,
+      "grad_norm": 1.4402879476547241,
+      "learning_rate": 2e-05,
+      "loss": 0.0516119,
+      "step": 13738
+    },
+    {
+      "epoch": 27.478,
+      "grad_norm": 1.3633337020874023,
+      "learning_rate": 2e-05,
+      "loss": 0.05501125,
+      "step": 13739
+    },
+    {
+      "epoch": 27.48,
+      "grad_norm": 1.403083324432373,
+      "learning_rate": 2e-05,
+      "loss": 0.07498935,
+      "step": 13740
+    },
+    {
+      "epoch": 27.482,
+      "grad_norm": 1.22272527217865,
+      "learning_rate": 2e-05,
+      "loss": 0.04192954,
+      "step": 13741
+    },
+    {
+      "epoch": 27.484,
+      "grad_norm": 0.9646196961402893,
+      "learning_rate": 2e-05,
+      "loss": 0.03863492,
+      "step": 13742
+    },
+    {
+      "epoch": 27.486,
+      "grad_norm": 1.4368653297424316,
+      "learning_rate": 2e-05,
+      "loss": 0.05678754,
+      "step": 13743
+    },
+    {
+      "epoch": 27.488,
+      "grad_norm": 1.0323776006698608,
+      "learning_rate": 2e-05,
+      "loss": 0.03164582,
+      "step": 13744
+    },
+    {
+      "epoch": 27.49,
+      "grad_norm": 1.335216760635376,
+      "learning_rate": 2e-05,
+      "loss": 0.04490231,
+      "step": 13745
+    },
+    {
+      "epoch": 27.492,
+      "grad_norm": 1.462903618812561,
+      "learning_rate": 2e-05,
+      "loss": 0.05864259,
+      "step": 13746
+    },
+    {
+      "epoch": 27.494,
+      "grad_norm": 1.1962298154830933,
+      "learning_rate": 2e-05,
+      "loss": 0.03484888,
+      "step": 13747
+    },
+    {
+      "epoch": 27.496,
+      "grad_norm": 1.8432676792144775,
+      "learning_rate": 2e-05,
+      "loss": 0.04844757,
+      "step": 13748
+    },
+    {
+      "epoch": 27.498,
+      "grad_norm": 1.9054256677627563,
+      "learning_rate": 2e-05,
+      "loss": 0.03610344,
+      "step": 13749
+    },
+    {
+      "epoch": 27.5,
+      "grad_norm": 1.8628199100494385,
+      "learning_rate": 2e-05,
+      "loss": 0.03124732,
+      "step": 13750
+    },
+    {
+      "epoch": 27.502,
+      "grad_norm": 1.300698161125183,
+      "learning_rate": 2e-05,
+      "loss": 0.03852872,
+      "step": 13751
+    },
+    {
+      "epoch": 27.504,
+      "grad_norm": 1.4470487833023071,
+      "learning_rate": 2e-05,
+      "loss": 0.03881678,
+      "step": 13752
+    },
+    {
+      "epoch": 27.506,
+      "grad_norm": 1.4798719882965088,
+      "learning_rate": 2e-05,
+      "loss": 0.0504625,
+      "step": 13753
+    },
+    {
+      "epoch": 27.508,
+      "grad_norm": 0.9879271388053894,
+      "learning_rate": 2e-05,
+      "loss": 0.0390942,
+      "step": 13754
+    },
+    {
+      "epoch": 27.51,
+      "grad_norm": 1.9289140701293945,
+      "learning_rate": 2e-05,
+      "loss": 0.04204384,
+      "step": 13755
+    },
+    {
+      "epoch": 27.512,
+      "grad_norm": 2.8271074295043945,
+      "learning_rate": 2e-05,
+      "loss": 0.0529681,
+      "step": 13756
+    },
+    {
+      "epoch": 27.514,
+      "grad_norm": 4.042783260345459,
+      "learning_rate": 2e-05,
+      "loss": 0.04220014,
+      "step": 13757
+    },
+    {
+      "epoch": 27.516,
+      "grad_norm": 1.2601898908615112,
+      "learning_rate": 2e-05,
+      "loss": 0.03472359,
+      "step": 13758
+    },
+    {
+      "epoch": 27.518,
+      "grad_norm": 1.2109252214431763,
+      "learning_rate": 2e-05,
+      "loss": 0.04545831,
+      "step": 13759
+    },
+    {
+      "epoch": 27.52,
+      "grad_norm": 1.168678641319275,
+      "learning_rate": 2e-05,
+      "loss": 0.03518002,
+      "step": 13760
+    },
+    {
+      "epoch": 27.522,
+      "grad_norm": 1.7536158561706543,
+      "learning_rate": 2e-05,
+      "loss": 0.05130105,
+      "step": 13761
+    },
+    {
+      "epoch": 27.524,
+      "grad_norm": 1.5836191177368164,
+      "learning_rate": 2e-05,
+      "loss": 0.03332604,
+      "step": 13762
+    },
+    {
+      "epoch": 27.526,
+      "grad_norm": 2.656456232070923,
+      "learning_rate": 2e-05,
+      "loss": 0.05432518,
+      "step": 13763
+    },
+    {
+      "epoch": 27.528,
+      "grad_norm": 2.7838237285614014,
+      "learning_rate": 2e-05,
+      "loss": 0.03519111,
+      "step": 13764
+    },
+    {
+      "epoch": 27.53,
+      "grad_norm": 2.392075777053833,
+      "learning_rate": 2e-05,
+      "loss": 0.04325394,
+      "step": 13765
+    },
+    {
+      "epoch": 27.532,
+      "grad_norm": 1.0115920305252075,
+      "learning_rate": 2e-05,
+      "loss": 0.03398691,
+      "step": 13766
+    },
+    {
+      "epoch": 27.534,
+      "grad_norm": 1.0626614093780518,
+      "learning_rate": 2e-05,
+      "loss": 0.03680531,
+      "step": 13767
+    },
+    {
+      "epoch": 27.536,
+      "grad_norm": 1.1814494132995605,
+      "learning_rate": 2e-05,
+      "loss": 0.0340856,
+      "step": 13768
+    },
+    {
+      "epoch": 27.538,
+      "grad_norm": 1.5384715795516968,
+      "learning_rate": 2e-05,
+      "loss": 0.04477349,
+      "step": 13769
+    },
+    {
+      "epoch": 27.54,
+      "grad_norm": 1.392616868019104,
+      "learning_rate": 2e-05,
+      "loss": 0.03679165,
+      "step": 13770
+    },
+    {
+      "epoch": 27.542,
+      "grad_norm": 2.3772451877593994,
+      "learning_rate": 2e-05,
+      "loss": 0.04677385,
+      "step": 13771
+    },
+    {
+      "epoch": 27.544,
+      "grad_norm": 2.6153879165649414,
+      "learning_rate": 2e-05,
+      "loss": 0.05070829,
+      "step": 13772
+    },
+    {
+      "epoch": 27.546,
+      "grad_norm": 1.1042912006378174,
+      "learning_rate": 2e-05,
+      "loss": 0.03895772,
+      "step": 13773
+    },
+    {
+      "epoch": 27.548000000000002,
+      "grad_norm": 2.243035316467285,
+      "learning_rate": 2e-05,
+      "loss": 0.04948647,
+      "step": 13774
+    },
+    {
+      "epoch": 27.55,
+      "grad_norm": 1.1408405303955078,
+      "learning_rate": 2e-05,
+      "loss": 0.03680218,
+      "step": 13775
+    },
+    {
+      "epoch": 27.552,
+      "grad_norm": 1.123650074005127,
+      "learning_rate": 2e-05,
+      "loss": 0.04027671,
+      "step": 13776
+    },
+    {
+      "epoch": 27.554,
+      "grad_norm": 1.9606479406356812,
+      "learning_rate": 2e-05,
+      "loss": 0.05564614,
+      "step": 13777
+    },
+    {
+      "epoch": 27.556,
+      "grad_norm": 1.310119867324829,
+      "learning_rate": 2e-05,
+      "loss": 0.03973632,
+      "step": 13778
+    },
+    {
+      "epoch": 27.558,
+      "grad_norm": 1.5927740335464478,
+      "learning_rate": 2e-05,
+      "loss": 0.04547199,
+      "step": 13779
+    },
+    {
+      "epoch": 27.56,
+      "grad_norm": 1.2601191997528076,
+      "learning_rate": 2e-05,
+      "loss": 0.05317003,
+      "step": 13780
+    },
+    {
+      "epoch": 27.562,
+      "grad_norm": 1.647443175315857,
+      "learning_rate": 2e-05,
+      "loss": 0.0479614,
+      "step": 13781
+    },
+    {
+      "epoch": 27.564,
+      "grad_norm": 1.0511468648910522,
+      "learning_rate": 2e-05,
+      "loss": 0.03439099,
+      "step": 13782
+    },
+    {
+      "epoch": 27.566,
+      "grad_norm": 1.2322078943252563,
+      "learning_rate": 2e-05,
+      "loss": 0.04574615,
+      "step": 13783
+    },
+    {
+      "epoch": 27.568,
+      "grad_norm": 1.0778727531433105,
+      "learning_rate": 2e-05,
+      "loss": 0.03921001,
+      "step": 13784
+    },
+    {
+      "epoch": 27.57,
+      "grad_norm": 1.158125877380371,
+      "learning_rate": 2e-05,
+      "loss": 0.0469446,
+      "step": 13785
+    },
+    {
+      "epoch": 27.572,
+      "grad_norm": 1.4670352935791016,
+      "learning_rate": 2e-05,
+      "loss": 0.04167302,
+      "step": 13786
+    },
+    {
+      "epoch": 27.574,
+      "grad_norm": 2.666126251220703,
+      "learning_rate": 2e-05,
+      "loss": 0.05114643,
+      "step": 13787
+    },
+    {
+      "epoch": 27.576,
+      "grad_norm": 1.1133148670196533,
+      "learning_rate": 2e-05,
+      "loss": 0.03147897,
+      "step": 13788
+    },
+    {
+      "epoch": 27.578,
+      "grad_norm": 1.703502893447876,
+      "learning_rate": 2e-05,
+      "loss": 0.03852395,
+      "step": 13789
+    },
+    {
+      "epoch": 27.58,
+      "grad_norm": 1.3457422256469727,
+      "learning_rate": 2e-05,
+      "loss": 0.0443656,
+      "step": 13790
+    },
+    {
+      "epoch": 27.582,
+      "grad_norm": 1.0474896430969238,
+      "learning_rate": 2e-05,
+      "loss": 0.02956739,
+      "step": 13791
+    },
+    {
+      "epoch": 27.584,
+      "grad_norm": 1.2296174764633179,
+      "learning_rate": 2e-05,
+      "loss": 0.03405485,
+      "step": 13792
+    },
+    {
+      "epoch": 27.586,
+      "grad_norm": 0.9992415308952332,
+      "learning_rate": 2e-05,
+      "loss": 0.02881895,
+      "step": 13793
+    },
+    {
+      "epoch": 27.588,
+      "grad_norm": 1.370805263519287,
+      "learning_rate": 2e-05,
+      "loss": 0.0559724,
+      "step": 13794
+    },
+    {
+      "epoch": 27.59,
+      "grad_norm": 2.4789819717407227,
+      "learning_rate": 2e-05,
+      "loss": 0.04826446,
+      "step": 13795
+    },
+    {
+      "epoch": 27.592,
+      "grad_norm": 2.2076499462127686,
+      "learning_rate": 2e-05,
+      "loss": 0.05645463,
+      "step": 13796
+    },
+    {
+      "epoch": 27.594,
+      "grad_norm": 1.5076375007629395,
+      "learning_rate": 2e-05,
+      "loss": 0.04553786,
+      "step": 13797
+    },
+    {
+      "epoch": 27.596,
+      "grad_norm": 1.0695239305496216,
+      "learning_rate": 2e-05,
+      "loss": 0.0458618,
+      "step": 13798
+    },
+    {
+      "epoch": 27.598,
+      "grad_norm": 1.2451701164245605,
+      "learning_rate": 2e-05,
+      "loss": 0.04357987,
+      "step": 13799
+    },
+    {
+      "epoch": 27.6,
+      "grad_norm": 1.1719753742218018,
+      "learning_rate": 2e-05,
+      "loss": 0.05280523,
+      "step": 13800
+    },
+    {
+      "epoch": 27.602,
+      "grad_norm": 1.9771161079406738,
+      "learning_rate": 2e-05,
+      "loss": 0.0490317,
+      "step": 13801
+    },
+    {
+      "epoch": 27.604,
+      "grad_norm": 1.2304954528808594,
+      "learning_rate": 2e-05,
+      "loss": 0.04574135,
+      "step": 13802
+    },
+    {
+      "epoch": 27.606,
+      "grad_norm": 1.237014651298523,
+      "learning_rate": 2e-05,
+      "loss": 0.03605729,
+      "step": 13803
+    },
+    {
+      "epoch": 27.608,
+      "grad_norm": 3.3194026947021484,
+      "learning_rate": 2e-05,
+      "loss": 0.05870149,
+      "step": 13804
+    },
+    {
+      "epoch": 27.61,
+      "grad_norm": 1.4011256694793701,
+      "learning_rate": 2e-05,
+      "loss": 0.03893548,
+      "step": 13805
+    },
+    {
+      "epoch": 27.612,
+      "grad_norm": 1.604568362236023,
+      "learning_rate": 2e-05,
+      "loss": 0.0501047,
+      "step": 13806
+    },
+    {
+      "epoch": 27.614,
+      "grad_norm": 1.0760208368301392,
+      "learning_rate": 2e-05,
+      "loss": 0.03730212,
+      "step": 13807
+    },
+    {
+      "epoch": 27.616,
+      "grad_norm": 1.54534912109375,
+      "learning_rate": 2e-05,
+      "loss": 0.05426643,
+      "step": 13808
+    },
+    {
+      "epoch": 27.618,
+      "grad_norm": 1.1083625555038452,
+      "learning_rate": 2e-05,
+      "loss": 0.03855055,
+      "step": 13809
+    },
+    {
+      "epoch": 27.62,
+      "grad_norm": 1.462074875831604,
+      "learning_rate": 2e-05,
+      "loss": 0.04273522,
+      "step": 13810
+    },
+    {
+      "epoch": 27.622,
+      "grad_norm": 2.044818878173828,
+      "learning_rate": 2e-05,
+      "loss": 0.06233047,
+      "step": 13811
+    },
+    {
+      "epoch": 27.624,
+      "grad_norm": 1.4091565608978271,
+      "learning_rate": 2e-05,
+      "loss": 0.04076818,
+      "step": 13812
+    },
+    {
+      "epoch": 27.626,
+      "grad_norm": 1.1645406484603882,
+      "learning_rate": 2e-05,
+      "loss": 0.0466489,
+      "step": 13813
+    },
+    {
+      "epoch": 27.628,
+      "grad_norm": 1.4516359567642212,
+      "learning_rate": 2e-05,
+      "loss": 0.04340706,
+      "step": 13814
+    },
+    {
+      "epoch": 27.63,
+      "grad_norm": 1.0174524784088135,
+      "learning_rate": 2e-05,
+      "loss": 0.02986944,
+      "step": 13815
+    },
+    {
+      "epoch": 27.632,
+      "grad_norm": 1.4557236433029175,
+      "learning_rate": 2e-05,
+      "loss": 0.04444972,
+      "step": 13816
+    },
+    {
+      "epoch": 27.634,
+      "grad_norm": 1.2507081031799316,
+      "learning_rate": 2e-05,
+      "loss": 0.0360889,
+      "step": 13817
+    },
+    {
+      "epoch": 27.636,
+      "grad_norm": 2.815284013748169,
+      "learning_rate": 2e-05,
+      "loss": 0.04525167,
+      "step": 13818
+    },
+    {
+      "epoch": 27.638,
+      "grad_norm": 0.9910860061645508,
+      "learning_rate": 2e-05,
+      "loss": 0.03063339,
+      "step": 13819
+    },
+    {
+      "epoch": 27.64,
+      "grad_norm": 2.006213426589966,
+      "learning_rate": 2e-05,
+      "loss": 0.04587614,
+      "step": 13820
+    },
+    {
+      "epoch": 27.642,
+      "grad_norm": 1.6014063358306885,
+      "learning_rate": 2e-05,
+      "loss": 0.05192234,
+      "step": 13821
+    },
+    {
+      "epoch": 27.644,
+      "grad_norm": 1.4815598726272583,
+      "learning_rate": 2e-05,
+      "loss": 0.05399399,
+      "step": 13822
+    },
+    {
+      "epoch": 27.646,
+      "grad_norm": 1.2255171537399292,
+      "learning_rate": 2e-05,
+      "loss": 0.04327129,
+      "step": 13823
+    },
+    {
+      "epoch": 27.648,
+      "grad_norm": 1.0979926586151123,
+      "learning_rate": 2e-05,
+      "loss": 0.03967486,
+      "step": 13824
+    },
+    {
+      "epoch": 27.65,
+      "grad_norm": 1.2818236351013184,
+      "learning_rate": 2e-05,
+      "loss": 0.05102768,
+      "step": 13825
+    },
+    {
+      "epoch": 27.652,
+      "grad_norm": 1.1726670265197754,
+      "learning_rate": 2e-05,
+      "loss": 0.04293338,
+      "step": 13826
+    },
+    {
+      "epoch": 27.654,
+      "grad_norm": 1.7466474771499634,
+      "learning_rate": 2e-05,
+      "loss": 0.05429044,
+      "step": 13827
+    },
+    {
+      "epoch": 27.656,
+      "grad_norm": 1.266982078552246,
+      "learning_rate": 2e-05,
+      "loss": 0.05010791,
+      "step": 13828
+    },
+    {
+      "epoch": 27.658,
+      "grad_norm": 1.0520493984222412,
+      "learning_rate": 2e-05,
+      "loss": 0.0411777,
+      "step": 13829
+    },
+    {
+      "epoch": 27.66,
+      "grad_norm": 1.2513072490692139,
+      "learning_rate": 2e-05,
+      "loss": 0.0348096,
+      "step": 13830
+    },
+    {
+      "epoch": 27.662,
+      "grad_norm": 0.8475068211555481,
+      "learning_rate": 2e-05,
+      "loss": 0.02205401,
+      "step": 13831
+    },
+    {
+      "epoch": 27.664,
+      "grad_norm": 1.2510443925857544,
+      "learning_rate": 2e-05,
+      "loss": 0.04548548,
+      "step": 13832
+    },
+    {
+      "epoch": 27.666,
+      "grad_norm": 1.2824064493179321,
+      "learning_rate": 2e-05,
+      "loss": 0.03442299,
+      "step": 13833
+    },
+    {
+      "epoch": 27.668,
+      "grad_norm": 1.1054766178131104,
+      "learning_rate": 2e-05,
+      "loss": 0.04257656,
+      "step": 13834
+    },
+    {
+      "epoch": 27.67,
+      "grad_norm": 1.853498935699463,
+      "learning_rate": 2e-05,
+      "loss": 0.05419736,
+      "step": 13835
+    },
+    {
+      "epoch": 27.672,
+      "grad_norm": 4.41054630279541,
+      "learning_rate": 2e-05,
+      "loss": 0.05251537,
+      "step": 13836
+    },
+    {
+      "epoch": 27.674,
+      "grad_norm": 1.2487452030181885,
+      "learning_rate": 2e-05,
+      "loss": 0.04494422,
+      "step": 13837
+    },
+    {
+      "epoch": 27.676,
+      "grad_norm": 1.9383426904678345,
+      "learning_rate": 2e-05,
+      "loss": 0.05443368,
+      "step": 13838
+    },
+    {
+      "epoch": 27.678,
+      "grad_norm": 1.7748433351516724,
+      "learning_rate": 2e-05,
+      "loss": 0.04427961,
+      "step": 13839
+    },
+    {
+      "epoch": 27.68,
+      "grad_norm": 0.8062875866889954,
+      "learning_rate": 2e-05,
+      "loss": 0.02282286,
+      "step": 13840
+    },
+    {
+      "epoch": 27.682,
+      "grad_norm": 2.103724479675293,
+      "learning_rate": 2e-05,
+      "loss": 0.05308349,
+      "step": 13841
+    },
+    {
+      "epoch": 27.684,
+      "grad_norm": 1.1036003828048706,
+      "learning_rate": 2e-05,
+      "loss": 0.0406925,
+      "step": 13842
+    },
+    {
+      "epoch": 27.686,
+      "grad_norm": 1.1960515975952148,
+      "learning_rate": 2e-05,
+      "loss": 0.04112234,
+      "step": 13843
+    },
+    {
+      "epoch": 27.688,
+      "grad_norm": 1.1153143644332886,
+      "learning_rate": 2e-05,
+      "loss": 0.02981465,
+      "step": 13844
+    },
+    {
+      "epoch": 27.69,
+      "grad_norm": 1.2569992542266846,
+      "learning_rate": 2e-05,
+      "loss": 0.0435634,
+      "step": 13845
+    },
+    {
+      "epoch": 27.692,
+      "grad_norm": 1.384782314300537,
+      "learning_rate": 2e-05,
+      "loss": 0.03404148,
+      "step": 13846
+    },
+    {
+      "epoch": 27.694,
+      "grad_norm": 1.6214118003845215,
+      "learning_rate": 2e-05,
+      "loss": 0.04686133,
+      "step": 13847
+    },
+    {
+      "epoch": 27.696,
+      "grad_norm": 1.5570800304412842,
+      "learning_rate": 2e-05,
+      "loss": 0.05804433,
+      "step": 13848
+    },
+    {
+      "epoch": 27.698,
+      "grad_norm": 1.5227315425872803,
+      "learning_rate": 2e-05,
+      "loss": 0.04167802,
+      "step": 13849
+    },
+    {
+      "epoch": 27.7,
+      "grad_norm": 1.578632116317749,
+      "learning_rate": 2e-05,
+      "loss": 0.04537342,
+      "step": 13850
+    },
+    {
+      "epoch": 27.701999999999998,
+      "grad_norm": 1.34732186794281,
+      "learning_rate": 2e-05,
+      "loss": 0.05119549,
+      "step": 13851
+    },
+    {
+      "epoch": 27.704,
+      "grad_norm": 2.011436939239502,
+      "learning_rate": 2e-05,
+      "loss": 0.08780254,
+      "step": 13852
+    },
+    {
+      "epoch": 27.706,
+      "grad_norm": 1.0843994617462158,
+      "learning_rate": 2e-05,
+      "loss": 0.03971418,
+      "step": 13853
+    },
+    {
+      "epoch": 27.708,
+      "grad_norm": 1.1688135862350464,
+      "learning_rate": 2e-05,
+      "loss": 0.04056138,
+      "step": 13854
+    },
+    {
+      "epoch": 27.71,
+      "grad_norm": 0.9493677616119385,
+      "learning_rate": 2e-05,
+      "loss": 0.03187769,
+      "step": 13855
+    },
+    {
+      "epoch": 27.712,
+      "grad_norm": 1.0012264251708984,
+      "learning_rate": 2e-05,
+      "loss": 0.04027501,
+      "step": 13856
+    },
+    {
+      "epoch": 27.714,
+      "grad_norm": 0.8952470421791077,
+      "learning_rate": 2e-05,
+      "loss": 0.03563171,
+      "step": 13857
+    },
+    {
+      "epoch": 27.716,
+      "grad_norm": 1.1415109634399414,
+      "learning_rate": 2e-05,
+      "loss": 0.03367986,
+      "step": 13858
+    },
+    {
+      "epoch": 27.718,
+      "grad_norm": 1.085681676864624,
+      "learning_rate": 2e-05,
+      "loss": 0.04318651,
+      "step": 13859
+    },
+    {
+      "epoch": 27.72,
+      "grad_norm": 1.0478479862213135,
+      "learning_rate": 2e-05,
+      "loss": 0.03802703,
+      "step": 13860
+    },
+    {
+      "epoch": 27.722,
+      "grad_norm": 1.5198477506637573,
+      "learning_rate": 2e-05,
+      "loss": 0.04132574,
+      "step": 13861
+    },
+    {
+      "epoch": 27.724,
+      "grad_norm": 1.0840024948120117,
+      "learning_rate": 2e-05,
+      "loss": 0.04059497,
+      "step": 13862
+    },
+    {
+      "epoch": 27.726,
+      "grad_norm": 1.027240514755249,
+      "learning_rate": 2e-05,
+      "loss": 0.03974399,
+      "step": 13863
+    },
+    {
+      "epoch": 27.728,
+      "grad_norm": 1.053519368171692,
+      "learning_rate": 2e-05,
+      "loss": 0.03394579,
+      "step": 13864
+    },
+    {
+      "epoch": 27.73,
+      "grad_norm": 1.8327863216400146,
+      "learning_rate": 2e-05,
+      "loss": 0.03920068,
+      "step": 13865
+    },
+    {
+      "epoch": 27.732,
+      "grad_norm": 2.2196757793426514,
+      "learning_rate": 2e-05,
+      "loss": 0.04846443,
+      "step": 13866
+    },
+    {
+      "epoch": 27.734,
+      "grad_norm": 1.4028538465499878,
+      "learning_rate": 2e-05,
+      "loss": 0.04624466,
+      "step": 13867
+    },
+    {
+      "epoch": 27.736,
+      "grad_norm": 1.165058970451355,
+      "learning_rate": 2e-05,
+      "loss": 0.0364709,
+      "step": 13868
+    },
+    {
+      "epoch": 27.738,
+      "grad_norm": 4.092937469482422,
+      "learning_rate": 2e-05,
+      "loss": 0.05190163,
+      "step": 13869
+    },
+    {
+      "epoch": 27.74,
+      "grad_norm": 1.328935146331787,
+      "learning_rate": 2e-05,
+      "loss": 0.05347368,
+      "step": 13870
+    },
+    {
+      "epoch": 27.742,
+      "grad_norm": 2.0258467197418213,
+      "learning_rate": 2e-05,
+      "loss": 0.03681408,
+      "step": 13871
+    },
+    {
+      "epoch": 27.744,
+      "grad_norm": 1.4326660633087158,
+      "learning_rate": 2e-05,
+      "loss": 0.06321809,
+      "step": 13872
+    },
+    {
+      "epoch": 27.746,
+      "grad_norm": 1.131027340888977,
+      "learning_rate": 2e-05,
+      "loss": 0.0508743,
+      "step": 13873
+    },
+    {
+      "epoch": 27.748,
+      "grad_norm": 1.18997061252594,
+      "learning_rate": 2e-05,
+      "loss": 0.03275201,
+      "step": 13874
+    },
+    {
+      "epoch": 27.75,
+      "grad_norm": 2.4015719890594482,
+      "learning_rate": 2e-05,
+      "loss": 0.03576087,
+      "step": 13875
+    },
+    {
+      "epoch": 27.752,
+      "grad_norm": 2.24355149269104,
+      "learning_rate": 2e-05,
+      "loss": 0.05075097,
+      "step": 13876
+    },
+    {
+      "epoch": 27.754,
+      "grad_norm": 1.1765543222427368,
+      "learning_rate": 2e-05,
+      "loss": 0.04541855,
+      "step": 13877
+    },
+    {
+      "epoch": 27.756,
+      "grad_norm": 1.5028316974639893,
+      "learning_rate": 2e-05,
+      "loss": 0.05536481,
+      "step": 13878
+    },
+    {
+      "epoch": 27.758,
+      "grad_norm": 1.1628483533859253,
+      "learning_rate": 2e-05,
+      "loss": 0.04130168,
+      "step": 13879
+    },
+    {
+      "epoch": 27.76,
+      "grad_norm": 1.752406120300293,
+      "learning_rate": 2e-05,
+      "loss": 0.05474552,
+      "step": 13880
+    },
+    {
+      "epoch": 27.762,
+      "grad_norm": 1.0870063304901123,
+      "learning_rate": 2e-05,
+      "loss": 0.04335282,
+      "step": 13881
+    },
+    {
+      "epoch": 27.764,
+      "grad_norm": 2.0513229370117188,
+      "learning_rate": 2e-05,
+      "loss": 0.03874574,
+      "step": 13882
+    },
+    {
+      "epoch": 27.766,
+      "grad_norm": 1.2223379611968994,
+      "learning_rate": 2e-05,
+      "loss": 0.05046418,
+      "step": 13883
+    },
+    {
+      "epoch": 27.768,
+      "grad_norm": 1.059180498123169,
+      "learning_rate": 2e-05,
+      "loss": 0.03355616,
+      "step": 13884
+    },
+    {
+      "epoch": 27.77,
+      "grad_norm": 1.3236563205718994,
+      "learning_rate": 2e-05,
+      "loss": 0.0412216,
+      "step": 13885
+    },
+    {
+      "epoch": 27.772,
+      "grad_norm": 1.560232162475586,
+      "learning_rate": 2e-05,
+      "loss": 0.03983791,
+      "step": 13886
+    },
+    {
+      "epoch": 27.774,
+      "grad_norm": 1.3728487491607666,
+      "learning_rate": 2e-05,
+      "loss": 0.04270254,
+      "step": 13887
+    },
+    {
+      "epoch": 27.776,
+      "grad_norm": 1.7650214433670044,
+      "learning_rate": 2e-05,
+      "loss": 0.03499018,
+      "step": 13888
+    },
+    {
+      "epoch": 27.778,
+      "grad_norm": 1.0585721731185913,
+      "learning_rate": 2e-05,
+      "loss": 0.03514803,
+      "step": 13889
+    },
+    {
+      "epoch": 27.78,
+      "grad_norm": 1.339477300643921,
+      "learning_rate": 2e-05,
+      "loss": 0.05417306,
+      "step": 13890
+    },
+    {
+      "epoch": 27.782,
+      "grad_norm": 1.4370014667510986,
+      "learning_rate": 2e-05,
+      "loss": 0.05690622,
+      "step": 13891
+    },
+    {
+      "epoch": 27.784,
+      "grad_norm": 1.5644543170928955,
+      "learning_rate": 2e-05,
+      "loss": 0.04291898,
+      "step": 13892
+    },
+    {
+      "epoch": 27.786,
+      "grad_norm": 1.0327696800231934,
+      "learning_rate": 2e-05,
+      "loss": 0.03720927,
+      "step": 13893
+    },
+    {
+      "epoch": 27.788,
+      "grad_norm": 1.0730547904968262,
+      "learning_rate": 2e-05,
+      "loss": 0.0368498,
+      "step": 13894
+    },
+    {
+      "epoch": 27.79,
+      "grad_norm": 1.1899945735931396,
+      "learning_rate": 2e-05,
+      "loss": 0.04839388,
+      "step": 13895
+    },
+    {
+      "epoch": 27.792,
+      "grad_norm": 1.9402941465377808,
+      "learning_rate": 2e-05,
+      "loss": 0.05492554,
+      "step": 13896
+    },
+    {
+      "epoch": 27.794,
+      "grad_norm": 1.4402939081192017,
+      "learning_rate": 2e-05,
+      "loss": 0.05447104,
+      "step": 13897
+    },
+    {
+      "epoch": 27.796,
+      "grad_norm": 2.334462881088257,
+      "learning_rate": 2e-05,
+      "loss": 0.05614328,
+      "step": 13898
+    },
+    {
+      "epoch": 27.798000000000002,
+      "grad_norm": 1.5610952377319336,
+      "learning_rate": 2e-05,
+      "loss": 0.05484813,
+      "step": 13899
+    },
+    {
+      "epoch": 27.8,
+      "grad_norm": 2.178652763366699,
+      "learning_rate": 2e-05,
+      "loss": 0.06258977,
+      "step": 13900
+    },
+    {
+      "epoch": 27.802,
+      "grad_norm": 1.075310468673706,
+      "learning_rate": 2e-05,
+      "loss": 0.03375444,
+      "step": 13901
+    },
+    {
+      "epoch": 27.804,
+      "grad_norm": 1.107140064239502,
+      "learning_rate": 2e-05,
+      "loss": 0.03990953,
+      "step": 13902
+    },
+    {
+      "epoch": 27.806,
+      "grad_norm": 1.2043249607086182,
+      "learning_rate": 2e-05,
+      "loss": 0.05589031,
+      "step": 13903
+    },
+    {
+      "epoch": 27.808,
+      "grad_norm": 3.252065658569336,
+      "learning_rate": 2e-05,
+      "loss": 0.04488713,
+      "step": 13904
+    },
+    {
+      "epoch": 27.81,
+      "grad_norm": 1.4825037717819214,
+      "learning_rate": 2e-05,
+      "loss": 0.04118239,
+      "step": 13905
+    },
+    {
+      "epoch": 27.812,
+      "grad_norm": 1.1587483882904053,
+      "learning_rate": 2e-05,
+      "loss": 0.03698997,
+      "step": 13906
+    },
+    {
+      "epoch": 27.814,
+      "grad_norm": 1.0998228788375854,
+      "learning_rate": 2e-05,
+      "loss": 0.03384409,
+      "step": 13907
+    },
+    {
+      "epoch": 27.816,
+      "grad_norm": 1.2784723043441772,
+      "learning_rate": 2e-05,
+      "loss": 0.04363027,
+      "step": 13908
+    },
+    {
+      "epoch": 27.818,
+      "grad_norm": 1.6805578470230103,
+      "learning_rate": 2e-05,
+      "loss": 0.05093968,
+      "step": 13909
+    },
+    {
+      "epoch": 27.82,
+      "grad_norm": 4.108209609985352,
+      "learning_rate": 2e-05,
+      "loss": 0.04699673,
+      "step": 13910
+    },
+    {
+      "epoch": 27.822,
+      "grad_norm": 1.0773557424545288,
+      "learning_rate": 2e-05,
+      "loss": 0.03354843,
+      "step": 13911
+    },
+    {
+      "epoch": 27.824,
+      "grad_norm": 2.1529176235198975,
+      "learning_rate": 2e-05,
+      "loss": 0.05770917,
+      "step": 13912
+    },
+    {
+      "epoch": 27.826,
+      "grad_norm": 1.2121992111206055,
+      "learning_rate": 2e-05,
+      "loss": 0.04613957,
+      "step": 13913
+    },
+    {
+      "epoch": 27.828,
+      "grad_norm": 1.3433780670166016,
+      "learning_rate": 2e-05,
+      "loss": 0.06019507,
+      "step": 13914
+    },
+    {
+      "epoch": 27.83,
+      "grad_norm": 1.4100244045257568,
+      "learning_rate": 2e-05,
+      "loss": 0.04316228,
+      "step": 13915
+    },
+    {
+      "epoch": 27.832,
+      "grad_norm": 2.21895694732666,
+      "learning_rate": 2e-05,
+      "loss": 0.04332146,
+      "step": 13916
+    },
+    {
+      "epoch": 27.834,
+      "grad_norm": 1.132088303565979,
+      "learning_rate": 2e-05,
+      "loss": 0.03310034,
+      "step": 13917
+    },
+    {
+      "epoch": 27.836,
+      "grad_norm": 0.9991299510002136,
+      "learning_rate": 2e-05,
+      "loss": 0.03843431,
+      "step": 13918
+    },
+    {
+      "epoch": 27.838,
+      "grad_norm": 1.3979414701461792,
+      "learning_rate": 2e-05,
+      "loss": 0.04863296,
+      "step": 13919
+    },
+    {
+      "epoch": 27.84,
+      "grad_norm": 1.2485167980194092,
+      "learning_rate": 2e-05,
+      "loss": 0.03016427,
+      "step": 13920
+    },
+    {
+      "epoch": 27.842,
+      "grad_norm": 2.4594881534576416,
+      "learning_rate": 2e-05,
+      "loss": 0.0523196,
+      "step": 13921
+    },
+    {
+      "epoch": 27.844,
+      "grad_norm": 1.169227957725525,
+      "learning_rate": 2e-05,
+      "loss": 0.04624384,
+      "step": 13922
+    },
+    {
+      "epoch": 27.846,
+      "grad_norm": 1.015640377998352,
+      "learning_rate": 2e-05,
+      "loss": 0.03382251,
+      "step": 13923
+    },
+    {
+      "epoch": 27.848,
+      "grad_norm": 1.3113818168640137,
+      "learning_rate": 2e-05,
+      "loss": 0.05841094,
+      "step": 13924
+    },
+    {
+      "epoch": 27.85,
+      "grad_norm": 1.174110770225525,
+      "learning_rate": 2e-05,
+      "loss": 0.03625581,
+      "step": 13925
+    },
+    {
+      "epoch": 27.852,
+      "grad_norm": 1.187264084815979,
+      "learning_rate": 2e-05,
+      "loss": 0.04166123,
+      "step": 13926
+    },
+    {
+      "epoch": 27.854,
+      "grad_norm": 1.0158615112304688,
+      "learning_rate": 2e-05,
+      "loss": 0.0311013,
+      "step": 13927
+    },
+    {
+      "epoch": 27.856,
+      "grad_norm": 1.158115267753601,
+      "learning_rate": 2e-05,
+      "loss": 0.03714382,
+      "step": 13928
+    },
+    {
+      "epoch": 27.858,
+      "grad_norm": 1.1665765047073364,
+      "learning_rate": 2e-05,
+      "loss": 0.04791439,
+      "step": 13929
+    },
+    {
+      "epoch": 27.86,
+      "grad_norm": 1.0975072383880615,
+      "learning_rate": 2e-05,
+      "loss": 0.05305796,
+      "step": 13930
+    },
+    {
+      "epoch": 27.862,
+      "grad_norm": 1.2020257711410522,
+      "learning_rate": 2e-05,
+      "loss": 0.03673447,
+      "step": 13931
+    },
+    {
+      "epoch": 27.864,
+      "grad_norm": 1.7321451902389526,
+      "learning_rate": 2e-05,
+      "loss": 0.03766854,
+      "step": 13932
+    },
+    {
+      "epoch": 27.866,
+      "grad_norm": 1.0557488203048706,
+      "learning_rate": 2e-05,
+      "loss": 0.03491737,
+      "step": 13933
+    },
+    {
+      "epoch": 27.868,
+      "grad_norm": 1.1876567602157593,
+      "learning_rate": 2e-05,
+      "loss": 0.02960133,
+      "step": 13934
+    },
+    {
+      "epoch": 27.87,
+      "grad_norm": 1.1305729150772095,
+      "learning_rate": 2e-05,
+      "loss": 0.02485805,
+      "step": 13935
+    },
+    {
+      "epoch": 27.872,
+      "grad_norm": 1.0962995290756226,
+      "learning_rate": 2e-05,
+      "loss": 0.05305071,
+      "step": 13936
+    },
+    {
+      "epoch": 27.874,
+      "grad_norm": 1.6864302158355713,
+      "learning_rate": 2e-05,
+      "loss": 0.05625251,
+      "step": 13937
+    },
+    {
+      "epoch": 27.876,
+      "grad_norm": 1.1173137426376343,
+      "learning_rate": 2e-05,
+      "loss": 0.04091568,
+      "step": 13938
+    },
+    {
+      "epoch": 27.878,
+      "grad_norm": 1.3211218118667603,
+      "learning_rate": 2e-05,
+      "loss": 0.05462737,
+      "step": 13939
+    },
+    {
+      "epoch": 27.88,
+      "grad_norm": 2.1169698238372803,
+      "learning_rate": 2e-05,
+      "loss": 0.07160946,
+      "step": 13940
+    },
+    {
+      "epoch": 27.882,
+      "grad_norm": 2.254796028137207,
+      "learning_rate": 2e-05,
+      "loss": 0.05478616,
+      "step": 13941
+    },
+    {
+      "epoch": 27.884,
+      "grad_norm": 1.0552836656570435,
+      "learning_rate": 2e-05,
+      "loss": 0.0460465,
+      "step": 13942
+    },
+    {
+      "epoch": 27.886,
+      "grad_norm": 1.2649129629135132,
+      "learning_rate": 2e-05,
+      "loss": 0.05034371,
+      "step": 13943
+    },
+    {
+      "epoch": 27.888,
+      "grad_norm": 1.4517078399658203,
+      "learning_rate": 2e-05,
+      "loss": 0.05885756,
+      "step": 13944
+    },
+    {
+      "epoch": 27.89,
+      "grad_norm": 1.3323506116867065,
+      "learning_rate": 2e-05,
+      "loss": 0.03592032,
+      "step": 13945
+    },
+    {
+      "epoch": 27.892,
+      "grad_norm": 1.004231333732605,
+      "learning_rate": 2e-05,
+      "loss": 0.03064884,
+      "step": 13946
+    },
+    {
+      "epoch": 27.894,
+      "grad_norm": 3.4080281257629395,
+      "learning_rate": 2e-05,
+      "loss": 0.04444969,
+      "step": 13947
+    },
+    {
+      "epoch": 27.896,
+      "grad_norm": 1.0597387552261353,
+      "learning_rate": 2e-05,
+      "loss": 0.05533614,
+      "step": 13948
+    },
+    {
+      "epoch": 27.898,
+      "grad_norm": 3.37524151802063,
+      "learning_rate": 2e-05,
+      "loss": 0.03681152,
+      "step": 13949
+    },
+    {
+      "epoch": 27.9,
+      "grad_norm": 1.2154083251953125,
+      "learning_rate": 2e-05,
+      "loss": 0.04092389,
+      "step": 13950
+    },
+    {
+      "epoch": 27.902,
+      "grad_norm": 1.1490893363952637,
+      "learning_rate": 2e-05,
+      "loss": 0.04752996,
+      "step": 13951
+    },
+    {
+      "epoch": 27.904,
+      "grad_norm": 2.7081265449523926,
+      "learning_rate": 2e-05,
+      "loss": 0.07245349,
+      "step": 13952
+    },
+    {
+      "epoch": 27.906,
+      "grad_norm": 0.9910997748374939,
+      "learning_rate": 2e-05,
+      "loss": 0.03678518,
+      "step": 13953
+    },
+    {
+      "epoch": 27.908,
+      "grad_norm": 1.374592900276184,
+      "learning_rate": 2e-05,
+      "loss": 0.05117041,
+      "step": 13954
+    },
+    {
+      "epoch": 27.91,
+      "grad_norm": 1.7255548238754272,
+      "learning_rate": 2e-05,
+      "loss": 0.04001204,
+      "step": 13955
+    },
+    {
+      "epoch": 27.912,
+      "grad_norm": 1.196602463722229,
+      "learning_rate": 2e-05,
+      "loss": 0.02974688,
+      "step": 13956
+    },
+    {
+      "epoch": 27.914,
+      "grad_norm": 1.4762085676193237,
+      "learning_rate": 2e-05,
+      "loss": 0.04770838,
+      "step": 13957
+    },
+    {
+      "epoch": 27.916,
+      "grad_norm": 1.137497067451477,
+      "learning_rate": 2e-05,
+      "loss": 0.04189302,
+      "step": 13958
+    },
+    {
+      "epoch": 27.918,
+      "grad_norm": 1.2305675745010376,
+      "learning_rate": 2e-05,
+      "loss": 0.03438773,
+      "step": 13959
+    },
+    {
+      "epoch": 27.92,
+      "grad_norm": 1.2307065725326538,
+      "learning_rate": 2e-05,
+      "loss": 0.0470115,
+      "step": 13960
+    },
+    {
+      "epoch": 27.922,
+      "grad_norm": 1.3657495975494385,
+      "learning_rate": 2e-05,
+      "loss": 0.04345737,
+      "step": 13961
+    },
+    {
+      "epoch": 27.924,
+      "grad_norm": 1.3622334003448486,
+      "learning_rate": 2e-05,
+      "loss": 0.04275557,
+      "step": 13962
+    },
+    {
+      "epoch": 27.926,
+      "grad_norm": 1.8857320547103882,
+      "learning_rate": 2e-05,
+      "loss": 0.04838826,
+      "step": 13963
+    },
+    {
+      "epoch": 27.928,
+      "grad_norm": 1.1448822021484375,
+      "learning_rate": 2e-05,
+      "loss": 0.04292282,
+      "step": 13964
+    },
+    {
+      "epoch": 27.93,
+      "grad_norm": 1.3071253299713135,
+      "learning_rate": 2e-05,
+      "loss": 0.04572289,
+      "step": 13965
+    },
+    {
+      "epoch": 27.932,
+      "grad_norm": 1.3378077745437622,
+      "learning_rate": 2e-05,
+      "loss": 0.04263856,
+      "step": 13966
+    },
+    {
+      "epoch": 27.934,
+      "grad_norm": 1.6963642835617065,
+      "learning_rate": 2e-05,
+      "loss": 0.05341676,
+      "step": 13967
+    },
+    {
+      "epoch": 27.936,
+      "grad_norm": 3.9375112056732178,
+      "learning_rate": 2e-05,
+      "loss": 0.05422161,
+      "step": 13968
+    },
+    {
+      "epoch": 27.938,
+      "grad_norm": 1.3592469692230225,
+      "learning_rate": 2e-05,
+      "loss": 0.06483057,
+      "step": 13969
+    },
+    {
+      "epoch": 27.94,
+      "grad_norm": 1.460677981376648,
+      "learning_rate": 2e-05,
+      "loss": 0.05142547,
+      "step": 13970
+    },
+    {
+      "epoch": 27.942,
+      "grad_norm": 1.2273732423782349,
+      "learning_rate": 2e-05,
+      "loss": 0.03599922,
+      "step": 13971
+    },
+    {
+      "epoch": 27.944,
+      "grad_norm": 1.2938038110733032,
+      "learning_rate": 2e-05,
+      "loss": 0.04321476,
+      "step": 13972
+    },
+    {
+      "epoch": 27.946,
+      "grad_norm": 1.007976770401001,
+      "learning_rate": 2e-05,
+      "loss": 0.04248318,
+      "step": 13973
+    },
+    {
+      "epoch": 27.948,
+      "grad_norm": 2.465160846710205,
+      "learning_rate": 2e-05,
+      "loss": 0.04337021,
+      "step": 13974
+    },
+    {
+      "epoch": 27.95,
+      "grad_norm": 3.6507928371429443,
+      "learning_rate": 2e-05,
+      "loss": 0.06409168,
+      "step": 13975
+    },
+    {
+      "epoch": 27.951999999999998,
+      "grad_norm": 1.0644190311431885,
+      "learning_rate": 2e-05,
+      "loss": 0.04959762,
+      "step": 13976
+    },
+    {
+      "epoch": 27.954,
+      "grad_norm": 1.2026413679122925,
+      "learning_rate": 2e-05,
+      "loss": 0.04680315,
+      "step": 13977
+    },
+    {
+      "epoch": 27.956,
+      "grad_norm": 0.87662672996521,
+      "learning_rate": 2e-05,
+      "loss": 0.03997586,
+      "step": 13978
+    },
+    {
+      "epoch": 27.958,
+      "grad_norm": 1.1587470769882202,
+      "learning_rate": 2e-05,
+      "loss": 0.0403148,
+      "step": 13979
+    },
+    {
+      "epoch": 27.96,
+      "grad_norm": 1.5906647443771362,
+      "learning_rate": 2e-05,
+      "loss": 0.05821125,
+      "step": 13980
+    },
+    {
+      "epoch": 27.962,
+      "grad_norm": 2.6488704681396484,
+      "learning_rate": 2e-05,
+      "loss": 0.05198992,
+      "step": 13981
+    },
+    {
+      "epoch": 27.964,
+      "grad_norm": 1.068997859954834,
+      "learning_rate": 2e-05,
+      "loss": 0.02909309,
+      "step": 13982
+    },
+    {
+      "epoch": 27.966,
+      "grad_norm": 1.0853064060211182,
+      "learning_rate": 2e-05,
+      "loss": 0.03459357,
+      "step": 13983
+    },
+    {
+      "epoch": 27.968,
+      "grad_norm": 1.227805733680725,
+      "learning_rate": 2e-05,
+      "loss": 0.04584693,
+      "step": 13984
+    },
+    {
+      "epoch": 27.97,
+      "grad_norm": 1.665924072265625,
+      "learning_rate": 2e-05,
+      "loss": 0.04428935,
+      "step": 13985
+    },
+    {
+      "epoch": 27.972,
+      "grad_norm": 1.8217002153396606,
+      "learning_rate": 2e-05,
+      "loss": 0.04603599,
+      "step": 13986
+    },
+    {
+      "epoch": 27.974,
+      "grad_norm": 1.1141926050186157,
+      "learning_rate": 2e-05,
+      "loss": 0.03780417,
+      "step": 13987
+    },
+    {
+      "epoch": 27.976,
+      "grad_norm": 1.0562472343444824,
+      "learning_rate": 2e-05,
+      "loss": 0.04129979,
+      "step": 13988
+    },
+    {
+      "epoch": 27.978,
+      "grad_norm": 1.1564449071884155,
+      "learning_rate": 2e-05,
+      "loss": 0.05021477,
+      "step": 13989
+    },
+    {
+      "epoch": 27.98,
+      "grad_norm": 1.0061001777648926,
+      "learning_rate": 2e-05,
+      "loss": 0.03430205,
+      "step": 13990
+    },
+    {
+      "epoch": 27.982,
+      "grad_norm": 3.191288709640503,
+      "learning_rate": 2e-05,
+      "loss": 0.04336466,
+      "step": 13991
+    },
+    {
+      "epoch": 27.984,
+      "grad_norm": 1.3708455562591553,
+      "learning_rate": 2e-05,
+      "loss": 0.03357625,
+      "step": 13992
+    },
+    {
+      "epoch": 27.986,
+      "grad_norm": 1.410815954208374,
+      "learning_rate": 2e-05,
+      "loss": 0.03761749,
+      "step": 13993
+    },
+    {
+      "epoch": 27.988,
+      "grad_norm": 1.0445529222488403,
+      "learning_rate": 2e-05,
+      "loss": 0.03375229,
+      "step": 13994
+    },
+    {
+      "epoch": 27.99,
+      "grad_norm": 1.9780248403549194,
+      "learning_rate": 2e-05,
+      "loss": 0.05313461,
+      "step": 13995
+    },
+    {
+      "epoch": 27.992,
+      "grad_norm": 1.082139492034912,
+      "learning_rate": 2e-05,
+      "loss": 0.04322237,
+      "step": 13996
+    },
+    {
+      "epoch": 27.994,
+      "grad_norm": 1.701316475868225,
+      "learning_rate": 2e-05,
+      "loss": 0.04349414,
+      "step": 13997
+    },
+    {
+      "epoch": 27.996,
+      "grad_norm": 1.5152575969696045,
+      "learning_rate": 2e-05,
+      "loss": 0.04271465,
+      "step": 13998
+    },
+    {
+      "epoch": 27.998,
+      "grad_norm": 1.3154436349868774,
+      "learning_rate": 2e-05,
+      "loss": 0.04630833,
+      "step": 13999
+    },
+    {
+      "epoch": 28.0,
+      "grad_norm": 1.4491382837295532,
+      "learning_rate": 2e-05,
+      "loss": 0.05390036,
+      "step": 14000
+    },
+    {
+      "epoch": 28.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9880239520958084,
+        "Equal_1": 0.998,
+        "Equal_2": 0.9820359281437125,
+        "Equal_3": 0.9840319361277445,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9940119760479041,
+        "Parallel_1": 0.9939879759519038,
+        "Parallel_2": 0.9979959919839679,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 1.0,
+        "Perpendicular_2": 0.988,
+        "Perpendicular_3": 0.8677354709418837,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 0.998,
+        "PointLiesOnCircle_3": 0.9876666666666667,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9860279441117764
+      },
+      "eval_runtime": 319.0556,
+      "eval_samples_per_second": 32.91,
+      "eval_steps_per_second": 0.658,
+      "step": 14000
+    },
+    {
+      "epoch": 28.002,
+      "grad_norm": 1.4470635652542114,
+      "learning_rate": 2e-05,
+      "loss": 0.03747822,
+      "step": 14001
+    },
+    {
+      "epoch": 28.004,
+      "grad_norm": 1.0949195623397827,
+      "learning_rate": 2e-05,
+      "loss": 0.04181771,
+      "step": 14002
+    },
+    {
+      "epoch": 28.006,
+      "grad_norm": 1.7039893865585327,
+      "learning_rate": 2e-05,
+      "loss": 0.04626131,
+      "step": 14003
+    },
+    {
+      "epoch": 28.008,
+      "grad_norm": 2.289400100708008,
+      "learning_rate": 2e-05,
+      "loss": 0.05722143,
+      "step": 14004
+    },
+    {
+      "epoch": 28.01,
+      "grad_norm": 1.7248917818069458,
+      "learning_rate": 2e-05,
+      "loss": 0.04162604,
+      "step": 14005
+    },
+    {
+      "epoch": 28.012,
+      "grad_norm": 1.235735297203064,
+      "learning_rate": 2e-05,
+      "loss": 0.05281135,
+      "step": 14006
+    },
+    {
+      "epoch": 28.014,
+      "grad_norm": 1.0774388313293457,
+      "learning_rate": 2e-05,
+      "loss": 0.03179125,
+      "step": 14007
+    },
+    {
+      "epoch": 28.016,
+      "grad_norm": 1.8190655708312988,
+      "learning_rate": 2e-05,
+      "loss": 0.04784362,
+      "step": 14008
+    },
+    {
+      "epoch": 28.018,
+      "grad_norm": 1.644675612449646,
+      "learning_rate": 2e-05,
+      "loss": 0.04483693,
+      "step": 14009
+    },
+    {
+      "epoch": 28.02,
+      "grad_norm": 1.27074134349823,
+      "learning_rate": 2e-05,
+      "loss": 0.05767696,
+      "step": 14010
+    },
+    {
+      "epoch": 28.022,
+      "grad_norm": 1.3586097955703735,
+      "learning_rate": 2e-05,
+      "loss": 0.03288449,
+      "step": 14011
+    },
+    {
+      "epoch": 28.024,
+      "grad_norm": 0.9760823845863342,
+      "learning_rate": 2e-05,
+      "loss": 0.03223011,
+      "step": 14012
+    },
+    {
+      "epoch": 28.026,
+      "grad_norm": 1.9947422742843628,
+      "learning_rate": 2e-05,
+      "loss": 0.06774095,
+      "step": 14013
+    },
+    {
+      "epoch": 28.028,
+      "grad_norm": 1.3853563070297241,
+      "learning_rate": 2e-05,
+      "loss": 0.03108654,
+      "step": 14014
+    },
+    {
+      "epoch": 28.03,
+      "grad_norm": 1.166358470916748,
+      "learning_rate": 2e-05,
+      "loss": 0.04371829,
+      "step": 14015
+    },
+    {
+      "epoch": 28.032,
+      "grad_norm": 1.3257215023040771,
+      "learning_rate": 2e-05,
+      "loss": 0.05647626,
+      "step": 14016
+    },
+    {
+      "epoch": 28.034,
+      "grad_norm": 1.8312715291976929,
+      "learning_rate": 2e-05,
+      "loss": 0.04174164,
+      "step": 14017
+    },
+    {
+      "epoch": 28.036,
+      "grad_norm": 1.1713063716888428,
+      "learning_rate": 2e-05,
+      "loss": 0.04272055,
+      "step": 14018
+    },
+    {
+      "epoch": 28.038,
+      "grad_norm": 1.022841453552246,
+      "learning_rate": 2e-05,
+      "loss": 0.04106878,
+      "step": 14019
+    },
+    {
+      "epoch": 28.04,
+      "grad_norm": 1.123375654220581,
+      "learning_rate": 2e-05,
+      "loss": 0.04132652,
+      "step": 14020
+    },
+    {
+      "epoch": 28.042,
+      "grad_norm": 0.9225612878799438,
+      "learning_rate": 2e-05,
+      "loss": 0.03926476,
+      "step": 14021
+    },
+    {
+      "epoch": 28.044,
+      "grad_norm": 1.3327795267105103,
+      "learning_rate": 2e-05,
+      "loss": 0.04138635,
+      "step": 14022
+    },
+    {
+      "epoch": 28.046,
+      "grad_norm": 1.0123158693313599,
+      "learning_rate": 2e-05,
+      "loss": 0.03201216,
+      "step": 14023
+    },
+    {
+      "epoch": 28.048,
+      "grad_norm": 1.0678902864456177,
+      "learning_rate": 2e-05,
+      "loss": 0.03800803,
+      "step": 14024
+    },
+    {
+      "epoch": 28.05,
+      "grad_norm": 1.6009563207626343,
+      "learning_rate": 2e-05,
+      "loss": 0.05565003,
+      "step": 14025
+    },
+    {
+      "epoch": 28.052,
+      "grad_norm": 1.7109339237213135,
+      "learning_rate": 2e-05,
+      "loss": 0.03978261,
+      "step": 14026
+    },
+    {
+      "epoch": 28.054,
+      "grad_norm": 0.9139453172683716,
+      "learning_rate": 2e-05,
+      "loss": 0.02852735,
+      "step": 14027
+    },
+    {
+      "epoch": 28.056,
+      "grad_norm": 1.2070707082748413,
+      "learning_rate": 2e-05,
+      "loss": 0.04159069,
+      "step": 14028
+    },
+    {
+      "epoch": 28.058,
+      "grad_norm": 1.1406387090682983,
+      "learning_rate": 2e-05,
+      "loss": 0.03497103,
+      "step": 14029
+    },
+    {
+      "epoch": 28.06,
+      "grad_norm": 1.5811642408370972,
+      "learning_rate": 2e-05,
+      "loss": 0.04424217,
+      "step": 14030
+    },
+    {
+      "epoch": 28.062,
+      "grad_norm": 2.008183479309082,
+      "learning_rate": 2e-05,
+      "loss": 0.0522799,
+      "step": 14031
+    },
+    {
+      "epoch": 28.064,
+      "grad_norm": 1.3083295822143555,
+      "learning_rate": 2e-05,
+      "loss": 0.04245078,
+      "step": 14032
+    },
+    {
+      "epoch": 28.066,
+      "grad_norm": 1.156947135925293,
+      "learning_rate": 2e-05,
+      "loss": 0.04440115,
+      "step": 14033
+    },
+    {
+      "epoch": 28.068,
+      "grad_norm": 1.2095609903335571,
+      "learning_rate": 2e-05,
+      "loss": 0.0388485,
+      "step": 14034
+    },
+    {
+      "epoch": 28.07,
+      "grad_norm": 1.2254390716552734,
+      "learning_rate": 2e-05,
+      "loss": 0.03347994,
+      "step": 14035
+    },
+    {
+      "epoch": 28.072,
+      "grad_norm": 1.1406832933425903,
+      "learning_rate": 2e-05,
+      "loss": 0.04151241,
+      "step": 14036
+    },
+    {
+      "epoch": 28.074,
+      "grad_norm": 1.8342012166976929,
+      "learning_rate": 2e-05,
+      "loss": 0.0339776,
+      "step": 14037
+    },
+    {
+      "epoch": 28.076,
+      "grad_norm": 1.5849714279174805,
+      "learning_rate": 2e-05,
+      "loss": 0.04472189,
+      "step": 14038
+    },
+    {
+      "epoch": 28.078,
+      "grad_norm": 1.7151453495025635,
+      "learning_rate": 2e-05,
+      "loss": 0.04204217,
+      "step": 14039
+    },
+    {
+      "epoch": 28.08,
+      "grad_norm": 1.2614927291870117,
+      "learning_rate": 2e-05,
+      "loss": 0.04573628,
+      "step": 14040
+    },
+    {
+      "epoch": 28.082,
+      "grad_norm": 0.9816139936447144,
+      "learning_rate": 2e-05,
+      "loss": 0.04488077,
+      "step": 14041
+    },
+    {
+      "epoch": 28.084,
+      "grad_norm": 1.832466959953308,
+      "learning_rate": 2e-05,
+      "loss": 0.03606717,
+      "step": 14042
+    },
+    {
+      "epoch": 28.086,
+      "grad_norm": 1.1039230823516846,
+      "learning_rate": 2e-05,
+      "loss": 0.03455424,
+      "step": 14043
+    },
+    {
+      "epoch": 28.088,
+      "grad_norm": 1.3150477409362793,
+      "learning_rate": 2e-05,
+      "loss": 0.03564423,
+      "step": 14044
+    },
+    {
+      "epoch": 28.09,
+      "grad_norm": 1.4359805583953857,
+      "learning_rate": 2e-05,
+      "loss": 0.04819302,
+      "step": 14045
+    },
+    {
+      "epoch": 28.092,
+      "grad_norm": 1.6399556398391724,
+      "learning_rate": 2e-05,
+      "loss": 0.06461568,
+      "step": 14046
+    },
+    {
+      "epoch": 28.094,
+      "grad_norm": 1.261181116104126,
+      "learning_rate": 2e-05,
+      "loss": 0.03727288,
+      "step": 14047
+    },
+    {
+      "epoch": 28.096,
+      "grad_norm": 1.3191063404083252,
+      "learning_rate": 2e-05,
+      "loss": 0.05051003,
+      "step": 14048
+    },
+    {
+      "epoch": 28.098,
+      "grad_norm": 2.9893617630004883,
+      "learning_rate": 2e-05,
+      "loss": 0.04233899,
+      "step": 14049
+    },
+    {
+      "epoch": 28.1,
+      "grad_norm": 1.1125555038452148,
+      "learning_rate": 2e-05,
+      "loss": 0.03035634,
+      "step": 14050
+    },
+    {
+      "epoch": 28.102,
+      "grad_norm": 1.1988699436187744,
+      "learning_rate": 2e-05,
+      "loss": 0.0434362,
+      "step": 14051
+    },
+    {
+      "epoch": 28.104,
+      "grad_norm": 1.1385446786880493,
+      "learning_rate": 2e-05,
+      "loss": 0.04272396,
+      "step": 14052
+    },
+    {
+      "epoch": 28.106,
+      "grad_norm": 1.5576671361923218,
+      "learning_rate": 2e-05,
+      "loss": 0.04542271,
+      "step": 14053
+    },
+    {
+      "epoch": 28.108,
+      "grad_norm": 0.9911758303642273,
+      "learning_rate": 2e-05,
+      "loss": 0.02662791,
+      "step": 14054
+    },
+    {
+      "epoch": 28.11,
+      "grad_norm": 1.0654523372650146,
+      "learning_rate": 2e-05,
+      "loss": 0.02663458,
+      "step": 14055
+    },
+    {
+      "epoch": 28.112,
+      "grad_norm": 2.447866916656494,
+      "learning_rate": 2e-05,
+      "loss": 0.06156901,
+      "step": 14056
+    },
+    {
+      "epoch": 28.114,
+      "grad_norm": 1.5394794940948486,
+      "learning_rate": 2e-05,
+      "loss": 0.03734413,
+      "step": 14057
+    },
+    {
+      "epoch": 28.116,
+      "grad_norm": 1.2493864297866821,
+      "learning_rate": 2e-05,
+      "loss": 0.04325924,
+      "step": 14058
+    },
+    {
+      "epoch": 28.118,
+      "grad_norm": 1.0674172639846802,
+      "learning_rate": 2e-05,
+      "loss": 0.03373833,
+      "step": 14059
+    },
+    {
+      "epoch": 28.12,
+      "grad_norm": 1.6935789585113525,
+      "learning_rate": 2e-05,
+      "loss": 0.06536911,
+      "step": 14060
+    },
+    {
+      "epoch": 28.122,
+      "grad_norm": 1.346823811531067,
+      "learning_rate": 2e-05,
+      "loss": 0.03599031,
+      "step": 14061
+    },
+    {
+      "epoch": 28.124,
+      "grad_norm": 1.2584784030914307,
+      "learning_rate": 2e-05,
+      "loss": 0.04426415,
+      "step": 14062
+    },
+    {
+      "epoch": 28.126,
+      "grad_norm": 1.5686087608337402,
+      "learning_rate": 2e-05,
+      "loss": 0.03532685,
+      "step": 14063
+    },
+    {
+      "epoch": 28.128,
+      "grad_norm": 1.0472480058670044,
+      "learning_rate": 2e-05,
+      "loss": 0.02852079,
+      "step": 14064
+    },
+    {
+      "epoch": 28.13,
+      "grad_norm": 1.4226044416427612,
+      "learning_rate": 2e-05,
+      "loss": 0.05019859,
+      "step": 14065
+    },
+    {
+      "epoch": 28.132,
+      "grad_norm": 1.4346106052398682,
+      "learning_rate": 2e-05,
+      "loss": 0.05388649,
+      "step": 14066
+    },
+    {
+      "epoch": 28.134,
+      "grad_norm": 2.281315803527832,
+      "learning_rate": 2e-05,
+      "loss": 0.0545974,
+      "step": 14067
+    },
+    {
+      "epoch": 28.136,
+      "grad_norm": 1.2159454822540283,
+      "learning_rate": 2e-05,
+      "loss": 0.03459051,
+      "step": 14068
+    },
+    {
+      "epoch": 28.138,
+      "grad_norm": 1.2951035499572754,
+      "learning_rate": 2e-05,
+      "loss": 0.0595167,
+      "step": 14069
+    },
+    {
+      "epoch": 28.14,
+      "grad_norm": 1.1814157962799072,
+      "learning_rate": 2e-05,
+      "loss": 0.03543475,
+      "step": 14070
+    },
+    {
+      "epoch": 28.142,
+      "grad_norm": 1.0047551393508911,
+      "learning_rate": 2e-05,
+      "loss": 0.03344669,
+      "step": 14071
+    },
+    {
+      "epoch": 28.144,
+      "grad_norm": 0.9211133718490601,
+      "learning_rate": 2e-05,
+      "loss": 0.03054952,
+      "step": 14072
+    },
+    {
+      "epoch": 28.146,
+      "grad_norm": 0.922265887260437,
+      "learning_rate": 2e-05,
+      "loss": 0.03004513,
+      "step": 14073
+    },
+    {
+      "epoch": 28.148,
+      "grad_norm": 1.5452007055282593,
+      "learning_rate": 2e-05,
+      "loss": 0.0379502,
+      "step": 14074
+    },
+    {
+      "epoch": 28.15,
+      "grad_norm": 1.1936745643615723,
+      "learning_rate": 2e-05,
+      "loss": 0.0383325,
+      "step": 14075
+    },
+    {
+      "epoch": 28.152,
+      "grad_norm": 0.9541253447532654,
+      "learning_rate": 2e-05,
+      "loss": 0.03261286,
+      "step": 14076
+    },
+    {
+      "epoch": 28.154,
+      "grad_norm": 1.3197306394577026,
+      "learning_rate": 2e-05,
+      "loss": 0.05566575,
+      "step": 14077
+    },
+    {
+      "epoch": 28.156,
+      "grad_norm": 1.5139483213424683,
+      "learning_rate": 2e-05,
+      "loss": 0.04211396,
+      "step": 14078
+    },
+    {
+      "epoch": 28.158,
+      "grad_norm": 2.144613742828369,
+      "learning_rate": 2e-05,
+      "loss": 0.05478022,
+      "step": 14079
+    },
+    {
+      "epoch": 28.16,
+      "grad_norm": 1.2877733707427979,
+      "learning_rate": 2e-05,
+      "loss": 0.04739144,
+      "step": 14080
+    },
+    {
+      "epoch": 28.162,
+      "grad_norm": 0.9685249924659729,
+      "learning_rate": 2e-05,
+      "loss": 0.02751518,
+      "step": 14081
+    },
+    {
+      "epoch": 28.164,
+      "grad_norm": 1.1108578443527222,
+      "learning_rate": 2e-05,
+      "loss": 0.04293613,
+      "step": 14082
+    },
+    {
+      "epoch": 28.166,
+      "grad_norm": 0.9321027994155884,
+      "learning_rate": 2e-05,
+      "loss": 0.03453288,
+      "step": 14083
+    },
+    {
+      "epoch": 28.168,
+      "grad_norm": 1.8604676723480225,
+      "learning_rate": 2e-05,
+      "loss": 0.03813556,
+      "step": 14084
+    },
+    {
+      "epoch": 28.17,
+      "grad_norm": 1.3579093217849731,
+      "learning_rate": 2e-05,
+      "loss": 0.03601374,
+      "step": 14085
+    },
+    {
+      "epoch": 28.172,
+      "grad_norm": 1.3795456886291504,
+      "learning_rate": 2e-05,
+      "loss": 0.05737968,
+      "step": 14086
+    },
+    {
+      "epoch": 28.174,
+      "grad_norm": 1.3870776891708374,
+      "learning_rate": 2e-05,
+      "loss": 0.03918634,
+      "step": 14087
+    },
+    {
+      "epoch": 28.176,
+      "grad_norm": 1.5775760412216187,
+      "learning_rate": 2e-05,
+      "loss": 0.04204413,
+      "step": 14088
+    },
+    {
+      "epoch": 28.178,
+      "grad_norm": 1.3750708103179932,
+      "learning_rate": 2e-05,
+      "loss": 0.03528126,
+      "step": 14089
+    },
+    {
+      "epoch": 28.18,
+      "grad_norm": 1.1607511043548584,
+      "learning_rate": 2e-05,
+      "loss": 0.0513554,
+      "step": 14090
+    },
+    {
+      "epoch": 28.182,
+      "grad_norm": 1.7518668174743652,
+      "learning_rate": 2e-05,
+      "loss": 0.05958167,
+      "step": 14091
+    },
+    {
+      "epoch": 28.184,
+      "grad_norm": 0.8187534213066101,
+      "learning_rate": 2e-05,
+      "loss": 0.02375086,
+      "step": 14092
+    },
+    {
+      "epoch": 28.186,
+      "grad_norm": 1.1586307287216187,
+      "learning_rate": 2e-05,
+      "loss": 0.03455898,
+      "step": 14093
+    },
+    {
+      "epoch": 28.188,
+      "grad_norm": 2.6694910526275635,
+      "learning_rate": 2e-05,
+      "loss": 0.04185054,
+      "step": 14094
+    },
+    {
+      "epoch": 28.19,
+      "grad_norm": 1.0022153854370117,
+      "learning_rate": 2e-05,
+      "loss": 0.03717811,
+      "step": 14095
+    },
+    {
+      "epoch": 28.192,
+      "grad_norm": 1.133339762687683,
+      "learning_rate": 2e-05,
+      "loss": 0.02898293,
+      "step": 14096
+    },
+    {
+      "epoch": 28.194,
+      "grad_norm": 0.9910487532615662,
+      "learning_rate": 2e-05,
+      "loss": 0.02521181,
+      "step": 14097
+    },
+    {
+      "epoch": 28.196,
+      "grad_norm": 1.2302407026290894,
+      "learning_rate": 2e-05,
+      "loss": 0.04486918,
+      "step": 14098
+    },
+    {
+      "epoch": 28.198,
+      "grad_norm": 1.398140788078308,
+      "learning_rate": 2e-05,
+      "loss": 0.02154693,
+      "step": 14099
+    },
+    {
+      "epoch": 28.2,
+      "grad_norm": 2.0596702098846436,
+      "learning_rate": 2e-05,
+      "loss": 0.03849518,
+      "step": 14100
+    },
+    {
+      "epoch": 28.202,
+      "grad_norm": 1.3504750728607178,
+      "learning_rate": 2e-05,
+      "loss": 0.0484315,
+      "step": 14101
+    },
+    {
+      "epoch": 28.204,
+      "grad_norm": 1.0688234567642212,
+      "learning_rate": 2e-05,
+      "loss": 0.0351585,
+      "step": 14102
+    },
+    {
+      "epoch": 28.206,
+      "grad_norm": 1.4983340501785278,
+      "learning_rate": 2e-05,
+      "loss": 0.04929278,
+      "step": 14103
+    },
+    {
+      "epoch": 28.208,
+      "grad_norm": 1.1455981731414795,
+      "learning_rate": 2e-05,
+      "loss": 0.04318264,
+      "step": 14104
+    },
+    {
+      "epoch": 28.21,
+      "grad_norm": 1.3201831579208374,
+      "learning_rate": 2e-05,
+      "loss": 0.03962881,
+      "step": 14105
+    },
+    {
+      "epoch": 28.212,
+      "grad_norm": 1.7087925672531128,
+      "learning_rate": 2e-05,
+      "loss": 0.05958664,
+      "step": 14106
+    },
+    {
+      "epoch": 28.214,
+      "grad_norm": 1.799774169921875,
+      "learning_rate": 2e-05,
+      "loss": 0.04615483,
+      "step": 14107
+    },
+    {
+      "epoch": 28.216,
+      "grad_norm": 2.0174641609191895,
+      "learning_rate": 2e-05,
+      "loss": 0.03510554,
+      "step": 14108
+    },
+    {
+      "epoch": 28.218,
+      "grad_norm": 1.7066216468811035,
+      "learning_rate": 2e-05,
+      "loss": 0.04433943,
+      "step": 14109
+    },
+    {
+      "epoch": 28.22,
+      "grad_norm": 1.1175912618637085,
+      "learning_rate": 2e-05,
+      "loss": 0.04259259,
+      "step": 14110
+    },
+    {
+      "epoch": 28.222,
+      "grad_norm": 1.3678927421569824,
+      "learning_rate": 2e-05,
+      "loss": 0.0313809,
+      "step": 14111
+    },
+    {
+      "epoch": 28.224,
+      "grad_norm": 1.330997109413147,
+      "learning_rate": 2e-05,
+      "loss": 0.03561566,
+      "step": 14112
+    },
+    {
+      "epoch": 28.226,
+      "grad_norm": 1.0542383193969727,
+      "learning_rate": 2e-05,
+      "loss": 0.03285637,
+      "step": 14113
+    },
+    {
+      "epoch": 28.228,
+      "grad_norm": 1.6406633853912354,
+      "learning_rate": 2e-05,
+      "loss": 0.03652732,
+      "step": 14114
+    },
+    {
+      "epoch": 28.23,
+      "grad_norm": 1.4692353010177612,
+      "learning_rate": 2e-05,
+      "loss": 0.04664152,
+      "step": 14115
+    },
+    {
+      "epoch": 28.232,
+      "grad_norm": 3.2253143787384033,
+      "learning_rate": 2e-05,
+      "loss": 0.052161,
+      "step": 14116
+    },
+    {
+      "epoch": 28.234,
+      "grad_norm": 1.0887095928192139,
+      "learning_rate": 2e-05,
+      "loss": 0.03422359,
+      "step": 14117
+    },
+    {
+      "epoch": 28.236,
+      "grad_norm": 1.4389132261276245,
+      "learning_rate": 2e-05,
+      "loss": 0.0335054,
+      "step": 14118
+    },
+    {
+      "epoch": 28.238,
+      "grad_norm": 1.1777735948562622,
+      "learning_rate": 2e-05,
+      "loss": 0.0480193,
+      "step": 14119
+    },
+    {
+      "epoch": 28.24,
+      "grad_norm": 1.2068471908569336,
+      "learning_rate": 2e-05,
+      "loss": 0.03690059,
+      "step": 14120
+    },
+    {
+      "epoch": 28.242,
+      "grad_norm": 0.9290897846221924,
+      "learning_rate": 2e-05,
+      "loss": 0.03334764,
+      "step": 14121
+    },
+    {
+      "epoch": 28.244,
+      "grad_norm": 1.0152655839920044,
+      "learning_rate": 2e-05,
+      "loss": 0.04105952,
+      "step": 14122
+    },
+    {
+      "epoch": 28.246,
+      "grad_norm": 1.3126530647277832,
+      "learning_rate": 2e-05,
+      "loss": 0.03845873,
+      "step": 14123
+    },
+    {
+      "epoch": 28.248,
+      "grad_norm": 1.0583008527755737,
+      "learning_rate": 2e-05,
+      "loss": 0.02750574,
+      "step": 14124
+    },
+    {
+      "epoch": 28.25,
+      "grad_norm": 1.3489996194839478,
+      "learning_rate": 2e-05,
+      "loss": 0.03019518,
+      "step": 14125
+    },
+    {
+      "epoch": 28.252,
+      "grad_norm": 1.9256123304367065,
+      "learning_rate": 2e-05,
+      "loss": 0.0618353,
+      "step": 14126
+    },
+    {
+      "epoch": 28.254,
+      "grad_norm": 1.2744817733764648,
+      "learning_rate": 2e-05,
+      "loss": 0.05729286,
+      "step": 14127
+    },
+    {
+      "epoch": 28.256,
+      "grad_norm": 1.2284598350524902,
+      "learning_rate": 2e-05,
+      "loss": 0.0571045,
+      "step": 14128
+    },
+    {
+      "epoch": 28.258,
+      "grad_norm": 1.3543322086334229,
+      "learning_rate": 2e-05,
+      "loss": 0.04329966,
+      "step": 14129
+    },
+    {
+      "epoch": 28.26,
+      "grad_norm": 1.3076565265655518,
+      "learning_rate": 2e-05,
+      "loss": 0.04858845,
+      "step": 14130
+    },
+    {
+      "epoch": 28.262,
+      "grad_norm": 1.8700453042984009,
+      "learning_rate": 2e-05,
+      "loss": 0.06022802,
+      "step": 14131
+    },
+    {
+      "epoch": 28.264,
+      "grad_norm": 1.6746340990066528,
+      "learning_rate": 2e-05,
+      "loss": 0.05681061,
+      "step": 14132
+    },
+    {
+      "epoch": 28.266,
+      "grad_norm": 1.4257715940475464,
+      "learning_rate": 2e-05,
+      "loss": 0.04958214,
+      "step": 14133
+    },
+    {
+      "epoch": 28.268,
+      "grad_norm": 1.2631217241287231,
+      "learning_rate": 2e-05,
+      "loss": 0.03757473,
+      "step": 14134
+    },
+    {
+      "epoch": 28.27,
+      "grad_norm": 1.4865784645080566,
+      "learning_rate": 2e-05,
+      "loss": 0.0356946,
+      "step": 14135
+    },
+    {
+      "epoch": 28.272,
+      "grad_norm": 1.2895032167434692,
+      "learning_rate": 2e-05,
+      "loss": 0.03698609,
+      "step": 14136
+    },
+    {
+      "epoch": 28.274,
+      "grad_norm": 1.0424330234527588,
+      "learning_rate": 2e-05,
+      "loss": 0.03331292,
+      "step": 14137
+    },
+    {
+      "epoch": 28.276,
+      "grad_norm": 0.9681985378265381,
+      "learning_rate": 2e-05,
+      "loss": 0.02686777,
+      "step": 14138
+    },
+    {
+      "epoch": 28.278,
+      "grad_norm": 1.1451689004898071,
+      "learning_rate": 2e-05,
+      "loss": 0.04804129,
+      "step": 14139
+    },
+    {
+      "epoch": 28.28,
+      "grad_norm": 1.2859004735946655,
+      "learning_rate": 2e-05,
+      "loss": 0.03937485,
+      "step": 14140
+    },
+    {
+      "epoch": 28.282,
+      "grad_norm": 1.9779037237167358,
+      "learning_rate": 2e-05,
+      "loss": 0.04771383,
+      "step": 14141
+    },
+    {
+      "epoch": 28.284,
+      "grad_norm": 1.6972112655639648,
+      "learning_rate": 2e-05,
+      "loss": 0.04609143,
+      "step": 14142
+    },
+    {
+      "epoch": 28.286,
+      "grad_norm": 1.5024956464767456,
+      "learning_rate": 2e-05,
+      "loss": 0.04176302,
+      "step": 14143
+    },
+    {
+      "epoch": 28.288,
+      "grad_norm": 2.7473392486572266,
+      "learning_rate": 2e-05,
+      "loss": 0.06857879,
+      "step": 14144
+    },
+    {
+      "epoch": 28.29,
+      "grad_norm": 1.476321816444397,
+      "learning_rate": 2e-05,
+      "loss": 0.04816295,
+      "step": 14145
+    },
+    {
+      "epoch": 28.292,
+      "grad_norm": 1.9630745649337769,
+      "learning_rate": 2e-05,
+      "loss": 0.04113162,
+      "step": 14146
+    },
+    {
+      "epoch": 28.294,
+      "grad_norm": 1.5762404203414917,
+      "learning_rate": 2e-05,
+      "loss": 0.04521403,
+      "step": 14147
+    },
+    {
+      "epoch": 28.296,
+      "grad_norm": 1.8791242837905884,
+      "learning_rate": 2e-05,
+      "loss": 0.05574446,
+      "step": 14148
+    },
+    {
+      "epoch": 28.298,
+      "grad_norm": 1.3232239484786987,
+      "learning_rate": 2e-05,
+      "loss": 0.04592151,
+      "step": 14149
+    },
+    {
+      "epoch": 28.3,
+      "grad_norm": 1.0860222578048706,
+      "learning_rate": 2e-05,
+      "loss": 0.03201285,
+      "step": 14150
+    },
+    {
+      "epoch": 28.302,
+      "grad_norm": 1.1450574398040771,
+      "learning_rate": 2e-05,
+      "loss": 0.0431738,
+      "step": 14151
+    },
+    {
+      "epoch": 28.304,
+      "grad_norm": 1.2073911428451538,
+      "learning_rate": 2e-05,
+      "loss": 0.04106464,
+      "step": 14152
+    },
+    {
+      "epoch": 28.306,
+      "grad_norm": 2.2250092029571533,
+      "learning_rate": 2e-05,
+      "loss": 0.03272331,
+      "step": 14153
+    },
+    {
+      "epoch": 28.308,
+      "grad_norm": 1.2592679262161255,
+      "learning_rate": 2e-05,
+      "loss": 0.03767652,
+      "step": 14154
+    },
+    {
+      "epoch": 28.31,
+      "grad_norm": 1.481679081916809,
+      "learning_rate": 2e-05,
+      "loss": 0.04322655,
+      "step": 14155
+    },
+    {
+      "epoch": 28.312,
+      "grad_norm": 1.8476687669754028,
+      "learning_rate": 2e-05,
+      "loss": 0.02506692,
+      "step": 14156
+    },
+    {
+      "epoch": 28.314,
+      "grad_norm": 2.533881664276123,
+      "learning_rate": 2e-05,
+      "loss": 0.05798201,
+      "step": 14157
+    },
+    {
+      "epoch": 28.316,
+      "grad_norm": 1.16340172290802,
+      "learning_rate": 2e-05,
+      "loss": 0.04741621,
+      "step": 14158
+    },
+    {
+      "epoch": 28.318,
+      "grad_norm": 1.0567313432693481,
+      "learning_rate": 2e-05,
+      "loss": 0.03934213,
+      "step": 14159
+    },
+    {
+      "epoch": 28.32,
+      "grad_norm": 1.8486943244934082,
+      "learning_rate": 2e-05,
+      "loss": 0.04713926,
+      "step": 14160
+    },
+    {
+      "epoch": 28.322,
+      "grad_norm": 1.489852786064148,
+      "learning_rate": 2e-05,
+      "loss": 0.0400935,
+      "step": 14161
+    },
+    {
+      "epoch": 28.324,
+      "grad_norm": 3.173983097076416,
+      "learning_rate": 2e-05,
+      "loss": 0.05666588,
+      "step": 14162
+    },
+    {
+      "epoch": 28.326,
+      "grad_norm": 1.2710679769515991,
+      "learning_rate": 2e-05,
+      "loss": 0.04324594,
+      "step": 14163
+    },
+    {
+      "epoch": 28.328,
+      "grad_norm": 2.056770086288452,
+      "learning_rate": 2e-05,
+      "loss": 0.0498774,
+      "step": 14164
+    },
+    {
+      "epoch": 28.33,
+      "grad_norm": 1.194877028465271,
+      "learning_rate": 2e-05,
+      "loss": 0.033945,
+      "step": 14165
+    },
+    {
+      "epoch": 28.332,
+      "grad_norm": 1.5242279767990112,
+      "learning_rate": 2e-05,
+      "loss": 0.04730607,
+      "step": 14166
+    },
+    {
+      "epoch": 28.334,
+      "grad_norm": 1.2096294164657593,
+      "learning_rate": 2e-05,
+      "loss": 0.04421799,
+      "step": 14167
+    },
+    {
+      "epoch": 28.336,
+      "grad_norm": 2.0177338123321533,
+      "learning_rate": 2e-05,
+      "loss": 0.05057753,
+      "step": 14168
+    },
+    {
+      "epoch": 28.338,
+      "grad_norm": 1.4078340530395508,
+      "learning_rate": 2e-05,
+      "loss": 0.04942624,
+      "step": 14169
+    },
+    {
+      "epoch": 28.34,
+      "grad_norm": 2.0044713020324707,
+      "learning_rate": 2e-05,
+      "loss": 0.05860762,
+      "step": 14170
+    },
+    {
+      "epoch": 28.342,
+      "grad_norm": 1.4696686267852783,
+      "learning_rate": 2e-05,
+      "loss": 0.05207686,
+      "step": 14171
+    },
+    {
+      "epoch": 28.344,
+      "grad_norm": 1.4194014072418213,
+      "learning_rate": 2e-05,
+      "loss": 0.03678628,
+      "step": 14172
+    },
+    {
+      "epoch": 28.346,
+      "grad_norm": 1.3729615211486816,
+      "learning_rate": 2e-05,
+      "loss": 0.07338053,
+      "step": 14173
+    },
+    {
+      "epoch": 28.348,
+      "grad_norm": 1.4968407154083252,
+      "learning_rate": 2e-05,
+      "loss": 0.05206795,
+      "step": 14174
+    },
+    {
+      "epoch": 28.35,
+      "grad_norm": 0.9872764348983765,
+      "learning_rate": 2e-05,
+      "loss": 0.03436659,
+      "step": 14175
+    },
+    {
+      "epoch": 28.352,
+      "grad_norm": 1.0700467824935913,
+      "learning_rate": 2e-05,
+      "loss": 0.02983767,
+      "step": 14176
+    },
+    {
+      "epoch": 28.354,
+      "grad_norm": 1.9751871824264526,
+      "learning_rate": 2e-05,
+      "loss": 0.04803991,
+      "step": 14177
+    },
+    {
+      "epoch": 28.356,
+      "grad_norm": 1.7464139461517334,
+      "learning_rate": 2e-05,
+      "loss": 0.03985727,
+      "step": 14178
+    },
+    {
+      "epoch": 28.358,
+      "grad_norm": 1.93489670753479,
+      "learning_rate": 2e-05,
+      "loss": 0.03951417,
+      "step": 14179
+    },
+    {
+      "epoch": 28.36,
+      "grad_norm": 1.8252264261245728,
+      "learning_rate": 2e-05,
+      "loss": 0.05627768,
+      "step": 14180
+    },
+    {
+      "epoch": 28.362,
+      "grad_norm": 1.957201361656189,
+      "learning_rate": 2e-05,
+      "loss": 0.03644243,
+      "step": 14181
+    },
+    {
+      "epoch": 28.364,
+      "grad_norm": 1.244791865348816,
+      "learning_rate": 2e-05,
+      "loss": 0.05136299,
+      "step": 14182
+    },
+    {
+      "epoch": 28.366,
+      "grad_norm": 1.722070336341858,
+      "learning_rate": 2e-05,
+      "loss": 0.05756142,
+      "step": 14183
+    },
+    {
+      "epoch": 28.368,
+      "grad_norm": 1.8835808038711548,
+      "learning_rate": 2e-05,
+      "loss": 0.05007068,
+      "step": 14184
+    },
+    {
+      "epoch": 28.37,
+      "grad_norm": 1.2348777055740356,
+      "learning_rate": 2e-05,
+      "loss": 0.03819226,
+      "step": 14185
+    },
+    {
+      "epoch": 28.372,
+      "grad_norm": 1.257193684577942,
+      "learning_rate": 2e-05,
+      "loss": 0.05539931,
+      "step": 14186
+    },
+    {
+      "epoch": 28.374,
+      "grad_norm": 2.638650894165039,
+      "learning_rate": 2e-05,
+      "loss": 0.03770338,
+      "step": 14187
+    },
+    {
+      "epoch": 28.376,
+      "grad_norm": 1.6504069566726685,
+      "learning_rate": 2e-05,
+      "loss": 0.05300485,
+      "step": 14188
+    },
+    {
+      "epoch": 28.378,
+      "grad_norm": 1.1879500150680542,
+      "learning_rate": 2e-05,
+      "loss": 0.04724121,
+      "step": 14189
+    },
+    {
+      "epoch": 28.38,
+      "grad_norm": 1.1239627599716187,
+      "learning_rate": 2e-05,
+      "loss": 0.04006843,
+      "step": 14190
+    },
+    {
+      "epoch": 28.382,
+      "grad_norm": 1.3962266445159912,
+      "learning_rate": 2e-05,
+      "loss": 0.04130918,
+      "step": 14191
+    },
+    {
+      "epoch": 28.384,
+      "grad_norm": 1.0262913703918457,
+      "learning_rate": 2e-05,
+      "loss": 0.03978799,
+      "step": 14192
+    },
+    {
+      "epoch": 28.386,
+      "grad_norm": 1.3596129417419434,
+      "learning_rate": 2e-05,
+      "loss": 0.06046655,
+      "step": 14193
+    },
+    {
+      "epoch": 28.388,
+      "grad_norm": 1.1571170091629028,
+      "learning_rate": 2e-05,
+      "loss": 0.05023224,
+      "step": 14194
+    },
+    {
+      "epoch": 28.39,
+      "grad_norm": 1.3431873321533203,
+      "learning_rate": 2e-05,
+      "loss": 0.05262809,
+      "step": 14195
+    },
+    {
+      "epoch": 28.392,
+      "grad_norm": 2.0367238521575928,
+      "learning_rate": 2e-05,
+      "loss": 0.04556445,
+      "step": 14196
+    },
+    {
+      "epoch": 28.394,
+      "grad_norm": 1.3604267835617065,
+      "learning_rate": 2e-05,
+      "loss": 0.04495424,
+      "step": 14197
+    },
+    {
+      "epoch": 28.396,
+      "grad_norm": 1.0701876878738403,
+      "learning_rate": 2e-05,
+      "loss": 0.03906007,
+      "step": 14198
+    },
+    {
+      "epoch": 28.398,
+      "grad_norm": 1.1933096647262573,
+      "learning_rate": 2e-05,
+      "loss": 0.05860349,
+      "step": 14199
+    },
+    {
+      "epoch": 28.4,
+      "grad_norm": 1.3504351377487183,
+      "learning_rate": 2e-05,
+      "loss": 0.04702557,
+      "step": 14200
+    },
+    {
+      "epoch": 28.402,
+      "grad_norm": 1.2822675704956055,
+      "learning_rate": 2e-05,
+      "loss": 0.03257849,
+      "step": 14201
+    },
+    {
+      "epoch": 28.404,
+      "grad_norm": 1.6527397632598877,
+      "learning_rate": 2e-05,
+      "loss": 0.05858765,
+      "step": 14202
+    },
+    {
+      "epoch": 28.406,
+      "grad_norm": 1.0486624240875244,
+      "learning_rate": 2e-05,
+      "loss": 0.03177363,
+      "step": 14203
+    },
+    {
+      "epoch": 28.408,
+      "grad_norm": 1.128495216369629,
+      "learning_rate": 2e-05,
+      "loss": 0.04799465,
+      "step": 14204
+    },
+    {
+      "epoch": 28.41,
+      "grad_norm": 1.0047136545181274,
+      "learning_rate": 2e-05,
+      "loss": 0.03385249,
+      "step": 14205
+    },
+    {
+      "epoch": 28.412,
+      "grad_norm": 1.4099255800247192,
+      "learning_rate": 2e-05,
+      "loss": 0.0562316,
+      "step": 14206
+    },
+    {
+      "epoch": 28.414,
+      "grad_norm": 2.193967580795288,
+      "learning_rate": 2e-05,
+      "loss": 0.07079425,
+      "step": 14207
+    },
+    {
+      "epoch": 28.416,
+      "grad_norm": 1.1035521030426025,
+      "learning_rate": 2e-05,
+      "loss": 0.04856515,
+      "step": 14208
+    },
+    {
+      "epoch": 28.418,
+      "grad_norm": 1.315811038017273,
+      "learning_rate": 2e-05,
+      "loss": 0.03866019,
+      "step": 14209
+    },
+    {
+      "epoch": 28.42,
+      "grad_norm": 1.1560747623443604,
+      "learning_rate": 2e-05,
+      "loss": 0.04573543,
+      "step": 14210
+    },
+    {
+      "epoch": 28.422,
+      "grad_norm": 1.3213006258010864,
+      "learning_rate": 2e-05,
+      "loss": 0.04729943,
+      "step": 14211
+    },
+    {
+      "epoch": 28.424,
+      "grad_norm": 1.467768907546997,
+      "learning_rate": 2e-05,
+      "loss": 0.0497423,
+      "step": 14212
+    },
+    {
+      "epoch": 28.426,
+      "grad_norm": 1.0412020683288574,
+      "learning_rate": 2e-05,
+      "loss": 0.03104374,
+      "step": 14213
+    },
+    {
+      "epoch": 28.428,
+      "grad_norm": 1.289223074913025,
+      "learning_rate": 2e-05,
+      "loss": 0.04878268,
+      "step": 14214
+    },
+    {
+      "epoch": 28.43,
+      "grad_norm": 1.0868279933929443,
+      "learning_rate": 2e-05,
+      "loss": 0.04009105,
+      "step": 14215
+    },
+    {
+      "epoch": 28.432,
+      "grad_norm": 1.506838321685791,
+      "learning_rate": 2e-05,
+      "loss": 0.04779705,
+      "step": 14216
+    },
+    {
+      "epoch": 28.434,
+      "grad_norm": 1.3598910570144653,
+      "learning_rate": 2e-05,
+      "loss": 0.0399574,
+      "step": 14217
+    },
+    {
+      "epoch": 28.436,
+      "grad_norm": 1.165126085281372,
+      "learning_rate": 2e-05,
+      "loss": 0.03589812,
+      "step": 14218
+    },
+    {
+      "epoch": 28.438,
+      "grad_norm": 1.403108835220337,
+      "learning_rate": 2e-05,
+      "loss": 0.03950584,
+      "step": 14219
+    },
+    {
+      "epoch": 28.44,
+      "grad_norm": 0.9602160453796387,
+      "learning_rate": 2e-05,
+      "loss": 0.0343546,
+      "step": 14220
+    },
+    {
+      "epoch": 28.442,
+      "grad_norm": 1.707650065422058,
+      "learning_rate": 2e-05,
+      "loss": 0.04669677,
+      "step": 14221
+    },
+    {
+      "epoch": 28.444,
+      "grad_norm": 1.5494041442871094,
+      "learning_rate": 2e-05,
+      "loss": 0.05189268,
+      "step": 14222
+    },
+    {
+      "epoch": 28.446,
+      "grad_norm": 1.4422967433929443,
+      "learning_rate": 2e-05,
+      "loss": 0.04455651,
+      "step": 14223
+    },
+    {
+      "epoch": 28.448,
+      "grad_norm": 2.1481945514678955,
+      "learning_rate": 2e-05,
+      "loss": 0.04172315,
+      "step": 14224
+    },
+    {
+      "epoch": 28.45,
+      "grad_norm": 1.8955481052398682,
+      "learning_rate": 2e-05,
+      "loss": 0.05096827,
+      "step": 14225
+    },
+    {
+      "epoch": 28.452,
+      "grad_norm": 1.4757720232009888,
+      "learning_rate": 2e-05,
+      "loss": 0.0427974,
+      "step": 14226
+    },
+    {
+      "epoch": 28.454,
+      "grad_norm": 3.5340352058410645,
+      "learning_rate": 2e-05,
+      "loss": 0.03696814,
+      "step": 14227
+    },
+    {
+      "epoch": 28.456,
+      "grad_norm": 1.2981595993041992,
+      "learning_rate": 2e-05,
+      "loss": 0.04960819,
+      "step": 14228
+    },
+    {
+      "epoch": 28.458,
+      "grad_norm": 1.3688567876815796,
+      "learning_rate": 2e-05,
+      "loss": 0.04346091,
+      "step": 14229
+    },
+    {
+      "epoch": 28.46,
+      "grad_norm": 1.0863968133926392,
+      "learning_rate": 2e-05,
+      "loss": 0.04432026,
+      "step": 14230
+    },
+    {
+      "epoch": 28.462,
+      "grad_norm": 1.1809697151184082,
+      "learning_rate": 2e-05,
+      "loss": 0.04062311,
+      "step": 14231
+    },
+    {
+      "epoch": 28.464,
+      "grad_norm": 1.087384581565857,
+      "learning_rate": 2e-05,
+      "loss": 0.04801937,
+      "step": 14232
+    },
+    {
+      "epoch": 28.466,
+      "grad_norm": 3.0253777503967285,
+      "learning_rate": 2e-05,
+      "loss": 0.06829751,
+      "step": 14233
+    },
+    {
+      "epoch": 28.468,
+      "grad_norm": 0.9512907862663269,
+      "learning_rate": 2e-05,
+      "loss": 0.02574901,
+      "step": 14234
+    },
+    {
+      "epoch": 28.47,
+      "grad_norm": 1.2561564445495605,
+      "learning_rate": 2e-05,
+      "loss": 0.03712705,
+      "step": 14235
+    },
+    {
+      "epoch": 28.472,
+      "grad_norm": 1.114723801612854,
+      "learning_rate": 2e-05,
+      "loss": 0.04053148,
+      "step": 14236
+    },
+    {
+      "epoch": 28.474,
+      "grad_norm": 0.9256406426429749,
+      "learning_rate": 2e-05,
+      "loss": 0.0282863,
+      "step": 14237
+    },
+    {
+      "epoch": 28.476,
+      "grad_norm": 1.0956227779388428,
+      "learning_rate": 2e-05,
+      "loss": 0.03660049,
+      "step": 14238
+    },
+    {
+      "epoch": 28.478,
+      "grad_norm": 1.1929676532745361,
+      "learning_rate": 2e-05,
+      "loss": 0.03722402,
+      "step": 14239
+    },
+    {
+      "epoch": 28.48,
+      "grad_norm": 1.0275921821594238,
+      "learning_rate": 2e-05,
+      "loss": 0.03313165,
+      "step": 14240
+    },
+    {
+      "epoch": 28.482,
+      "grad_norm": 1.4167778491973877,
+      "learning_rate": 2e-05,
+      "loss": 0.04980793,
+      "step": 14241
+    },
+    {
+      "epoch": 28.484,
+      "grad_norm": 2.8569796085357666,
+      "learning_rate": 2e-05,
+      "loss": 0.04864322,
+      "step": 14242
+    },
+    {
+      "epoch": 28.486,
+      "grad_norm": 1.0566939115524292,
+      "learning_rate": 2e-05,
+      "loss": 0.03854397,
+      "step": 14243
+    },
+    {
+      "epoch": 28.488,
+      "grad_norm": 1.8737938404083252,
+      "learning_rate": 2e-05,
+      "loss": 0.06766672,
+      "step": 14244
+    },
+    {
+      "epoch": 28.49,
+      "grad_norm": 1.1267757415771484,
+      "learning_rate": 2e-05,
+      "loss": 0.03949556,
+      "step": 14245
+    },
+    {
+      "epoch": 28.492,
+      "grad_norm": 1.5439599752426147,
+      "learning_rate": 2e-05,
+      "loss": 0.03746489,
+      "step": 14246
+    },
+    {
+      "epoch": 28.494,
+      "grad_norm": 1.10368812084198,
+      "learning_rate": 2e-05,
+      "loss": 0.04418413,
+      "step": 14247
+    },
+    {
+      "epoch": 28.496,
+      "grad_norm": 1.8851401805877686,
+      "learning_rate": 2e-05,
+      "loss": 0.05599113,
+      "step": 14248
+    },
+    {
+      "epoch": 28.498,
+      "grad_norm": 1.218051552772522,
+      "learning_rate": 2e-05,
+      "loss": 0.0527468,
+      "step": 14249
+    },
+    {
+      "epoch": 28.5,
+      "grad_norm": 2.2494425773620605,
+      "learning_rate": 2e-05,
+      "loss": 0.05196273,
+      "step": 14250
+    },
+    {
+      "epoch": 28.502,
+      "grad_norm": 1.1959189176559448,
+      "learning_rate": 2e-05,
+      "loss": 0.04170246,
+      "step": 14251
+    },
+    {
+      "epoch": 28.504,
+      "grad_norm": 1.4841954708099365,
+      "learning_rate": 2e-05,
+      "loss": 0.04115407,
+      "step": 14252
+    },
+    {
+      "epoch": 28.506,
+      "grad_norm": 1.4579923152923584,
+      "learning_rate": 2e-05,
+      "loss": 0.06711036,
+      "step": 14253
+    },
+    {
+      "epoch": 28.508,
+      "grad_norm": 1.3409069776535034,
+      "learning_rate": 2e-05,
+      "loss": 0.04993474,
+      "step": 14254
+    },
+    {
+      "epoch": 28.51,
+      "grad_norm": 0.9282925128936768,
+      "learning_rate": 2e-05,
+      "loss": 0.02424309,
+      "step": 14255
+    },
+    {
+      "epoch": 28.512,
+      "grad_norm": 1.0717816352844238,
+      "learning_rate": 2e-05,
+      "loss": 0.03973263,
+      "step": 14256
+    },
+    {
+      "epoch": 28.514,
+      "grad_norm": 1.3361554145812988,
+      "learning_rate": 2e-05,
+      "loss": 0.03955629,
+      "step": 14257
+    },
+    {
+      "epoch": 28.516,
+      "grad_norm": 2.1933910846710205,
+      "learning_rate": 2e-05,
+      "loss": 0.05748781,
+      "step": 14258
+    },
+    {
+      "epoch": 28.518,
+      "grad_norm": 1.1817597150802612,
+      "learning_rate": 2e-05,
+      "loss": 0.04961357,
+      "step": 14259
+    },
+    {
+      "epoch": 28.52,
+      "grad_norm": 1.298127293586731,
+      "learning_rate": 2e-05,
+      "loss": 0.04155416,
+      "step": 14260
+    },
+    {
+      "epoch": 28.522,
+      "grad_norm": 1.060938835144043,
+      "learning_rate": 2e-05,
+      "loss": 0.04268425,
+      "step": 14261
+    },
+    {
+      "epoch": 28.524,
+      "grad_norm": 1.8386595249176025,
+      "learning_rate": 2e-05,
+      "loss": 0.04054004,
+      "step": 14262
+    },
+    {
+      "epoch": 28.526,
+      "grad_norm": 1.0489224195480347,
+      "learning_rate": 2e-05,
+      "loss": 0.0311989,
+      "step": 14263
+    },
+    {
+      "epoch": 28.528,
+      "grad_norm": 1.3312833309173584,
+      "learning_rate": 2e-05,
+      "loss": 0.04081573,
+      "step": 14264
+    },
+    {
+      "epoch": 28.53,
+      "grad_norm": 1.7386471033096313,
+      "learning_rate": 2e-05,
+      "loss": 0.05501184,
+      "step": 14265
+    },
+    {
+      "epoch": 28.532,
+      "grad_norm": 1.1394075155258179,
+      "learning_rate": 2e-05,
+      "loss": 0.0496092,
+      "step": 14266
+    },
+    {
+      "epoch": 28.534,
+      "grad_norm": 1.0800431966781616,
+      "learning_rate": 2e-05,
+      "loss": 0.04388725,
+      "step": 14267
+    },
+    {
+      "epoch": 28.536,
+      "grad_norm": 1.365820288658142,
+      "learning_rate": 2e-05,
+      "loss": 0.04674472,
+      "step": 14268
+    },
+    {
+      "epoch": 28.538,
+      "grad_norm": 1.6938581466674805,
+      "learning_rate": 2e-05,
+      "loss": 0.03524645,
+      "step": 14269
+    },
+    {
+      "epoch": 28.54,
+      "grad_norm": 1.0656059980392456,
+      "learning_rate": 2e-05,
+      "loss": 0.0350618,
+      "step": 14270
+    },
+    {
+      "epoch": 28.542,
+      "grad_norm": 1.300699234008789,
+      "learning_rate": 2e-05,
+      "loss": 0.0439024,
+      "step": 14271
+    },
+    {
+      "epoch": 28.544,
+      "grad_norm": 1.0407557487487793,
+      "learning_rate": 2e-05,
+      "loss": 0.03287885,
+      "step": 14272
+    },
+    {
+      "epoch": 28.546,
+      "grad_norm": 1.203810214996338,
+      "learning_rate": 2e-05,
+      "loss": 0.03666135,
+      "step": 14273
+    },
+    {
+      "epoch": 28.548000000000002,
+      "grad_norm": 1.3447908163070679,
+      "learning_rate": 2e-05,
+      "loss": 0.03037151,
+      "step": 14274
+    },
+    {
+      "epoch": 28.55,
+      "grad_norm": 1.6330796480178833,
+      "learning_rate": 2e-05,
+      "loss": 0.04389308,
+      "step": 14275
+    },
+    {
+      "epoch": 28.552,
+      "grad_norm": 1.453428030014038,
+      "learning_rate": 2e-05,
+      "loss": 0.05481421,
+      "step": 14276
+    },
+    {
+      "epoch": 28.554,
+      "grad_norm": 1.3805347681045532,
+      "learning_rate": 2e-05,
+      "loss": 0.05831959,
+      "step": 14277
+    },
+    {
+      "epoch": 28.556,
+      "grad_norm": 1.1240483522415161,
+      "learning_rate": 2e-05,
+      "loss": 0.03230035,
+      "step": 14278
+    },
+    {
+      "epoch": 28.558,
+      "grad_norm": 1.5682629346847534,
+      "learning_rate": 2e-05,
+      "loss": 0.04137525,
+      "step": 14279
+    },
+    {
+      "epoch": 28.56,
+      "grad_norm": 1.067922830581665,
+      "learning_rate": 2e-05,
+      "loss": 0.03095679,
+      "step": 14280
+    },
+    {
+      "epoch": 28.562,
+      "grad_norm": 1.3047091960906982,
+      "learning_rate": 2e-05,
+      "loss": 0.040992,
+      "step": 14281
+    },
+    {
+      "epoch": 28.564,
+      "grad_norm": 1.1715583801269531,
+      "learning_rate": 2e-05,
+      "loss": 0.04538528,
+      "step": 14282
+    },
+    {
+      "epoch": 28.566,
+      "grad_norm": 2.710317373275757,
+      "learning_rate": 2e-05,
+      "loss": 0.04703784,
+      "step": 14283
+    },
+    {
+      "epoch": 28.568,
+      "grad_norm": 1.1046556234359741,
+      "learning_rate": 2e-05,
+      "loss": 0.03596055,
+      "step": 14284
+    },
+    {
+      "epoch": 28.57,
+      "grad_norm": 1.5328203439712524,
+      "learning_rate": 2e-05,
+      "loss": 0.05959145,
+      "step": 14285
+    },
+    {
+      "epoch": 28.572,
+      "grad_norm": 1.1067321300506592,
+      "learning_rate": 2e-05,
+      "loss": 0.03717593,
+      "step": 14286
+    },
+    {
+      "epoch": 28.574,
+      "grad_norm": 1.5921152830123901,
+      "learning_rate": 2e-05,
+      "loss": 0.04818038,
+      "step": 14287
+    },
+    {
+      "epoch": 28.576,
+      "grad_norm": 1.2906737327575684,
+      "learning_rate": 2e-05,
+      "loss": 0.04694307,
+      "step": 14288
+    },
+    {
+      "epoch": 28.578,
+      "grad_norm": 1.4389523267745972,
+      "learning_rate": 2e-05,
+      "loss": 0.06396471,
+      "step": 14289
+    },
+    {
+      "epoch": 28.58,
+      "grad_norm": 1.250801920890808,
+      "learning_rate": 2e-05,
+      "loss": 0.03378974,
+      "step": 14290
+    },
+    {
+      "epoch": 28.582,
+      "grad_norm": 1.8675917387008667,
+      "learning_rate": 2e-05,
+      "loss": 0.04210771,
+      "step": 14291
+    },
+    {
+      "epoch": 28.584,
+      "grad_norm": 1.2386926412582397,
+      "learning_rate": 2e-05,
+      "loss": 0.04732635,
+      "step": 14292
+    },
+    {
+      "epoch": 28.586,
+      "grad_norm": 2.474187135696411,
+      "learning_rate": 2e-05,
+      "loss": 0.06141835,
+      "step": 14293
+    },
+    {
+      "epoch": 28.588,
+      "grad_norm": 1.7406890392303467,
+      "learning_rate": 2e-05,
+      "loss": 0.0405802,
+      "step": 14294
+    },
+    {
+      "epoch": 28.59,
+      "grad_norm": 1.0976345539093018,
+      "learning_rate": 2e-05,
+      "loss": 0.03914689,
+      "step": 14295
+    },
+    {
+      "epoch": 28.592,
+      "grad_norm": 1.8601065874099731,
+      "learning_rate": 2e-05,
+      "loss": 0.04376506,
+      "step": 14296
+    },
+    {
+      "epoch": 28.594,
+      "grad_norm": 1.8947181701660156,
+      "learning_rate": 2e-05,
+      "loss": 0.05417305,
+      "step": 14297
+    },
+    {
+      "epoch": 28.596,
+      "grad_norm": 1.3904882669448853,
+      "learning_rate": 2e-05,
+      "loss": 0.05922655,
+      "step": 14298
+    },
+    {
+      "epoch": 28.598,
+      "grad_norm": 1.1092592477798462,
+      "learning_rate": 2e-05,
+      "loss": 0.04238461,
+      "step": 14299
+    },
+    {
+      "epoch": 28.6,
+      "grad_norm": 1.1075598001480103,
+      "learning_rate": 2e-05,
+      "loss": 0.03969569,
+      "step": 14300
+    },
+    {
+      "epoch": 28.602,
+      "grad_norm": 1.2379860877990723,
+      "learning_rate": 2e-05,
+      "loss": 0.05977491,
+      "step": 14301
+    },
+    {
+      "epoch": 28.604,
+      "grad_norm": 2.523383378982544,
+      "learning_rate": 2e-05,
+      "loss": 0.04559822,
+      "step": 14302
+    },
+    {
+      "epoch": 28.606,
+      "grad_norm": 1.390464186668396,
+      "learning_rate": 2e-05,
+      "loss": 0.03601509,
+      "step": 14303
+    },
+    {
+      "epoch": 28.608,
+      "grad_norm": 1.1424009799957275,
+      "learning_rate": 2e-05,
+      "loss": 0.04128938,
+      "step": 14304
+    },
+    {
+      "epoch": 28.61,
+      "grad_norm": 1.5949323177337646,
+      "learning_rate": 2e-05,
+      "loss": 0.03449958,
+      "step": 14305
+    },
+    {
+      "epoch": 28.612,
+      "grad_norm": 1.0766521692276,
+      "learning_rate": 2e-05,
+      "loss": 0.03662252,
+      "step": 14306
+    },
+    {
+      "epoch": 28.614,
+      "grad_norm": 1.684144377708435,
+      "learning_rate": 2e-05,
+      "loss": 0.03993898,
+      "step": 14307
+    },
+    {
+      "epoch": 28.616,
+      "grad_norm": 1.2283536195755005,
+      "learning_rate": 2e-05,
+      "loss": 0.0440466,
+      "step": 14308
+    },
+    {
+      "epoch": 28.618,
+      "grad_norm": 1.1252257823944092,
+      "learning_rate": 2e-05,
+      "loss": 0.03805107,
+      "step": 14309
+    },
+    {
+      "epoch": 28.62,
+      "grad_norm": 2.2137584686279297,
+      "learning_rate": 2e-05,
+      "loss": 0.04531762,
+      "step": 14310
+    },
+    {
+      "epoch": 28.622,
+      "grad_norm": 1.8632186651229858,
+      "learning_rate": 2e-05,
+      "loss": 0.04363943,
+      "step": 14311
+    },
+    {
+      "epoch": 28.624,
+      "grad_norm": 2.1854779720306396,
+      "learning_rate": 2e-05,
+      "loss": 0.04152116,
+      "step": 14312
+    },
+    {
+      "epoch": 28.626,
+      "grad_norm": 1.199179768562317,
+      "learning_rate": 2e-05,
+      "loss": 0.04882842,
+      "step": 14313
+    },
+    {
+      "epoch": 28.628,
+      "grad_norm": 1.9288642406463623,
+      "learning_rate": 2e-05,
+      "loss": 0.06294123,
+      "step": 14314
+    },
+    {
+      "epoch": 28.63,
+      "grad_norm": 0.9891992807388306,
+      "learning_rate": 2e-05,
+      "loss": 0.03864823,
+      "step": 14315
+    },
+    {
+      "epoch": 28.632,
+      "grad_norm": 1.1384472846984863,
+      "learning_rate": 2e-05,
+      "loss": 0.03697848,
+      "step": 14316
+    },
+    {
+      "epoch": 28.634,
+      "grad_norm": 1.3403377532958984,
+      "learning_rate": 2e-05,
+      "loss": 0.0520224,
+      "step": 14317
+    },
+    {
+      "epoch": 28.636,
+      "grad_norm": 1.8841863870620728,
+      "learning_rate": 2e-05,
+      "loss": 0.03481198,
+      "step": 14318
+    },
+    {
+      "epoch": 28.638,
+      "grad_norm": 2.253284215927124,
+      "learning_rate": 2e-05,
+      "loss": 0.05113673,
+      "step": 14319
+    },
+    {
+      "epoch": 28.64,
+      "grad_norm": 1.678731918334961,
+      "learning_rate": 2e-05,
+      "loss": 0.03881251,
+      "step": 14320
+    },
+    {
+      "epoch": 28.642,
+      "grad_norm": 1.419081211090088,
+      "learning_rate": 2e-05,
+      "loss": 0.03930322,
+      "step": 14321
+    },
+    {
+      "epoch": 28.644,
+      "grad_norm": 2.305083990097046,
+      "learning_rate": 2e-05,
+      "loss": 0.03415557,
+      "step": 14322
+    },
+    {
+      "epoch": 28.646,
+      "grad_norm": 1.7453691959381104,
+      "learning_rate": 2e-05,
+      "loss": 0.05402776,
+      "step": 14323
+    },
+    {
+      "epoch": 28.648,
+      "grad_norm": 1.6866177320480347,
+      "learning_rate": 2e-05,
+      "loss": 0.05624064,
+      "step": 14324
+    },
+    {
+      "epoch": 28.65,
+      "grad_norm": 2.8119149208068848,
+      "learning_rate": 2e-05,
+      "loss": 0.03970397,
+      "step": 14325
+    },
+    {
+      "epoch": 28.652,
+      "grad_norm": 1.4768586158752441,
+      "learning_rate": 2e-05,
+      "loss": 0.04976168,
+      "step": 14326
+    },
+    {
+      "epoch": 28.654,
+      "grad_norm": 1.2367441654205322,
+      "learning_rate": 2e-05,
+      "loss": 0.04001617,
+      "step": 14327
+    },
+    {
+      "epoch": 28.656,
+      "grad_norm": 1.3339909315109253,
+      "learning_rate": 2e-05,
+      "loss": 0.0409479,
+      "step": 14328
+    },
+    {
+      "epoch": 28.658,
+      "grad_norm": 1.3718701601028442,
+      "learning_rate": 2e-05,
+      "loss": 0.06367579,
+      "step": 14329
+    },
+    {
+      "epoch": 28.66,
+      "grad_norm": 1.323080062866211,
+      "learning_rate": 2e-05,
+      "loss": 0.0345336,
+      "step": 14330
+    },
+    {
+      "epoch": 28.662,
+      "grad_norm": 0.998365044593811,
+      "learning_rate": 2e-05,
+      "loss": 0.03076694,
+      "step": 14331
+    },
+    {
+      "epoch": 28.664,
+      "grad_norm": 1.3956735134124756,
+      "learning_rate": 2e-05,
+      "loss": 0.04270991,
+      "step": 14332
+    },
+    {
+      "epoch": 28.666,
+      "grad_norm": 1.0146839618682861,
+      "learning_rate": 2e-05,
+      "loss": 0.03931056,
+      "step": 14333
+    },
+    {
+      "epoch": 28.668,
+      "grad_norm": 1.7695027589797974,
+      "learning_rate": 2e-05,
+      "loss": 0.0385569,
+      "step": 14334
+    },
+    {
+      "epoch": 28.67,
+      "grad_norm": 1.3834205865859985,
+      "learning_rate": 2e-05,
+      "loss": 0.05658727,
+      "step": 14335
+    },
+    {
+      "epoch": 28.672,
+      "grad_norm": 1.1219671964645386,
+      "learning_rate": 2e-05,
+      "loss": 0.03792579,
+      "step": 14336
+    },
+    {
+      "epoch": 28.674,
+      "grad_norm": 1.3171169757843018,
+      "learning_rate": 2e-05,
+      "loss": 0.03911427,
+      "step": 14337
+    },
+    {
+      "epoch": 28.676,
+      "grad_norm": 1.2529704570770264,
+      "learning_rate": 2e-05,
+      "loss": 0.05144939,
+      "step": 14338
+    },
+    {
+      "epoch": 28.678,
+      "grad_norm": 0.909761905670166,
+      "learning_rate": 2e-05,
+      "loss": 0.02666837,
+      "step": 14339
+    },
+    {
+      "epoch": 28.68,
+      "grad_norm": 1.0001680850982666,
+      "learning_rate": 2e-05,
+      "loss": 0.03465949,
+      "step": 14340
+    },
+    {
+      "epoch": 28.682,
+      "grad_norm": 1.1715612411499023,
+      "learning_rate": 2e-05,
+      "loss": 0.03337407,
+      "step": 14341
+    },
+    {
+      "epoch": 28.684,
+      "grad_norm": 1.56963050365448,
+      "learning_rate": 2e-05,
+      "loss": 0.04966843,
+      "step": 14342
+    },
+    {
+      "epoch": 28.686,
+      "grad_norm": 1.6177550554275513,
+      "learning_rate": 2e-05,
+      "loss": 0.06656225,
+      "step": 14343
+    },
+    {
+      "epoch": 28.688,
+      "grad_norm": 1.37739896774292,
+      "learning_rate": 2e-05,
+      "loss": 0.04131253,
+      "step": 14344
+    },
+    {
+      "epoch": 28.69,
+      "grad_norm": 1.133405327796936,
+      "learning_rate": 2e-05,
+      "loss": 0.04604138,
+      "step": 14345
+    },
+    {
+      "epoch": 28.692,
+      "grad_norm": 1.2516900300979614,
+      "learning_rate": 2e-05,
+      "loss": 0.04709205,
+      "step": 14346
+    },
+    {
+      "epoch": 28.694,
+      "grad_norm": 1.7897838354110718,
+      "learning_rate": 2e-05,
+      "loss": 0.04589042,
+      "step": 14347
+    },
+    {
+      "epoch": 28.696,
+      "grad_norm": 0.9966497421264648,
+      "learning_rate": 2e-05,
+      "loss": 0.0285507,
+      "step": 14348
+    },
+    {
+      "epoch": 28.698,
+      "grad_norm": 1.126671314239502,
+      "learning_rate": 2e-05,
+      "loss": 0.04037992,
+      "step": 14349
+    },
+    {
+      "epoch": 28.7,
+      "grad_norm": 1.076209306716919,
+      "learning_rate": 2e-05,
+      "loss": 0.04209936,
+      "step": 14350
+    },
+    {
+      "epoch": 28.701999999999998,
+      "grad_norm": 1.2333250045776367,
+      "learning_rate": 2e-05,
+      "loss": 0.04853529,
+      "step": 14351
+    },
+    {
+      "epoch": 28.704,
+      "grad_norm": 1.4717018604278564,
+      "learning_rate": 2e-05,
+      "loss": 0.04341109,
+      "step": 14352
+    },
+    {
+      "epoch": 28.706,
+      "grad_norm": 1.9151333570480347,
+      "learning_rate": 2e-05,
+      "loss": 0.04162496,
+      "step": 14353
+    },
+    {
+      "epoch": 28.708,
+      "grad_norm": 1.7289716005325317,
+      "learning_rate": 2e-05,
+      "loss": 0.04978495,
+      "step": 14354
+    },
+    {
+      "epoch": 28.71,
+      "grad_norm": 1.104732632637024,
+      "learning_rate": 2e-05,
+      "loss": 0.04172861,
+      "step": 14355
+    },
+    {
+      "epoch": 28.712,
+      "grad_norm": 1.1524678468704224,
+      "learning_rate": 2e-05,
+      "loss": 0.03477337,
+      "step": 14356
+    },
+    {
+      "epoch": 28.714,
+      "grad_norm": 1.5342556238174438,
+      "learning_rate": 2e-05,
+      "loss": 0.04556607,
+      "step": 14357
+    },
+    {
+      "epoch": 28.716,
+      "grad_norm": 1.1712950468063354,
+      "learning_rate": 2e-05,
+      "loss": 0.04763829,
+      "step": 14358
+    },
+    {
+      "epoch": 28.718,
+      "grad_norm": 1.091592788696289,
+      "learning_rate": 2e-05,
+      "loss": 0.03833639,
+      "step": 14359
+    },
+    {
+      "epoch": 28.72,
+      "grad_norm": 0.793989360332489,
+      "learning_rate": 2e-05,
+      "loss": 0.025282,
+      "step": 14360
+    },
+    {
+      "epoch": 28.722,
+      "grad_norm": 1.4406577348709106,
+      "learning_rate": 2e-05,
+      "loss": 0.03637164,
+      "step": 14361
+    },
+    {
+      "epoch": 28.724,
+      "grad_norm": 1.957209825515747,
+      "learning_rate": 2e-05,
+      "loss": 0.03893009,
+      "step": 14362
+    },
+    {
+      "epoch": 28.726,
+      "grad_norm": 1.1229599714279175,
+      "learning_rate": 2e-05,
+      "loss": 0.03854847,
+      "step": 14363
+    },
+    {
+      "epoch": 28.728,
+      "grad_norm": 1.9506926536560059,
+      "learning_rate": 2e-05,
+      "loss": 0.03666751,
+      "step": 14364
+    },
+    {
+      "epoch": 28.73,
+      "grad_norm": 1.435692310333252,
+      "learning_rate": 2e-05,
+      "loss": 0.03390378,
+      "step": 14365
+    },
+    {
+      "epoch": 28.732,
+      "grad_norm": 0.955740749835968,
+      "learning_rate": 2e-05,
+      "loss": 0.02934472,
+      "step": 14366
+    },
+    {
+      "epoch": 28.734,
+      "grad_norm": 1.1466535329818726,
+      "learning_rate": 2e-05,
+      "loss": 0.05017526,
+      "step": 14367
+    },
+    {
+      "epoch": 28.736,
+      "grad_norm": 1.218614101409912,
+      "learning_rate": 2e-05,
+      "loss": 0.05039631,
+      "step": 14368
+    },
+    {
+      "epoch": 28.738,
+      "grad_norm": 1.093510389328003,
+      "learning_rate": 2e-05,
+      "loss": 0.03176686,
+      "step": 14369
+    },
+    {
+      "epoch": 28.74,
+      "grad_norm": 1.144060492515564,
+      "learning_rate": 2e-05,
+      "loss": 0.04076057,
+      "step": 14370
+    },
+    {
+      "epoch": 28.742,
+      "grad_norm": 1.5863828659057617,
+      "learning_rate": 2e-05,
+      "loss": 0.04815026,
+      "step": 14371
+    },
+    {
+      "epoch": 28.744,
+      "grad_norm": 0.9931338429450989,
+      "learning_rate": 2e-05,
+      "loss": 0.03469016,
+      "step": 14372
+    },
+    {
+      "epoch": 28.746,
+      "grad_norm": 1.6943480968475342,
+      "learning_rate": 2e-05,
+      "loss": 0.05469589,
+      "step": 14373
+    },
+    {
+      "epoch": 28.748,
+      "grad_norm": 1.2340564727783203,
+      "learning_rate": 2e-05,
+      "loss": 0.04292573,
+      "step": 14374
+    },
+    {
+      "epoch": 28.75,
+      "grad_norm": 1.6473150253295898,
+      "learning_rate": 2e-05,
+      "loss": 0.06950572,
+      "step": 14375
+    },
+    {
+      "epoch": 28.752,
+      "grad_norm": 1.900185465812683,
+      "learning_rate": 2e-05,
+      "loss": 0.05023978,
+      "step": 14376
+    },
+    {
+      "epoch": 28.754,
+      "grad_norm": 1.381726622581482,
+      "learning_rate": 2e-05,
+      "loss": 0.05448072,
+      "step": 14377
+    },
+    {
+      "epoch": 28.756,
+      "grad_norm": 1.0618399381637573,
+      "learning_rate": 2e-05,
+      "loss": 0.03553692,
+      "step": 14378
+    },
+    {
+      "epoch": 28.758,
+      "grad_norm": 1.158182978630066,
+      "learning_rate": 2e-05,
+      "loss": 0.04157444,
+      "step": 14379
+    },
+    {
+      "epoch": 28.76,
+      "grad_norm": 1.0218384265899658,
+      "learning_rate": 2e-05,
+      "loss": 0.03774838,
+      "step": 14380
+    },
+    {
+      "epoch": 28.762,
+      "grad_norm": 1.4672837257385254,
+      "learning_rate": 2e-05,
+      "loss": 0.04652436,
+      "step": 14381
+    },
+    {
+      "epoch": 28.764,
+      "grad_norm": 2.9982128143310547,
+      "learning_rate": 2e-05,
+      "loss": 0.05198572,
+      "step": 14382
+    },
+    {
+      "epoch": 28.766,
+      "grad_norm": 1.5844800472259521,
+      "learning_rate": 2e-05,
+      "loss": 0.04013401,
+      "step": 14383
+    },
+    {
+      "epoch": 28.768,
+      "grad_norm": 2.2007713317871094,
+      "learning_rate": 2e-05,
+      "loss": 0.0491824,
+      "step": 14384
+    },
+    {
+      "epoch": 28.77,
+      "grad_norm": 1.0713263750076294,
+      "learning_rate": 2e-05,
+      "loss": 0.03637439,
+      "step": 14385
+    },
+    {
+      "epoch": 28.772,
+      "grad_norm": 2.0838675498962402,
+      "learning_rate": 2e-05,
+      "loss": 0.04036938,
+      "step": 14386
+    },
+    {
+      "epoch": 28.774,
+      "grad_norm": 1.051755428314209,
+      "learning_rate": 2e-05,
+      "loss": 0.03982811,
+      "step": 14387
+    },
+    {
+      "epoch": 28.776,
+      "grad_norm": 1.3116508722305298,
+      "learning_rate": 2e-05,
+      "loss": 0.04132722,
+      "step": 14388
+    },
+    {
+      "epoch": 28.778,
+      "grad_norm": 1.0713746547698975,
+      "learning_rate": 2e-05,
+      "loss": 0.0432993,
+      "step": 14389
+    },
+    {
+      "epoch": 28.78,
+      "grad_norm": 0.9972354173660278,
+      "learning_rate": 2e-05,
+      "loss": 0.04672929,
+      "step": 14390
+    },
+    {
+      "epoch": 28.782,
+      "grad_norm": 0.9369496703147888,
+      "learning_rate": 2e-05,
+      "loss": 0.02907495,
+      "step": 14391
+    },
+    {
+      "epoch": 28.784,
+      "grad_norm": 1.1444287300109863,
+      "learning_rate": 2e-05,
+      "loss": 0.03748661,
+      "step": 14392
+    },
+    {
+      "epoch": 28.786,
+      "grad_norm": 1.1183111667633057,
+      "learning_rate": 2e-05,
+      "loss": 0.02960247,
+      "step": 14393
+    },
+    {
+      "epoch": 28.788,
+      "grad_norm": 1.0258089303970337,
+      "learning_rate": 2e-05,
+      "loss": 0.03007736,
+      "step": 14394
+    },
+    {
+      "epoch": 28.79,
+      "grad_norm": 0.9938691258430481,
+      "learning_rate": 2e-05,
+      "loss": 0.03597631,
+      "step": 14395
+    },
+    {
+      "epoch": 28.792,
+      "grad_norm": 1.1507741212844849,
+      "learning_rate": 2e-05,
+      "loss": 0.04170668,
+      "step": 14396
+    },
+    {
+      "epoch": 28.794,
+      "grad_norm": 3.4133851528167725,
+      "learning_rate": 2e-05,
+      "loss": 0.04272288,
+      "step": 14397
+    },
+    {
+      "epoch": 28.796,
+      "grad_norm": 1.7486724853515625,
+      "learning_rate": 2e-05,
+      "loss": 0.06500375,
+      "step": 14398
+    },
+    {
+      "epoch": 28.798000000000002,
+      "grad_norm": 1.4064818620681763,
+      "learning_rate": 2e-05,
+      "loss": 0.04955719,
+      "step": 14399
+    },
+    {
+      "epoch": 28.8,
+      "grad_norm": 1.6619688272476196,
+      "learning_rate": 2e-05,
+      "loss": 0.04951192,
+      "step": 14400
+    },
+    {
+      "epoch": 28.802,
+      "grad_norm": 1.1675795316696167,
+      "learning_rate": 2e-05,
+      "loss": 0.037157,
+      "step": 14401
+    },
+    {
+      "epoch": 28.804,
+      "grad_norm": 1.6489543914794922,
+      "learning_rate": 2e-05,
+      "loss": 0.04647279,
+      "step": 14402
+    },
+    {
+      "epoch": 28.806,
+      "grad_norm": 1.7396496534347534,
+      "learning_rate": 2e-05,
+      "loss": 0.04493985,
+      "step": 14403
+    },
+    {
+      "epoch": 28.808,
+      "grad_norm": 1.156243920326233,
+      "learning_rate": 2e-05,
+      "loss": 0.04273091,
+      "step": 14404
+    },
+    {
+      "epoch": 28.81,
+      "grad_norm": 1.0005199909210205,
+      "learning_rate": 2e-05,
+      "loss": 0.03280544,
+      "step": 14405
+    },
+    {
+      "epoch": 28.812,
+      "grad_norm": 3.4301085472106934,
+      "learning_rate": 2e-05,
+      "loss": 0.03465569,
+      "step": 14406
+    },
+    {
+      "epoch": 28.814,
+      "grad_norm": 1.50563383102417,
+      "learning_rate": 2e-05,
+      "loss": 0.07086978,
+      "step": 14407
+    },
+    {
+      "epoch": 28.816,
+      "grad_norm": 1.5838433504104614,
+      "learning_rate": 2e-05,
+      "loss": 0.03891544,
+      "step": 14408
+    },
+    {
+      "epoch": 28.818,
+      "grad_norm": 1.1893781423568726,
+      "learning_rate": 2e-05,
+      "loss": 0.04945404,
+      "step": 14409
+    },
+    {
+      "epoch": 28.82,
+      "grad_norm": 0.8621784448623657,
+      "learning_rate": 2e-05,
+      "loss": 0.03163891,
+      "step": 14410
+    },
+    {
+      "epoch": 28.822,
+      "grad_norm": 1.1594724655151367,
+      "learning_rate": 2e-05,
+      "loss": 0.05155197,
+      "step": 14411
+    },
+    {
+      "epoch": 28.824,
+      "grad_norm": 1.443181037902832,
+      "learning_rate": 2e-05,
+      "loss": 0.05630941,
+      "step": 14412
+    },
+    {
+      "epoch": 28.826,
+      "grad_norm": 1.5907959938049316,
+      "learning_rate": 2e-05,
+      "loss": 0.04138833,
+      "step": 14413
+    },
+    {
+      "epoch": 28.828,
+      "grad_norm": 1.6216257810592651,
+      "learning_rate": 2e-05,
+      "loss": 0.04543457,
+      "step": 14414
+    },
+    {
+      "epoch": 28.83,
+      "grad_norm": 1.4624403715133667,
+      "learning_rate": 2e-05,
+      "loss": 0.05380312,
+      "step": 14415
+    },
+    {
+      "epoch": 28.832,
+      "grad_norm": 1.0963571071624756,
+      "learning_rate": 2e-05,
+      "loss": 0.04702737,
+      "step": 14416
+    },
+    {
+      "epoch": 28.834,
+      "grad_norm": 1.7015491724014282,
+      "learning_rate": 2e-05,
+      "loss": 0.03770491,
+      "step": 14417
+    },
+    {
+      "epoch": 28.836,
+      "grad_norm": 1.20378839969635,
+      "learning_rate": 2e-05,
+      "loss": 0.04444302,
+      "step": 14418
+    },
+    {
+      "epoch": 28.838,
+      "grad_norm": 1.72189199924469,
+      "learning_rate": 2e-05,
+      "loss": 0.05816022,
+      "step": 14419
+    },
+    {
+      "epoch": 28.84,
+      "grad_norm": 1.1528189182281494,
+      "learning_rate": 2e-05,
+      "loss": 0.03968474,
+      "step": 14420
+    },
+    {
+      "epoch": 28.842,
+      "grad_norm": 1.2541671991348267,
+      "learning_rate": 2e-05,
+      "loss": 0.03849109,
+      "step": 14421
+    },
+    {
+      "epoch": 28.844,
+      "grad_norm": 2.8099963665008545,
+      "learning_rate": 2e-05,
+      "loss": 0.05084788,
+      "step": 14422
+    },
+    {
+      "epoch": 28.846,
+      "grad_norm": 0.9781857132911682,
+      "learning_rate": 2e-05,
+      "loss": 0.0285262,
+      "step": 14423
+    },
+    {
+      "epoch": 28.848,
+      "grad_norm": 1.4903124570846558,
+      "learning_rate": 2e-05,
+      "loss": 0.05180159,
+      "step": 14424
+    },
+    {
+      "epoch": 28.85,
+      "grad_norm": 1.7416338920593262,
+      "learning_rate": 2e-05,
+      "loss": 0.03428745,
+      "step": 14425
+    },
+    {
+      "epoch": 28.852,
+      "grad_norm": 1.521485686302185,
+      "learning_rate": 2e-05,
+      "loss": 0.04813501,
+      "step": 14426
+    },
+    {
+      "epoch": 28.854,
+      "grad_norm": 1.2684760093688965,
+      "learning_rate": 2e-05,
+      "loss": 0.04741008,
+      "step": 14427
+    },
+    {
+      "epoch": 28.856,
+      "grad_norm": 1.1217941045761108,
+      "learning_rate": 2e-05,
+      "loss": 0.04172241,
+      "step": 14428
+    },
+    {
+      "epoch": 28.858,
+      "grad_norm": 1.3092353343963623,
+      "learning_rate": 2e-05,
+      "loss": 0.04935054,
+      "step": 14429
+    },
+    {
+      "epoch": 28.86,
+      "grad_norm": 1.0041587352752686,
+      "learning_rate": 2e-05,
+      "loss": 0.03718605,
+      "step": 14430
+    },
+    {
+      "epoch": 28.862,
+      "grad_norm": 2.0443413257598877,
+      "learning_rate": 2e-05,
+      "loss": 0.04975209,
+      "step": 14431
+    },
+    {
+      "epoch": 28.864,
+      "grad_norm": 0.8965648412704468,
+      "learning_rate": 2e-05,
+      "loss": 0.03016404,
+      "step": 14432
+    },
+    {
+      "epoch": 28.866,
+      "grad_norm": 1.8447829484939575,
+      "learning_rate": 2e-05,
+      "loss": 0.05125307,
+      "step": 14433
+    },
+    {
+      "epoch": 28.868,
+      "grad_norm": 1.2474303245544434,
+      "learning_rate": 2e-05,
+      "loss": 0.04687266,
+      "step": 14434
+    },
+    {
+      "epoch": 28.87,
+      "grad_norm": 1.745761752128601,
+      "learning_rate": 2e-05,
+      "loss": 0.07663966,
+      "step": 14435
+    },
+    {
+      "epoch": 28.872,
+      "grad_norm": 1.3358268737792969,
+      "learning_rate": 2e-05,
+      "loss": 0.04147369,
+      "step": 14436
+    },
+    {
+      "epoch": 28.874,
+      "grad_norm": 1.2408902645111084,
+      "learning_rate": 2e-05,
+      "loss": 0.04460159,
+      "step": 14437
+    },
+    {
+      "epoch": 28.876,
+      "grad_norm": 1.9124995470046997,
+      "learning_rate": 2e-05,
+      "loss": 0.05355459,
+      "step": 14438
+    },
+    {
+      "epoch": 28.878,
+      "grad_norm": 1.334394931793213,
+      "learning_rate": 2e-05,
+      "loss": 0.04897958,
+      "step": 14439
+    },
+    {
+      "epoch": 28.88,
+      "grad_norm": 1.8889309167861938,
+      "learning_rate": 2e-05,
+      "loss": 0.04751597,
+      "step": 14440
+    },
+    {
+      "epoch": 28.882,
+      "grad_norm": 1.1469106674194336,
+      "learning_rate": 2e-05,
+      "loss": 0.04521649,
+      "step": 14441
+    },
+    {
+      "epoch": 28.884,
+      "grad_norm": 1.0258030891418457,
+      "learning_rate": 2e-05,
+      "loss": 0.04085048,
+      "step": 14442
+    },
+    {
+      "epoch": 28.886,
+      "grad_norm": 2.1602070331573486,
+      "learning_rate": 2e-05,
+      "loss": 0.05456547,
+      "step": 14443
+    },
+    {
+      "epoch": 28.888,
+      "grad_norm": 1.524979591369629,
+      "learning_rate": 2e-05,
+      "loss": 0.04410193,
+      "step": 14444
+    },
+    {
+      "epoch": 28.89,
+      "grad_norm": 1.0402649641036987,
+      "learning_rate": 2e-05,
+      "loss": 0.03135321,
+      "step": 14445
+    },
+    {
+      "epoch": 28.892,
+      "grad_norm": 1.7973295450210571,
+      "learning_rate": 2e-05,
+      "loss": 0.04395194,
+      "step": 14446
+    },
+    {
+      "epoch": 28.894,
+      "grad_norm": 1.0734200477600098,
+      "learning_rate": 2e-05,
+      "loss": 0.0351134,
+      "step": 14447
+    },
+    {
+      "epoch": 28.896,
+      "grad_norm": 1.271427035331726,
+      "learning_rate": 2e-05,
+      "loss": 0.0497748,
+      "step": 14448
+    },
+    {
+      "epoch": 28.898,
+      "grad_norm": 0.9508517384529114,
+      "learning_rate": 2e-05,
+      "loss": 0.03418116,
+      "step": 14449
+    },
+    {
+      "epoch": 28.9,
+      "grad_norm": 2.287792921066284,
+      "learning_rate": 2e-05,
+      "loss": 0.04537106,
+      "step": 14450
+    },
+    {
+      "epoch": 28.902,
+      "grad_norm": 2.9237358570098877,
+      "learning_rate": 2e-05,
+      "loss": 0.06133288,
+      "step": 14451
+    },
+    {
+      "epoch": 28.904,
+      "grad_norm": 1.5036847591400146,
+      "learning_rate": 2e-05,
+      "loss": 0.04737253,
+      "step": 14452
+    },
+    {
+      "epoch": 28.906,
+      "grad_norm": 1.2241439819335938,
+      "learning_rate": 2e-05,
+      "loss": 0.03472542,
+      "step": 14453
+    },
+    {
+      "epoch": 28.908,
+      "grad_norm": 1.3582462072372437,
+      "learning_rate": 2e-05,
+      "loss": 0.05366043,
+      "step": 14454
+    },
+    {
+      "epoch": 28.91,
+      "grad_norm": 1.204209327697754,
+      "learning_rate": 2e-05,
+      "loss": 0.04971019,
+      "step": 14455
+    },
+    {
+      "epoch": 28.912,
+      "grad_norm": 1.7010318040847778,
+      "learning_rate": 2e-05,
+      "loss": 0.05780742,
+      "step": 14456
+    },
+    {
+      "epoch": 28.914,
+      "grad_norm": 1.0974135398864746,
+      "learning_rate": 2e-05,
+      "loss": 0.03613953,
+      "step": 14457
+    },
+    {
+      "epoch": 28.916,
+      "grad_norm": 1.595192551612854,
+      "learning_rate": 2e-05,
+      "loss": 0.04265611,
+      "step": 14458
+    },
+    {
+      "epoch": 28.918,
+      "grad_norm": 1.5996463298797607,
+      "learning_rate": 2e-05,
+      "loss": 0.03614064,
+      "step": 14459
+    },
+    {
+      "epoch": 28.92,
+      "grad_norm": 0.9934185743331909,
+      "learning_rate": 2e-05,
+      "loss": 0.03452392,
+      "step": 14460
+    },
+    {
+      "epoch": 28.922,
+      "grad_norm": 1.1480003595352173,
+      "learning_rate": 2e-05,
+      "loss": 0.03976715,
+      "step": 14461
+    },
+    {
+      "epoch": 28.924,
+      "grad_norm": 0.7937719225883484,
+      "learning_rate": 2e-05,
+      "loss": 0.03106033,
+      "step": 14462
+    },
+    {
+      "epoch": 28.926,
+      "grad_norm": 0.9472975134849548,
+      "learning_rate": 2e-05,
+      "loss": 0.03592245,
+      "step": 14463
+    },
+    {
+      "epoch": 28.928,
+      "grad_norm": 1.2322276830673218,
+      "learning_rate": 2e-05,
+      "loss": 0.06219757,
+      "step": 14464
+    },
+    {
+      "epoch": 28.93,
+      "grad_norm": 1.1841628551483154,
+      "learning_rate": 2e-05,
+      "loss": 0.02757031,
+      "step": 14465
+    },
+    {
+      "epoch": 28.932,
+      "grad_norm": 2.2316744327545166,
+      "learning_rate": 2e-05,
+      "loss": 0.05290478,
+      "step": 14466
+    },
+    {
+      "epoch": 28.934,
+      "grad_norm": 1.1691445112228394,
+      "learning_rate": 2e-05,
+      "loss": 0.04423434,
+      "step": 14467
+    },
+    {
+      "epoch": 28.936,
+      "grad_norm": 1.2703580856323242,
+      "learning_rate": 2e-05,
+      "loss": 0.04534604,
+      "step": 14468
+    },
+    {
+      "epoch": 28.938,
+      "grad_norm": 1.8662946224212646,
+      "learning_rate": 2e-05,
+      "loss": 0.0505904,
+      "step": 14469
+    },
+    {
+      "epoch": 28.94,
+      "grad_norm": 1.984939455986023,
+      "learning_rate": 2e-05,
+      "loss": 0.04816742,
+      "step": 14470
+    },
+    {
+      "epoch": 28.942,
+      "grad_norm": 1.6010085344314575,
+      "learning_rate": 2e-05,
+      "loss": 0.03411138,
+      "step": 14471
+    },
+    {
+      "epoch": 28.944,
+      "grad_norm": 1.3114997148513794,
+      "learning_rate": 2e-05,
+      "loss": 0.03839431,
+      "step": 14472
+    },
+    {
+      "epoch": 28.946,
+      "grad_norm": 1.9743845462799072,
+      "learning_rate": 2e-05,
+      "loss": 0.06387973,
+      "step": 14473
+    },
+    {
+      "epoch": 28.948,
+      "grad_norm": 3.784247398376465,
+      "learning_rate": 2e-05,
+      "loss": 0.05783255,
+      "step": 14474
+    },
+    {
+      "epoch": 28.95,
+      "grad_norm": 1.5012797117233276,
+      "learning_rate": 2e-05,
+      "loss": 0.04049198,
+      "step": 14475
+    },
+    {
+      "epoch": 28.951999999999998,
+      "grad_norm": 0.9062674641609192,
+      "learning_rate": 2e-05,
+      "loss": 0.0387709,
+      "step": 14476
+    },
+    {
+      "epoch": 28.954,
+      "grad_norm": 1.216447353363037,
+      "learning_rate": 2e-05,
+      "loss": 0.03773759,
+      "step": 14477
+    },
+    {
+      "epoch": 28.956,
+      "grad_norm": 2.346182346343994,
+      "learning_rate": 2e-05,
+      "loss": 0.04593071,
+      "step": 14478
+    },
+    {
+      "epoch": 28.958,
+      "grad_norm": 2.298583984375,
+      "learning_rate": 2e-05,
+      "loss": 0.05919803,
+      "step": 14479
+    },
+    {
+      "epoch": 28.96,
+      "grad_norm": 0.9582846164703369,
+      "learning_rate": 2e-05,
+      "loss": 0.02871491,
+      "step": 14480
+    },
+    {
+      "epoch": 28.962,
+      "grad_norm": 0.7866615056991577,
+      "learning_rate": 2e-05,
+      "loss": 0.02259299,
+      "step": 14481
+    },
+    {
+      "epoch": 28.964,
+      "grad_norm": 1.4925915002822876,
+      "learning_rate": 2e-05,
+      "loss": 0.04916498,
+      "step": 14482
+    },
+    {
+      "epoch": 28.966,
+      "grad_norm": 1.4005680084228516,
+      "learning_rate": 2e-05,
+      "loss": 0.05168829,
+      "step": 14483
+    },
+    {
+      "epoch": 28.968,
+      "grad_norm": 4.166713237762451,
+      "learning_rate": 2e-05,
+      "loss": 0.04969958,
+      "step": 14484
+    },
+    {
+      "epoch": 28.97,
+      "grad_norm": 1.1008797883987427,
+      "learning_rate": 2e-05,
+      "loss": 0.03930159,
+      "step": 14485
+    },
+    {
+      "epoch": 28.972,
+      "grad_norm": 1.0620421171188354,
+      "learning_rate": 2e-05,
+      "loss": 0.03718957,
+      "step": 14486
+    },
+    {
+      "epoch": 28.974,
+      "grad_norm": 1.1578391790390015,
+      "learning_rate": 2e-05,
+      "loss": 0.04585633,
+      "step": 14487
+    },
+    {
+      "epoch": 28.976,
+      "grad_norm": 1.48446524143219,
+      "learning_rate": 2e-05,
+      "loss": 0.0247959,
+      "step": 14488
+    },
+    {
+      "epoch": 28.978,
+      "grad_norm": 1.0976674556732178,
+      "learning_rate": 2e-05,
+      "loss": 0.03197834,
+      "step": 14489
+    },
+    {
+      "epoch": 28.98,
+      "grad_norm": 1.139172077178955,
+      "learning_rate": 2e-05,
+      "loss": 0.04646052,
+      "step": 14490
+    },
+    {
+      "epoch": 28.982,
+      "grad_norm": 1.9785137176513672,
+      "learning_rate": 2e-05,
+      "loss": 0.05109739,
+      "step": 14491
+    },
+    {
+      "epoch": 28.984,
+      "grad_norm": 1.3987001180648804,
+      "learning_rate": 2e-05,
+      "loss": 0.0440822,
+      "step": 14492
+    },
+    {
+      "epoch": 28.986,
+      "grad_norm": 1.220787763595581,
+      "learning_rate": 2e-05,
+      "loss": 0.03669748,
+      "step": 14493
+    },
+    {
+      "epoch": 28.988,
+      "grad_norm": 2.083451986312866,
+      "learning_rate": 2e-05,
+      "loss": 0.05152371,
+      "step": 14494
+    },
+    {
+      "epoch": 28.99,
+      "grad_norm": 1.1273852586746216,
+      "learning_rate": 2e-05,
+      "loss": 0.03859162,
+      "step": 14495
+    },
+    {
+      "epoch": 28.992,
+      "grad_norm": 1.344372034072876,
+      "learning_rate": 2e-05,
+      "loss": 0.03838318,
+      "step": 14496
+    },
+    {
+      "epoch": 28.994,
+      "grad_norm": 1.728652000427246,
+      "learning_rate": 2e-05,
+      "loss": 0.03777511,
+      "step": 14497
+    },
+    {
+      "epoch": 28.996,
+      "grad_norm": 3.5054938793182373,
+      "learning_rate": 2e-05,
+      "loss": 0.04964066,
+      "step": 14498
+    },
+    {
+      "epoch": 28.998,
+      "grad_norm": 1.501190185546875,
+      "learning_rate": 2e-05,
+      "loss": 0.06090293,
+      "step": 14499
+    },
+    {
+      "epoch": 29.0,
+      "grad_norm": 1.227286696434021,
+      "learning_rate": 2e-05,
+      "loss": 0.04943068,
+      "step": 14500
+    },
+    {
+      "epoch": 29.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9780439121756487,
+        "Equal_1": 0.998,
+        "Equal_2": 0.9780439121756487,
+        "Equal_3": 0.9880239520958084,
+        "LineComparison_1": 0.998,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9960079840319361,
+        "Parallel_1": 0.9959919839679359,
+        "Parallel_2": 0.9979959919839679,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.984,
+        "Perpendicular_3": 0.8897795591182365,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9936666666666667,
+        "PointLiesOnCircle_3": 0.9876,
+        "PointLiesOnLine_1": 0.9899799599198397,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9820359281437125
+      },
+      "eval_runtime": 320.129,
+      "eval_samples_per_second": 32.799,
+      "eval_steps_per_second": 0.656,
+      "step": 14500
+    },
+    {
+      "epoch": 29.002,
+      "grad_norm": 1.3614428043365479,
+      "learning_rate": 2e-05,
+      "loss": 0.04904348,
+      "step": 14501
+    },
+    {
+      "epoch": 29.004,
+      "grad_norm": 0.990894079208374,
+      "learning_rate": 2e-05,
+      "loss": 0.03972761,
+      "step": 14502
+    },
+    {
+      "epoch": 29.006,
+      "grad_norm": 1.5582256317138672,
+      "learning_rate": 2e-05,
+      "loss": 0.04734577,
+      "step": 14503
+    },
+    {
+      "epoch": 29.008,
+      "grad_norm": 1.598989725112915,
+      "learning_rate": 2e-05,
+      "loss": 0.04445969,
+      "step": 14504
+    },
+    {
+      "epoch": 29.01,
+      "grad_norm": 1.4012812376022339,
+      "learning_rate": 2e-05,
+      "loss": 0.05144128,
+      "step": 14505
+    },
+    {
+      "epoch": 29.012,
+      "grad_norm": 1.633912444114685,
+      "learning_rate": 2e-05,
+      "loss": 0.04580737,
+      "step": 14506
+    },
+    {
+      "epoch": 29.014,
+      "grad_norm": 1.3634629249572754,
+      "learning_rate": 2e-05,
+      "loss": 0.03885691,
+      "step": 14507
+    },
+    {
+      "epoch": 29.016,
+      "grad_norm": 1.5107965469360352,
+      "learning_rate": 2e-05,
+      "loss": 0.05544426,
+      "step": 14508
+    },
+    {
+      "epoch": 29.018,
+      "grad_norm": 1.50559401512146,
+      "learning_rate": 2e-05,
+      "loss": 0.03723505,
+      "step": 14509
+    },
+    {
+      "epoch": 29.02,
+      "grad_norm": 1.2116549015045166,
+      "learning_rate": 2e-05,
+      "loss": 0.06099574,
+      "step": 14510
+    },
+    {
+      "epoch": 29.022,
+      "grad_norm": 1.4817736148834229,
+      "learning_rate": 2e-05,
+      "loss": 0.04306397,
+      "step": 14511
+    },
+    {
+      "epoch": 29.024,
+      "grad_norm": 1.6792289018630981,
+      "learning_rate": 2e-05,
+      "loss": 0.0461472,
+      "step": 14512
+    },
+    {
+      "epoch": 29.026,
+      "grad_norm": 1.1383700370788574,
+      "learning_rate": 2e-05,
+      "loss": 0.0400393,
+      "step": 14513
+    },
+    {
+      "epoch": 29.028,
+      "grad_norm": 1.6050753593444824,
+      "learning_rate": 2e-05,
+      "loss": 0.0388446,
+      "step": 14514
+    },
+    {
+      "epoch": 29.03,
+      "grad_norm": 2.475426435470581,
+      "learning_rate": 2e-05,
+      "loss": 0.04789076,
+      "step": 14515
+    },
+    {
+      "epoch": 29.032,
+      "grad_norm": 1.7298914194107056,
+      "learning_rate": 2e-05,
+      "loss": 0.04865304,
+      "step": 14516
+    },
+    {
+      "epoch": 29.034,
+      "grad_norm": 1.0967127084732056,
+      "learning_rate": 2e-05,
+      "loss": 0.02864125,
+      "step": 14517
+    },
+    {
+      "epoch": 29.036,
+      "grad_norm": 2.11590838432312,
+      "learning_rate": 2e-05,
+      "loss": 0.03580888,
+      "step": 14518
+    },
+    {
+      "epoch": 29.038,
+      "grad_norm": 1.135114073753357,
+      "learning_rate": 2e-05,
+      "loss": 0.04573719,
+      "step": 14519
+    },
+    {
+      "epoch": 29.04,
+      "grad_norm": 1.1727830171585083,
+      "learning_rate": 2e-05,
+      "loss": 0.04438477,
+      "step": 14520
+    },
+    {
+      "epoch": 29.042,
+      "grad_norm": 1.3196028470993042,
+      "learning_rate": 2e-05,
+      "loss": 0.04897299,
+      "step": 14521
+    },
+    {
+      "epoch": 29.044,
+      "grad_norm": 1.5784796476364136,
+      "learning_rate": 2e-05,
+      "loss": 0.06415123,
+      "step": 14522
+    },
+    {
+      "epoch": 29.046,
+      "grad_norm": 1.6042543649673462,
+      "learning_rate": 2e-05,
+      "loss": 0.03841857,
+      "step": 14523
+    },
+    {
+      "epoch": 29.048,
+      "grad_norm": 1.3283536434173584,
+      "learning_rate": 2e-05,
+      "loss": 0.06226228,
+      "step": 14524
+    },
+    {
+      "epoch": 29.05,
+      "grad_norm": 2.3464431762695312,
+      "learning_rate": 2e-05,
+      "loss": 0.05047053,
+      "step": 14525
+    },
+    {
+      "epoch": 29.052,
+      "grad_norm": 1.370872139930725,
+      "learning_rate": 2e-05,
+      "loss": 0.03550132,
+      "step": 14526
+    },
+    {
+      "epoch": 29.054,
+      "grad_norm": 1.3238565921783447,
+      "learning_rate": 2e-05,
+      "loss": 0.0407903,
+      "step": 14527
+    },
+    {
+      "epoch": 29.056,
+      "grad_norm": 1.16240656375885,
+      "learning_rate": 2e-05,
+      "loss": 0.03744307,
+      "step": 14528
+    },
+    {
+      "epoch": 29.058,
+      "grad_norm": 1.2564165592193604,
+      "learning_rate": 2e-05,
+      "loss": 0.03232381,
+      "step": 14529
+    },
+    {
+      "epoch": 29.06,
+      "grad_norm": 1.1828256845474243,
+      "learning_rate": 2e-05,
+      "loss": 0.04936057,
+      "step": 14530
+    },
+    {
+      "epoch": 29.062,
+      "grad_norm": 1.9137167930603027,
+      "learning_rate": 2e-05,
+      "loss": 0.05838963,
+      "step": 14531
+    },
+    {
+      "epoch": 29.064,
+      "grad_norm": 1.2713327407836914,
+      "learning_rate": 2e-05,
+      "loss": 0.04713174,
+      "step": 14532
+    },
+    {
+      "epoch": 29.066,
+      "grad_norm": 1.812070369720459,
+      "learning_rate": 2e-05,
+      "loss": 0.0491786,
+      "step": 14533
+    },
+    {
+      "epoch": 29.068,
+      "grad_norm": 1.9103559255599976,
+      "learning_rate": 2e-05,
+      "loss": 0.04707557,
+      "step": 14534
+    },
+    {
+      "epoch": 29.07,
+      "grad_norm": 1.211613655090332,
+      "learning_rate": 2e-05,
+      "loss": 0.0457197,
+      "step": 14535
+    },
+    {
+      "epoch": 29.072,
+      "grad_norm": 1.562606692314148,
+      "learning_rate": 2e-05,
+      "loss": 0.05934884,
+      "step": 14536
+    },
+    {
+      "epoch": 29.074,
+      "grad_norm": 1.6958134174346924,
+      "learning_rate": 2e-05,
+      "loss": 0.04679252,
+      "step": 14537
+    },
+    {
+      "epoch": 29.076,
+      "grad_norm": 1.0809810161590576,
+      "learning_rate": 2e-05,
+      "loss": 0.04069981,
+      "step": 14538
+    },
+    {
+      "epoch": 29.078,
+      "grad_norm": 1.0606014728546143,
+      "learning_rate": 2e-05,
+      "loss": 0.03034795,
+      "step": 14539
+    },
+    {
+      "epoch": 29.08,
+      "grad_norm": 1.3464914560317993,
+      "learning_rate": 2e-05,
+      "loss": 0.03704887,
+      "step": 14540
+    },
+    {
+      "epoch": 29.082,
+      "grad_norm": 2.9524738788604736,
+      "learning_rate": 2e-05,
+      "loss": 0.06308039,
+      "step": 14541
+    },
+    {
+      "epoch": 29.084,
+      "grad_norm": 1.7321432828903198,
+      "learning_rate": 2e-05,
+      "loss": 0.06781643,
+      "step": 14542
+    },
+    {
+      "epoch": 29.086,
+      "grad_norm": 1.9871220588684082,
+      "learning_rate": 2e-05,
+      "loss": 0.0448276,
+      "step": 14543
+    },
+    {
+      "epoch": 29.088,
+      "grad_norm": 1.0521864891052246,
+      "learning_rate": 2e-05,
+      "loss": 0.04277543,
+      "step": 14544
+    },
+    {
+      "epoch": 29.09,
+      "grad_norm": 1.6437976360321045,
+      "learning_rate": 2e-05,
+      "loss": 0.05797031,
+      "step": 14545
+    },
+    {
+      "epoch": 29.092,
+      "grad_norm": 1.883757472038269,
+      "learning_rate": 2e-05,
+      "loss": 0.03330542,
+      "step": 14546
+    },
+    {
+      "epoch": 29.094,
+      "grad_norm": 1.5868852138519287,
+      "learning_rate": 2e-05,
+      "loss": 0.05670642,
+      "step": 14547
+    },
+    {
+      "epoch": 29.096,
+      "grad_norm": 1.4247984886169434,
+      "learning_rate": 2e-05,
+      "loss": 0.04870617,
+      "step": 14548
+    },
+    {
+      "epoch": 29.098,
+      "grad_norm": 1.7102012634277344,
+      "learning_rate": 2e-05,
+      "loss": 0.0426238,
+      "step": 14549
+    },
+    {
+      "epoch": 29.1,
+      "grad_norm": 1.3921092748641968,
+      "learning_rate": 2e-05,
+      "loss": 0.06138652,
+      "step": 14550
+    },
+    {
+      "epoch": 29.102,
+      "grad_norm": 1.0594866275787354,
+      "learning_rate": 2e-05,
+      "loss": 0.0388628,
+      "step": 14551
+    },
+    {
+      "epoch": 29.104,
+      "grad_norm": 1.6638845205307007,
+      "learning_rate": 2e-05,
+      "loss": 0.04031256,
+      "step": 14552
+    },
+    {
+      "epoch": 29.106,
+      "grad_norm": 1.3251597881317139,
+      "learning_rate": 2e-05,
+      "loss": 0.0368821,
+      "step": 14553
+    },
+    {
+      "epoch": 29.108,
+      "grad_norm": 1.3504347801208496,
+      "learning_rate": 2e-05,
+      "loss": 0.04997741,
+      "step": 14554
+    },
+    {
+      "epoch": 29.11,
+      "grad_norm": 1.2452493906021118,
+      "learning_rate": 2e-05,
+      "loss": 0.04492478,
+      "step": 14555
+    },
+    {
+      "epoch": 29.112,
+      "grad_norm": 1.7179110050201416,
+      "learning_rate": 2e-05,
+      "loss": 0.06608713,
+      "step": 14556
+    },
+    {
+      "epoch": 29.114,
+      "grad_norm": 0.9290695190429688,
+      "learning_rate": 2e-05,
+      "loss": 0.02252526,
+      "step": 14557
+    },
+    {
+      "epoch": 29.116,
+      "grad_norm": 1.4510340690612793,
+      "learning_rate": 2e-05,
+      "loss": 0.03936088,
+      "step": 14558
+    },
+    {
+      "epoch": 29.118,
+      "grad_norm": 1.056706428527832,
+      "learning_rate": 2e-05,
+      "loss": 0.04137318,
+      "step": 14559
+    },
+    {
+      "epoch": 29.12,
+      "grad_norm": 1.5099283456802368,
+      "learning_rate": 2e-05,
+      "loss": 0.05786295,
+      "step": 14560
+    },
+    {
+      "epoch": 29.122,
+      "grad_norm": 1.028210997581482,
+      "learning_rate": 2e-05,
+      "loss": 0.03324694,
+      "step": 14561
+    },
+    {
+      "epoch": 29.124,
+      "grad_norm": 1.200790524482727,
+      "learning_rate": 2e-05,
+      "loss": 0.0418604,
+      "step": 14562
+    },
+    {
+      "epoch": 29.126,
+      "grad_norm": 1.2720760107040405,
+      "learning_rate": 2e-05,
+      "loss": 0.05241677,
+      "step": 14563
+    },
+    {
+      "epoch": 29.128,
+      "grad_norm": 1.2256773710250854,
+      "learning_rate": 2e-05,
+      "loss": 0.05769563,
+      "step": 14564
+    },
+    {
+      "epoch": 29.13,
+      "grad_norm": 1.125728964805603,
+      "learning_rate": 2e-05,
+      "loss": 0.03989631,
+      "step": 14565
+    },
+    {
+      "epoch": 29.132,
+      "grad_norm": 3.5522632598876953,
+      "learning_rate": 2e-05,
+      "loss": 0.04670805,
+      "step": 14566
+    },
+    {
+      "epoch": 29.134,
+      "grad_norm": 1.2549570798873901,
+      "learning_rate": 2e-05,
+      "loss": 0.03625032,
+      "step": 14567
+    },
+    {
+      "epoch": 29.136,
+      "grad_norm": 1.3453484773635864,
+      "learning_rate": 2e-05,
+      "loss": 0.04612465,
+      "step": 14568
+    },
+    {
+      "epoch": 29.138,
+      "grad_norm": 1.3071088790893555,
+      "learning_rate": 2e-05,
+      "loss": 0.04188219,
+      "step": 14569
+    },
+    {
+      "epoch": 29.14,
+      "grad_norm": 1.373300313949585,
+      "learning_rate": 2e-05,
+      "loss": 0.05468735,
+      "step": 14570
+    },
+    {
+      "epoch": 29.142,
+      "grad_norm": 1.1056849956512451,
+      "learning_rate": 2e-05,
+      "loss": 0.04801034,
+      "step": 14571
+    },
+    {
+      "epoch": 29.144,
+      "grad_norm": 2.6124298572540283,
+      "learning_rate": 2e-05,
+      "loss": 0.06286603,
+      "step": 14572
+    },
+    {
+      "epoch": 29.146,
+      "grad_norm": 1.090798020362854,
+      "learning_rate": 2e-05,
+      "loss": 0.04604357,
+      "step": 14573
+    },
+    {
+      "epoch": 29.148,
+      "grad_norm": 1.4949955940246582,
+      "learning_rate": 2e-05,
+      "loss": 0.05045891,
+      "step": 14574
+    },
+    {
+      "epoch": 29.15,
+      "grad_norm": 1.3205533027648926,
+      "learning_rate": 2e-05,
+      "loss": 0.04018179,
+      "step": 14575
+    },
+    {
+      "epoch": 29.152,
+      "grad_norm": 1.686370611190796,
+      "learning_rate": 2e-05,
+      "loss": 0.03889347,
+      "step": 14576
+    },
+    {
+      "epoch": 29.154,
+      "grad_norm": 0.989204466342926,
+      "learning_rate": 2e-05,
+      "loss": 0.04520692,
+      "step": 14577
+    },
+    {
+      "epoch": 29.156,
+      "grad_norm": 1.28013014793396,
+      "learning_rate": 2e-05,
+      "loss": 0.04326824,
+      "step": 14578
+    },
+    {
+      "epoch": 29.158,
+      "grad_norm": 1.1764912605285645,
+      "learning_rate": 2e-05,
+      "loss": 0.03642344,
+      "step": 14579
+    },
+    {
+      "epoch": 29.16,
+      "grad_norm": 1.8132061958312988,
+      "learning_rate": 2e-05,
+      "loss": 0.05279174,
+      "step": 14580
+    },
+    {
+      "epoch": 29.162,
+      "grad_norm": 1.478899359703064,
+      "learning_rate": 2e-05,
+      "loss": 0.05297806,
+      "step": 14581
+    },
+    {
+      "epoch": 29.164,
+      "grad_norm": 3.247129201889038,
+      "learning_rate": 2e-05,
+      "loss": 0.05896338,
+      "step": 14582
+    },
+    {
+      "epoch": 29.166,
+      "grad_norm": 1.3242168426513672,
+      "learning_rate": 2e-05,
+      "loss": 0.03942503,
+      "step": 14583
+    },
+    {
+      "epoch": 29.168,
+      "grad_norm": 1.1935416460037231,
+      "learning_rate": 2e-05,
+      "loss": 0.04639933,
+      "step": 14584
+    },
+    {
+      "epoch": 29.17,
+      "grad_norm": 1.0874605178833008,
+      "learning_rate": 2e-05,
+      "loss": 0.04434571,
+      "step": 14585
+    },
+    {
+      "epoch": 29.172,
+      "grad_norm": 1.1143393516540527,
+      "learning_rate": 2e-05,
+      "loss": 0.0469871,
+      "step": 14586
+    },
+    {
+      "epoch": 29.174,
+      "grad_norm": 1.423553228378296,
+      "learning_rate": 2e-05,
+      "loss": 0.03586338,
+      "step": 14587
+    },
+    {
+      "epoch": 29.176,
+      "grad_norm": 1.2574028968811035,
+      "learning_rate": 2e-05,
+      "loss": 0.038145,
+      "step": 14588
+    },
+    {
+      "epoch": 29.178,
+      "grad_norm": 1.1540359258651733,
+      "learning_rate": 2e-05,
+      "loss": 0.04100233,
+      "step": 14589
+    },
+    {
+      "epoch": 29.18,
+      "grad_norm": 1.3354766368865967,
+      "learning_rate": 2e-05,
+      "loss": 0.05904789,
+      "step": 14590
+    },
+    {
+      "epoch": 29.182,
+      "grad_norm": 1.163336992263794,
+      "learning_rate": 2e-05,
+      "loss": 0.04909441,
+      "step": 14591
+    },
+    {
+      "epoch": 29.184,
+      "grad_norm": 2.428375482559204,
+      "learning_rate": 2e-05,
+      "loss": 0.04465162,
+      "step": 14592
+    },
+    {
+      "epoch": 29.186,
+      "grad_norm": 1.6806772947311401,
+      "learning_rate": 2e-05,
+      "loss": 0.04538292,
+      "step": 14593
+    },
+    {
+      "epoch": 29.188,
+      "grad_norm": 1.1672495603561401,
+      "learning_rate": 2e-05,
+      "loss": 0.04777298,
+      "step": 14594
+    },
+    {
+      "epoch": 29.19,
+      "grad_norm": 1.0995149612426758,
+      "learning_rate": 2e-05,
+      "loss": 0.03148938,
+      "step": 14595
+    },
+    {
+      "epoch": 29.192,
+      "grad_norm": 1.1681532859802246,
+      "learning_rate": 2e-05,
+      "loss": 0.03608632,
+      "step": 14596
+    },
+    {
+      "epoch": 29.194,
+      "grad_norm": 1.3820486068725586,
+      "learning_rate": 2e-05,
+      "loss": 0.05060082,
+      "step": 14597
+    },
+    {
+      "epoch": 29.196,
+      "grad_norm": 1.0916459560394287,
+      "learning_rate": 2e-05,
+      "loss": 0.03177829,
+      "step": 14598
+    },
+    {
+      "epoch": 29.198,
+      "grad_norm": 1.2117968797683716,
+      "learning_rate": 2e-05,
+      "loss": 0.04310794,
+      "step": 14599
+    },
+    {
+      "epoch": 29.2,
+      "grad_norm": 1.3815139532089233,
+      "learning_rate": 2e-05,
+      "loss": 0.03904974,
+      "step": 14600
+    },
+    {
+      "epoch": 29.202,
+      "grad_norm": 1.455543875694275,
+      "learning_rate": 2e-05,
+      "loss": 0.04168963,
+      "step": 14601
+    },
+    {
+      "epoch": 29.204,
+      "grad_norm": 1.2078819274902344,
+      "learning_rate": 2e-05,
+      "loss": 0.05466375,
+      "step": 14602
+    },
+    {
+      "epoch": 29.206,
+      "grad_norm": 1.8423599004745483,
+      "learning_rate": 2e-05,
+      "loss": 0.03679687,
+      "step": 14603
+    },
+    {
+      "epoch": 29.208,
+      "grad_norm": 1.3103402853012085,
+      "learning_rate": 2e-05,
+      "loss": 0.0345687,
+      "step": 14604
+    },
+    {
+      "epoch": 29.21,
+      "grad_norm": 1.3979569673538208,
+      "learning_rate": 2e-05,
+      "loss": 0.04111661,
+      "step": 14605
+    },
+    {
+      "epoch": 29.212,
+      "grad_norm": 1.2820682525634766,
+      "learning_rate": 2e-05,
+      "loss": 0.05244845,
+      "step": 14606
+    },
+    {
+      "epoch": 29.214,
+      "grad_norm": 1.8376924991607666,
+      "learning_rate": 2e-05,
+      "loss": 0.06727045,
+      "step": 14607
+    },
+    {
+      "epoch": 29.216,
+      "grad_norm": 1.4911775588989258,
+      "learning_rate": 2e-05,
+      "loss": 0.03376488,
+      "step": 14608
+    },
+    {
+      "epoch": 29.218,
+      "grad_norm": 2.5098586082458496,
+      "learning_rate": 2e-05,
+      "loss": 0.07967003,
+      "step": 14609
+    },
+    {
+      "epoch": 29.22,
+      "grad_norm": 1.4665964841842651,
+      "learning_rate": 2e-05,
+      "loss": 0.05623496,
+      "step": 14610
+    },
+    {
+      "epoch": 29.222,
+      "grad_norm": 1.1030839681625366,
+      "learning_rate": 2e-05,
+      "loss": 0.04080631,
+      "step": 14611
+    },
+    {
+      "epoch": 29.224,
+      "grad_norm": 1.169342041015625,
+      "learning_rate": 2e-05,
+      "loss": 0.0512477,
+      "step": 14612
+    },
+    {
+      "epoch": 29.226,
+      "grad_norm": 5.032285690307617,
+      "learning_rate": 2e-05,
+      "loss": 0.04048228,
+      "step": 14613
+    },
+    {
+      "epoch": 29.228,
+      "grad_norm": 1.2166979312896729,
+      "learning_rate": 2e-05,
+      "loss": 0.04632211,
+      "step": 14614
+    },
+    {
+      "epoch": 29.23,
+      "grad_norm": 1.1534336805343628,
+      "learning_rate": 2e-05,
+      "loss": 0.03913743,
+      "step": 14615
+    },
+    {
+      "epoch": 29.232,
+      "grad_norm": 1.1760746240615845,
+      "learning_rate": 2e-05,
+      "loss": 0.04341678,
+      "step": 14616
+    },
+    {
+      "epoch": 29.234,
+      "grad_norm": 1.4027602672576904,
+      "learning_rate": 2e-05,
+      "loss": 0.04686597,
+      "step": 14617
+    },
+    {
+      "epoch": 29.236,
+      "grad_norm": 3.7930917739868164,
+      "learning_rate": 2e-05,
+      "loss": 0.05493877,
+      "step": 14618
+    },
+    {
+      "epoch": 29.238,
+      "grad_norm": 1.149735689163208,
+      "learning_rate": 2e-05,
+      "loss": 0.0425221,
+      "step": 14619
+    },
+    {
+      "epoch": 29.24,
+      "grad_norm": 1.6718671321868896,
+      "learning_rate": 2e-05,
+      "loss": 0.05708656,
+      "step": 14620
+    },
+    {
+      "epoch": 29.242,
+      "grad_norm": 1.2668569087982178,
+      "learning_rate": 2e-05,
+      "loss": 0.04664732,
+      "step": 14621
+    },
+    {
+      "epoch": 29.244,
+      "grad_norm": 1.033226490020752,
+      "learning_rate": 2e-05,
+      "loss": 0.02460607,
+      "step": 14622
+    },
+    {
+      "epoch": 29.246,
+      "grad_norm": 1.172166109085083,
+      "learning_rate": 2e-05,
+      "loss": 0.04102369,
+      "step": 14623
+    },
+    {
+      "epoch": 29.248,
+      "grad_norm": 1.5639601945877075,
+      "learning_rate": 2e-05,
+      "loss": 0.0346932,
+      "step": 14624
+    },
+    {
+      "epoch": 29.25,
+      "grad_norm": 1.2138320207595825,
+      "learning_rate": 2e-05,
+      "loss": 0.05353147,
+      "step": 14625
+    },
+    {
+      "epoch": 29.252,
+      "grad_norm": 1.075577974319458,
+      "learning_rate": 2e-05,
+      "loss": 0.04629251,
+      "step": 14626
+    },
+    {
+      "epoch": 29.254,
+      "grad_norm": 4.6314473152160645,
+      "learning_rate": 2e-05,
+      "loss": 0.05750819,
+      "step": 14627
+    },
+    {
+      "epoch": 29.256,
+      "grad_norm": 1.6789414882659912,
+      "learning_rate": 2e-05,
+      "loss": 0.04619577,
+      "step": 14628
+    },
+    {
+      "epoch": 29.258,
+      "grad_norm": 10.09363079071045,
+      "learning_rate": 2e-05,
+      "loss": 0.06505147,
+      "step": 14629
+    },
+    {
+      "epoch": 29.26,
+      "grad_norm": 1.3053208589553833,
+      "learning_rate": 2e-05,
+      "loss": 0.03836732,
+      "step": 14630
+    },
+    {
+      "epoch": 29.262,
+      "grad_norm": 1.1734479665756226,
+      "learning_rate": 2e-05,
+      "loss": 0.04292649,
+      "step": 14631
+    },
+    {
+      "epoch": 29.264,
+      "grad_norm": 1.0191097259521484,
+      "learning_rate": 2e-05,
+      "loss": 0.0362784,
+      "step": 14632
+    },
+    {
+      "epoch": 29.266,
+      "grad_norm": 1.4284632205963135,
+      "learning_rate": 2e-05,
+      "loss": 0.04708409,
+      "step": 14633
+    },
+    {
+      "epoch": 29.268,
+      "grad_norm": 1.857406735420227,
+      "learning_rate": 2e-05,
+      "loss": 0.04922803,
+      "step": 14634
+    },
+    {
+      "epoch": 29.27,
+      "grad_norm": 1.4661109447479248,
+      "learning_rate": 2e-05,
+      "loss": 0.032971,
+      "step": 14635
+    },
+    {
+      "epoch": 29.272,
+      "grad_norm": 1.6468689441680908,
+      "learning_rate": 2e-05,
+      "loss": 0.04467516,
+      "step": 14636
+    },
+    {
+      "epoch": 29.274,
+      "grad_norm": 1.1976274251937866,
+      "learning_rate": 2e-05,
+      "loss": 0.05661123,
+      "step": 14637
+    },
+    {
+      "epoch": 29.276,
+      "grad_norm": 1.0101951360702515,
+      "learning_rate": 2e-05,
+      "loss": 0.04431897,
+      "step": 14638
+    },
+    {
+      "epoch": 29.278,
+      "grad_norm": 1.9550938606262207,
+      "learning_rate": 2e-05,
+      "loss": 0.03428712,
+      "step": 14639
+    },
+    {
+      "epoch": 29.28,
+      "grad_norm": 1.263190746307373,
+      "learning_rate": 2e-05,
+      "loss": 0.05213517,
+      "step": 14640
+    },
+    {
+      "epoch": 29.282,
+      "grad_norm": 1.0882905721664429,
+      "learning_rate": 2e-05,
+      "loss": 0.04379845,
+      "step": 14641
+    },
+    {
+      "epoch": 29.284,
+      "grad_norm": 1.6553674936294556,
+      "learning_rate": 2e-05,
+      "loss": 0.04459314,
+      "step": 14642
+    },
+    {
+      "epoch": 29.286,
+      "grad_norm": 1.1553642749786377,
+      "learning_rate": 2e-05,
+      "loss": 0.04111927,
+      "step": 14643
+    },
+    {
+      "epoch": 29.288,
+      "grad_norm": 1.7052907943725586,
+      "learning_rate": 2e-05,
+      "loss": 0.04706088,
+      "step": 14644
+    },
+    {
+      "epoch": 29.29,
+      "grad_norm": 1.7893346548080444,
+      "learning_rate": 2e-05,
+      "loss": 0.04122625,
+      "step": 14645
+    },
+    {
+      "epoch": 29.292,
+      "grad_norm": 0.8917121887207031,
+      "learning_rate": 2e-05,
+      "loss": 0.02282002,
+      "step": 14646
+    },
+    {
+      "epoch": 29.294,
+      "grad_norm": 1.3831899166107178,
+      "learning_rate": 2e-05,
+      "loss": 0.06464884,
+      "step": 14647
+    },
+    {
+      "epoch": 29.296,
+      "grad_norm": 1.0708141326904297,
+      "learning_rate": 2e-05,
+      "loss": 0.03723243,
+      "step": 14648
+    },
+    {
+      "epoch": 29.298,
+      "grad_norm": 2.3124451637268066,
+      "learning_rate": 2e-05,
+      "loss": 0.03636015,
+      "step": 14649
+    },
+    {
+      "epoch": 29.3,
+      "grad_norm": 1.150187373161316,
+      "learning_rate": 2e-05,
+      "loss": 0.03666411,
+      "step": 14650
+    },
+    {
+      "epoch": 29.302,
+      "grad_norm": 1.4192458391189575,
+      "learning_rate": 2e-05,
+      "loss": 0.05449123,
+      "step": 14651
+    },
+    {
+      "epoch": 29.304,
+      "grad_norm": 1.0729455947875977,
+      "learning_rate": 2e-05,
+      "loss": 0.04042206,
+      "step": 14652
+    },
+    {
+      "epoch": 29.306,
+      "grad_norm": 1.1908464431762695,
+      "learning_rate": 2e-05,
+      "loss": 0.0445194,
+      "step": 14653
+    },
+    {
+      "epoch": 29.308,
+      "grad_norm": 1.1397109031677246,
+      "learning_rate": 2e-05,
+      "loss": 0.03711201,
+      "step": 14654
+    },
+    {
+      "epoch": 29.31,
+      "grad_norm": 1.309540867805481,
+      "learning_rate": 2e-05,
+      "loss": 0.04084215,
+      "step": 14655
+    },
+    {
+      "epoch": 29.312,
+      "grad_norm": 1.0415390729904175,
+      "learning_rate": 2e-05,
+      "loss": 0.03293938,
+      "step": 14656
+    },
+    {
+      "epoch": 29.314,
+      "grad_norm": 1.683875560760498,
+      "learning_rate": 2e-05,
+      "loss": 0.04933864,
+      "step": 14657
+    },
+    {
+      "epoch": 29.316,
+      "grad_norm": 1.3526222705841064,
+      "learning_rate": 2e-05,
+      "loss": 0.02989796,
+      "step": 14658
+    },
+    {
+      "epoch": 29.318,
+      "grad_norm": 1.5054032802581787,
+      "learning_rate": 2e-05,
+      "loss": 0.04449115,
+      "step": 14659
+    },
+    {
+      "epoch": 29.32,
+      "grad_norm": 1.2268929481506348,
+      "learning_rate": 2e-05,
+      "loss": 0.0320545,
+      "step": 14660
+    },
+    {
+      "epoch": 29.322,
+      "grad_norm": 1.3516592979431152,
+      "learning_rate": 2e-05,
+      "loss": 0.05182027,
+      "step": 14661
+    },
+    {
+      "epoch": 29.324,
+      "grad_norm": 1.1587803363800049,
+      "learning_rate": 2e-05,
+      "loss": 0.04185852,
+      "step": 14662
+    },
+    {
+      "epoch": 29.326,
+      "grad_norm": 1.048889398574829,
+      "learning_rate": 2e-05,
+      "loss": 0.02531143,
+      "step": 14663
+    },
+    {
+      "epoch": 29.328,
+      "grad_norm": 2.0608630180358887,
+      "learning_rate": 2e-05,
+      "loss": 0.04467651,
+      "step": 14664
+    },
+    {
+      "epoch": 29.33,
+      "grad_norm": 1.0219707489013672,
+      "learning_rate": 2e-05,
+      "loss": 0.02736703,
+      "step": 14665
+    },
+    {
+      "epoch": 29.332,
+      "grad_norm": 1.091489553451538,
+      "learning_rate": 2e-05,
+      "loss": 0.04337668,
+      "step": 14666
+    },
+    {
+      "epoch": 29.334,
+      "grad_norm": 1.2775390148162842,
+      "learning_rate": 2e-05,
+      "loss": 0.04817453,
+      "step": 14667
+    },
+    {
+      "epoch": 29.336,
+      "grad_norm": 1.760574221611023,
+      "learning_rate": 2e-05,
+      "loss": 0.04524562,
+      "step": 14668
+    },
+    {
+      "epoch": 29.338,
+      "grad_norm": 1.5247159004211426,
+      "learning_rate": 2e-05,
+      "loss": 0.04375861,
+      "step": 14669
+    },
+    {
+      "epoch": 29.34,
+      "grad_norm": 1.308775544166565,
+      "learning_rate": 2e-05,
+      "loss": 0.04865634,
+      "step": 14670
+    },
+    {
+      "epoch": 29.342,
+      "grad_norm": 1.2305173873901367,
+      "learning_rate": 2e-05,
+      "loss": 0.04637342,
+      "step": 14671
+    },
+    {
+      "epoch": 29.344,
+      "grad_norm": 1.3184024095535278,
+      "learning_rate": 2e-05,
+      "loss": 0.048801,
+      "step": 14672
+    },
+    {
+      "epoch": 29.346,
+      "grad_norm": 1.0286909341812134,
+      "learning_rate": 2e-05,
+      "loss": 0.03713179,
+      "step": 14673
+    },
+    {
+      "epoch": 29.348,
+      "grad_norm": 1.2571288347244263,
+      "learning_rate": 2e-05,
+      "loss": 0.03584104,
+      "step": 14674
+    },
+    {
+      "epoch": 29.35,
+      "grad_norm": 0.9635476469993591,
+      "learning_rate": 2e-05,
+      "loss": 0.03240035,
+      "step": 14675
+    },
+    {
+      "epoch": 29.352,
+      "grad_norm": 0.9458169341087341,
+      "learning_rate": 2e-05,
+      "loss": 0.03441823,
+      "step": 14676
+    },
+    {
+      "epoch": 29.354,
+      "grad_norm": 1.2421998977661133,
+      "learning_rate": 2e-05,
+      "loss": 0.02801424,
+      "step": 14677
+    },
+    {
+      "epoch": 29.356,
+      "grad_norm": 1.1887518167495728,
+      "learning_rate": 2e-05,
+      "loss": 0.03601447,
+      "step": 14678
+    },
+    {
+      "epoch": 29.358,
+      "grad_norm": 1.2935701608657837,
+      "learning_rate": 2e-05,
+      "loss": 0.05169864,
+      "step": 14679
+    },
+    {
+      "epoch": 29.36,
+      "grad_norm": 1.5155844688415527,
+      "learning_rate": 2e-05,
+      "loss": 0.05170265,
+      "step": 14680
+    },
+    {
+      "epoch": 29.362,
+      "grad_norm": 1.0864388942718506,
+      "learning_rate": 2e-05,
+      "loss": 0.03917462,
+      "step": 14681
+    },
+    {
+      "epoch": 29.364,
+      "grad_norm": 3.3979506492614746,
+      "learning_rate": 2e-05,
+      "loss": 0.06590277,
+      "step": 14682
+    },
+    {
+      "epoch": 29.366,
+      "grad_norm": 1.263260006904602,
+      "learning_rate": 2e-05,
+      "loss": 0.03736039,
+      "step": 14683
+    },
+    {
+      "epoch": 29.368,
+      "grad_norm": 1.6203467845916748,
+      "learning_rate": 2e-05,
+      "loss": 0.03045352,
+      "step": 14684
+    },
+    {
+      "epoch": 29.37,
+      "grad_norm": 1.185299038887024,
+      "learning_rate": 2e-05,
+      "loss": 0.0308273,
+      "step": 14685
+    },
+    {
+      "epoch": 29.372,
+      "grad_norm": 1.065706491470337,
+      "learning_rate": 2e-05,
+      "loss": 0.04142375,
+      "step": 14686
+    },
+    {
+      "epoch": 29.374,
+      "grad_norm": 1.071706771850586,
+      "learning_rate": 2e-05,
+      "loss": 0.02857183,
+      "step": 14687
+    },
+    {
+      "epoch": 29.376,
+      "grad_norm": 1.0515819787979126,
+      "learning_rate": 2e-05,
+      "loss": 0.03293607,
+      "step": 14688
+    },
+    {
+      "epoch": 29.378,
+      "grad_norm": 1.2863487005233765,
+      "learning_rate": 2e-05,
+      "loss": 0.06486448,
+      "step": 14689
+    },
+    {
+      "epoch": 29.38,
+      "grad_norm": 1.3283601999282837,
+      "learning_rate": 2e-05,
+      "loss": 0.0511212,
+      "step": 14690
+    },
+    {
+      "epoch": 29.382,
+      "grad_norm": 1.3435486555099487,
+      "learning_rate": 2e-05,
+      "loss": 0.04754411,
+      "step": 14691
+    },
+    {
+      "epoch": 29.384,
+      "grad_norm": 2.4031832218170166,
+      "learning_rate": 2e-05,
+      "loss": 0.02932245,
+      "step": 14692
+    },
+    {
+      "epoch": 29.386,
+      "grad_norm": 1.0233304500579834,
+      "learning_rate": 2e-05,
+      "loss": 0.02972196,
+      "step": 14693
+    },
+    {
+      "epoch": 29.388,
+      "grad_norm": 0.9224061369895935,
+      "learning_rate": 2e-05,
+      "loss": 0.03122431,
+      "step": 14694
+    },
+    {
+      "epoch": 29.39,
+      "grad_norm": 3.187535524368286,
+      "learning_rate": 2e-05,
+      "loss": 0.06037506,
+      "step": 14695
+    },
+    {
+      "epoch": 29.392,
+      "grad_norm": 2.3195505142211914,
+      "learning_rate": 2e-05,
+      "loss": 0.04076795,
+      "step": 14696
+    },
+    {
+      "epoch": 29.394,
+      "grad_norm": 1.0622916221618652,
+      "learning_rate": 2e-05,
+      "loss": 0.02912345,
+      "step": 14697
+    },
+    {
+      "epoch": 29.396,
+      "grad_norm": 1.1375831365585327,
+      "learning_rate": 2e-05,
+      "loss": 0.03395087,
+      "step": 14698
+    },
+    {
+      "epoch": 29.398,
+      "grad_norm": 4.46896505355835,
+      "learning_rate": 2e-05,
+      "loss": 0.08969009,
+      "step": 14699
+    },
+    {
+      "epoch": 29.4,
+      "grad_norm": 1.382957100868225,
+      "learning_rate": 2e-05,
+      "loss": 0.05825138,
+      "step": 14700
+    },
+    {
+      "epoch": 29.402,
+      "grad_norm": 3.7034122943878174,
+      "learning_rate": 2e-05,
+      "loss": 0.05984267,
+      "step": 14701
+    },
+    {
+      "epoch": 29.404,
+      "grad_norm": 1.288468837738037,
+      "learning_rate": 2e-05,
+      "loss": 0.04220777,
+      "step": 14702
+    },
+    {
+      "epoch": 29.406,
+      "grad_norm": 5.697216510772705,
+      "learning_rate": 2e-05,
+      "loss": 0.04966062,
+      "step": 14703
+    },
+    {
+      "epoch": 29.408,
+      "grad_norm": 1.3597912788391113,
+      "learning_rate": 2e-05,
+      "loss": 0.06009267,
+      "step": 14704
+    },
+    {
+      "epoch": 29.41,
+      "grad_norm": 1.123845100402832,
+      "learning_rate": 2e-05,
+      "loss": 0.03696072,
+      "step": 14705
+    },
+    {
+      "epoch": 29.412,
+      "grad_norm": 1.5596554279327393,
+      "learning_rate": 2e-05,
+      "loss": 0.03732103,
+      "step": 14706
+    },
+    {
+      "epoch": 29.414,
+      "grad_norm": 1.2256417274475098,
+      "learning_rate": 2e-05,
+      "loss": 0.0455781,
+      "step": 14707
+    },
+    {
+      "epoch": 29.416,
+      "grad_norm": 1.2588534355163574,
+      "learning_rate": 2e-05,
+      "loss": 0.04248727,
+      "step": 14708
+    },
+    {
+      "epoch": 29.418,
+      "grad_norm": 1.5599313974380493,
+      "learning_rate": 2e-05,
+      "loss": 0.05124461,
+      "step": 14709
+    },
+    {
+      "epoch": 29.42,
+      "grad_norm": 1.3829172849655151,
+      "learning_rate": 2e-05,
+      "loss": 0.05284266,
+      "step": 14710
+    },
+    {
+      "epoch": 29.422,
+      "grad_norm": 0.9758954644203186,
+      "learning_rate": 2e-05,
+      "loss": 0.03507173,
+      "step": 14711
+    },
+    {
+      "epoch": 29.424,
+      "grad_norm": 1.019920825958252,
+      "learning_rate": 2e-05,
+      "loss": 0.03660026,
+      "step": 14712
+    },
+    {
+      "epoch": 29.426,
+      "grad_norm": 1.6652642488479614,
+      "learning_rate": 2e-05,
+      "loss": 0.05362531,
+      "step": 14713
+    },
+    {
+      "epoch": 29.428,
+      "grad_norm": 3.9348857402801514,
+      "learning_rate": 2e-05,
+      "loss": 0.03445033,
+      "step": 14714
+    },
+    {
+      "epoch": 29.43,
+      "grad_norm": 0.9290486574172974,
+      "learning_rate": 2e-05,
+      "loss": 0.03240933,
+      "step": 14715
+    },
+    {
+      "epoch": 29.432,
+      "grad_norm": 1.4232569932937622,
+      "learning_rate": 2e-05,
+      "loss": 0.05291854,
+      "step": 14716
+    },
+    {
+      "epoch": 29.434,
+      "grad_norm": 1.1019871234893799,
+      "learning_rate": 2e-05,
+      "loss": 0.04499703,
+      "step": 14717
+    },
+    {
+      "epoch": 29.436,
+      "grad_norm": 1.0482898950576782,
+      "learning_rate": 2e-05,
+      "loss": 0.02875671,
+      "step": 14718
+    },
+    {
+      "epoch": 29.438,
+      "grad_norm": 2.0963454246520996,
+      "learning_rate": 2e-05,
+      "loss": 0.06372029,
+      "step": 14719
+    },
+    {
+      "epoch": 29.44,
+      "grad_norm": 1.4427788257598877,
+      "learning_rate": 2e-05,
+      "loss": 0.05994175,
+      "step": 14720
+    },
+    {
+      "epoch": 29.442,
+      "grad_norm": 1.1246081590652466,
+      "learning_rate": 2e-05,
+      "loss": 0.05035602,
+      "step": 14721
+    },
+    {
+      "epoch": 29.444,
+      "grad_norm": 1.080545425415039,
+      "learning_rate": 2e-05,
+      "loss": 0.03505052,
+      "step": 14722
+    },
+    {
+      "epoch": 29.446,
+      "grad_norm": 2.5756614208221436,
+      "learning_rate": 2e-05,
+      "loss": 0.0641964,
+      "step": 14723
+    },
+    {
+      "epoch": 29.448,
+      "grad_norm": 0.9659985899925232,
+      "learning_rate": 2e-05,
+      "loss": 0.0326273,
+      "step": 14724
+    },
+    {
+      "epoch": 29.45,
+      "grad_norm": 1.0880937576293945,
+      "learning_rate": 2e-05,
+      "loss": 0.03114274,
+      "step": 14725
+    },
+    {
+      "epoch": 29.452,
+      "grad_norm": 1.5530476570129395,
+      "learning_rate": 2e-05,
+      "loss": 0.0458639,
+      "step": 14726
+    },
+    {
+      "epoch": 29.454,
+      "grad_norm": 1.9776489734649658,
+      "learning_rate": 2e-05,
+      "loss": 0.04284504,
+      "step": 14727
+    },
+    {
+      "epoch": 29.456,
+      "grad_norm": 2.1341958045959473,
+      "learning_rate": 2e-05,
+      "loss": 0.06103675,
+      "step": 14728
+    },
+    {
+      "epoch": 29.458,
+      "grad_norm": 2.14361834526062,
+      "learning_rate": 2e-05,
+      "loss": 0.04144895,
+      "step": 14729
+    },
+    {
+      "epoch": 29.46,
+      "grad_norm": 1.2067841291427612,
+      "learning_rate": 2e-05,
+      "loss": 0.03508757,
+      "step": 14730
+    },
+    {
+      "epoch": 29.462,
+      "grad_norm": 1.6224901676177979,
+      "learning_rate": 2e-05,
+      "loss": 0.03946466,
+      "step": 14731
+    },
+    {
+      "epoch": 29.464,
+      "grad_norm": 1.643559455871582,
+      "learning_rate": 2e-05,
+      "loss": 0.04336884,
+      "step": 14732
+    },
+    {
+      "epoch": 29.466,
+      "grad_norm": 1.0944159030914307,
+      "learning_rate": 2e-05,
+      "loss": 0.05785472,
+      "step": 14733
+    },
+    {
+      "epoch": 29.468,
+      "grad_norm": 1.0807979106903076,
+      "learning_rate": 2e-05,
+      "loss": 0.03531633,
+      "step": 14734
+    },
+    {
+      "epoch": 29.47,
+      "grad_norm": 2.1468169689178467,
+      "learning_rate": 2e-05,
+      "loss": 0.05086451,
+      "step": 14735
+    },
+    {
+      "epoch": 29.472,
+      "grad_norm": 1.5246763229370117,
+      "learning_rate": 2e-05,
+      "loss": 0.05114653,
+      "step": 14736
+    },
+    {
+      "epoch": 29.474,
+      "grad_norm": 1.0691903829574585,
+      "learning_rate": 2e-05,
+      "loss": 0.04326432,
+      "step": 14737
+    },
+    {
+      "epoch": 29.476,
+      "grad_norm": 1.1668561697006226,
+      "learning_rate": 2e-05,
+      "loss": 0.05376833,
+      "step": 14738
+    },
+    {
+      "epoch": 29.478,
+      "grad_norm": 1.191372275352478,
+      "learning_rate": 2e-05,
+      "loss": 0.04350607,
+      "step": 14739
+    },
+    {
+      "epoch": 29.48,
+      "grad_norm": 1.1428472995758057,
+      "learning_rate": 2e-05,
+      "loss": 0.03237382,
+      "step": 14740
+    },
+    {
+      "epoch": 29.482,
+      "grad_norm": 1.0473662614822388,
+      "learning_rate": 2e-05,
+      "loss": 0.03938694,
+      "step": 14741
+    },
+    {
+      "epoch": 29.484,
+      "grad_norm": 0.9359759092330933,
+      "learning_rate": 2e-05,
+      "loss": 0.03335633,
+      "step": 14742
+    },
+    {
+      "epoch": 29.486,
+      "grad_norm": 1.2112761735916138,
+      "learning_rate": 2e-05,
+      "loss": 0.03153354,
+      "step": 14743
+    },
+    {
+      "epoch": 29.488,
+      "grad_norm": 0.8638390302658081,
+      "learning_rate": 2e-05,
+      "loss": 0.0252349,
+      "step": 14744
+    },
+    {
+      "epoch": 29.49,
+      "grad_norm": 0.9504496455192566,
+      "learning_rate": 2e-05,
+      "loss": 0.03327549,
+      "step": 14745
+    },
+    {
+      "epoch": 29.492,
+      "grad_norm": 4.146841526031494,
+      "learning_rate": 2e-05,
+      "loss": 0.04628399,
+      "step": 14746
+    },
+    {
+      "epoch": 29.494,
+      "grad_norm": 2.1307597160339355,
+      "learning_rate": 2e-05,
+      "loss": 0.06677426,
+      "step": 14747
+    },
+    {
+      "epoch": 29.496,
+      "grad_norm": 1.4499460458755493,
+      "learning_rate": 2e-05,
+      "loss": 0.04317267,
+      "step": 14748
+    },
+    {
+      "epoch": 29.498,
+      "grad_norm": 1.7764891386032104,
+      "learning_rate": 2e-05,
+      "loss": 0.06008489,
+      "step": 14749
+    },
+    {
+      "epoch": 29.5,
+      "grad_norm": 1.4432235956192017,
+      "learning_rate": 2e-05,
+      "loss": 0.06923801,
+      "step": 14750
+    },
+    {
+      "epoch": 29.502,
+      "grad_norm": 1.4468677043914795,
+      "learning_rate": 2e-05,
+      "loss": 0.05008958,
+      "step": 14751
+    },
+    {
+      "epoch": 29.504,
+      "grad_norm": 1.3012151718139648,
+      "learning_rate": 2e-05,
+      "loss": 0.04738281,
+      "step": 14752
+    },
+    {
+      "epoch": 29.506,
+      "grad_norm": 1.1001543998718262,
+      "learning_rate": 2e-05,
+      "loss": 0.04851674,
+      "step": 14753
+    },
+    {
+      "epoch": 29.508,
+      "grad_norm": 1.2774382829666138,
+      "learning_rate": 2e-05,
+      "loss": 0.03910155,
+      "step": 14754
+    },
+    {
+      "epoch": 29.51,
+      "grad_norm": 1.2547918558120728,
+      "learning_rate": 2e-05,
+      "loss": 0.05241001,
+      "step": 14755
+    },
+    {
+      "epoch": 29.512,
+      "grad_norm": 1.3092374801635742,
+      "learning_rate": 2e-05,
+      "loss": 0.04025935,
+      "step": 14756
+    },
+    {
+      "epoch": 29.514,
+      "grad_norm": 1.3792351484298706,
+      "learning_rate": 2e-05,
+      "loss": 0.0517666,
+      "step": 14757
+    },
+    {
+      "epoch": 29.516,
+      "grad_norm": 0.9093297719955444,
+      "learning_rate": 2e-05,
+      "loss": 0.03267594,
+      "step": 14758
+    },
+    {
+      "epoch": 29.518,
+      "grad_norm": 1.0705640316009521,
+      "learning_rate": 2e-05,
+      "loss": 0.04189052,
+      "step": 14759
+    },
+    {
+      "epoch": 29.52,
+      "grad_norm": 1.0387095212936401,
+      "learning_rate": 2e-05,
+      "loss": 0.04280265,
+      "step": 14760
+    },
+    {
+      "epoch": 29.522,
+      "grad_norm": 1.2232365608215332,
+      "learning_rate": 2e-05,
+      "loss": 0.04506699,
+      "step": 14761
+    },
+    {
+      "epoch": 29.524,
+      "grad_norm": 1.1054472923278809,
+      "learning_rate": 2e-05,
+      "loss": 0.04134972,
+      "step": 14762
+    },
+    {
+      "epoch": 29.526,
+      "grad_norm": 1.1075634956359863,
+      "learning_rate": 2e-05,
+      "loss": 0.03896327,
+      "step": 14763
+    },
+    {
+      "epoch": 29.528,
+      "grad_norm": 1.2247326374053955,
+      "learning_rate": 2e-05,
+      "loss": 0.0459715,
+      "step": 14764
+    },
+    {
+      "epoch": 29.53,
+      "grad_norm": 1.2403117418289185,
+      "learning_rate": 2e-05,
+      "loss": 0.03991577,
+      "step": 14765
+    },
+    {
+      "epoch": 29.532,
+      "grad_norm": 1.0780881643295288,
+      "learning_rate": 2e-05,
+      "loss": 0.03780124,
+      "step": 14766
+    },
+    {
+      "epoch": 29.534,
+      "grad_norm": 1.5894378423690796,
+      "learning_rate": 2e-05,
+      "loss": 0.03415346,
+      "step": 14767
+    },
+    {
+      "epoch": 29.536,
+      "grad_norm": 0.9309705495834351,
+      "learning_rate": 2e-05,
+      "loss": 0.02937978,
+      "step": 14768
+    },
+    {
+      "epoch": 29.538,
+      "grad_norm": 1.135497808456421,
+      "learning_rate": 2e-05,
+      "loss": 0.03995682,
+      "step": 14769
+    },
+    {
+      "epoch": 29.54,
+      "grad_norm": 1.4162516593933105,
+      "learning_rate": 2e-05,
+      "loss": 0.05143316,
+      "step": 14770
+    },
+    {
+      "epoch": 29.542,
+      "grad_norm": 1.3331693410873413,
+      "learning_rate": 2e-05,
+      "loss": 0.04119586,
+      "step": 14771
+    },
+    {
+      "epoch": 29.544,
+      "grad_norm": 1.2744102478027344,
+      "learning_rate": 2e-05,
+      "loss": 0.06154334,
+      "step": 14772
+    },
+    {
+      "epoch": 29.546,
+      "grad_norm": 1.3395334482192993,
+      "learning_rate": 2e-05,
+      "loss": 0.03931785,
+      "step": 14773
+    },
+    {
+      "epoch": 29.548000000000002,
+      "grad_norm": 1.2095366716384888,
+      "learning_rate": 2e-05,
+      "loss": 0.04690979,
+      "step": 14774
+    },
+    {
+      "epoch": 29.55,
+      "grad_norm": 1.7333354949951172,
+      "learning_rate": 2e-05,
+      "loss": 0.0398626,
+      "step": 14775
+    },
+    {
+      "epoch": 29.552,
+      "grad_norm": 1.404498815536499,
+      "learning_rate": 2e-05,
+      "loss": 0.03682255,
+      "step": 14776
+    },
+    {
+      "epoch": 29.554,
+      "grad_norm": 1.3003278970718384,
+      "learning_rate": 2e-05,
+      "loss": 0.03933616,
+      "step": 14777
+    },
+    {
+      "epoch": 29.556,
+      "grad_norm": 1.013600468635559,
+      "learning_rate": 2e-05,
+      "loss": 0.04233592,
+      "step": 14778
+    },
+    {
+      "epoch": 29.558,
+      "grad_norm": 0.9388518333435059,
+      "learning_rate": 2e-05,
+      "loss": 0.03290233,
+      "step": 14779
+    },
+    {
+      "epoch": 29.56,
+      "grad_norm": 1.076974868774414,
+      "learning_rate": 2e-05,
+      "loss": 0.03113963,
+      "step": 14780
+    },
+    {
+      "epoch": 29.562,
+      "grad_norm": 1.5698806047439575,
+      "learning_rate": 2e-05,
+      "loss": 0.03970818,
+      "step": 14781
+    },
+    {
+      "epoch": 29.564,
+      "grad_norm": 1.5527507066726685,
+      "learning_rate": 2e-05,
+      "loss": 0.05457158,
+      "step": 14782
+    },
+    {
+      "epoch": 29.566,
+      "grad_norm": 1.105551838874817,
+      "learning_rate": 2e-05,
+      "loss": 0.0364199,
+      "step": 14783
+    },
+    {
+      "epoch": 29.568,
+      "grad_norm": 1.2447288036346436,
+      "learning_rate": 2e-05,
+      "loss": 0.05043525,
+      "step": 14784
+    },
+    {
+      "epoch": 29.57,
+      "grad_norm": 1.6889266967773438,
+      "learning_rate": 2e-05,
+      "loss": 0.03643173,
+      "step": 14785
+    },
+    {
+      "epoch": 29.572,
+      "grad_norm": 1.1132603883743286,
+      "learning_rate": 2e-05,
+      "loss": 0.03821333,
+      "step": 14786
+    },
+    {
+      "epoch": 29.574,
+      "grad_norm": 1.628547191619873,
+      "learning_rate": 2e-05,
+      "loss": 0.02059213,
+      "step": 14787
+    },
+    {
+      "epoch": 29.576,
+      "grad_norm": 1.4750348329544067,
+      "learning_rate": 2e-05,
+      "loss": 0.05588658,
+      "step": 14788
+    },
+    {
+      "epoch": 29.578,
+      "grad_norm": 1.142136812210083,
+      "learning_rate": 2e-05,
+      "loss": 0.04343769,
+      "step": 14789
+    },
+    {
+      "epoch": 29.58,
+      "grad_norm": 2.035980463027954,
+      "learning_rate": 2e-05,
+      "loss": 0.04574387,
+      "step": 14790
+    },
+    {
+      "epoch": 29.582,
+      "grad_norm": 0.9724331498146057,
+      "learning_rate": 2e-05,
+      "loss": 0.0316183,
+      "step": 14791
+    },
+    {
+      "epoch": 29.584,
+      "grad_norm": 1.0555124282836914,
+      "learning_rate": 2e-05,
+      "loss": 0.03854711,
+      "step": 14792
+    },
+    {
+      "epoch": 29.586,
+      "grad_norm": 2.9631097316741943,
+      "learning_rate": 2e-05,
+      "loss": 0.07389696,
+      "step": 14793
+    },
+    {
+      "epoch": 29.588,
+      "grad_norm": 1.2569674253463745,
+      "learning_rate": 2e-05,
+      "loss": 0.04960248,
+      "step": 14794
+    },
+    {
+      "epoch": 29.59,
+      "grad_norm": 0.8959588408470154,
+      "learning_rate": 2e-05,
+      "loss": 0.02651092,
+      "step": 14795
+    },
+    {
+      "epoch": 29.592,
+      "grad_norm": 1.4212428331375122,
+      "learning_rate": 2e-05,
+      "loss": 0.05283296,
+      "step": 14796
+    },
+    {
+      "epoch": 29.594,
+      "grad_norm": 1.2432961463928223,
+      "learning_rate": 2e-05,
+      "loss": 0.03414245,
+      "step": 14797
+    },
+    {
+      "epoch": 29.596,
+      "grad_norm": 1.3017022609710693,
+      "learning_rate": 2e-05,
+      "loss": 0.04572171,
+      "step": 14798
+    },
+    {
+      "epoch": 29.598,
+      "grad_norm": 1.7550230026245117,
+      "learning_rate": 2e-05,
+      "loss": 0.04602732,
+      "step": 14799
+    },
+    {
+      "epoch": 29.6,
+      "grad_norm": 0.9701183438301086,
+      "learning_rate": 2e-05,
+      "loss": 0.02774266,
+      "step": 14800
+    },
+    {
+      "epoch": 29.602,
+      "grad_norm": 1.1324889659881592,
+      "learning_rate": 2e-05,
+      "loss": 0.03797245,
+      "step": 14801
+    },
+    {
+      "epoch": 29.604,
+      "grad_norm": 1.1381210088729858,
+      "learning_rate": 2e-05,
+      "loss": 0.043383,
+      "step": 14802
+    },
+    {
+      "epoch": 29.606,
+      "grad_norm": 1.3297442197799683,
+      "learning_rate": 2e-05,
+      "loss": 0.04047207,
+      "step": 14803
+    },
+    {
+      "epoch": 29.608,
+      "grad_norm": 1.389540433883667,
+      "learning_rate": 2e-05,
+      "loss": 0.04826028,
+      "step": 14804
+    },
+    {
+      "epoch": 29.61,
+      "grad_norm": 2.0742807388305664,
+      "learning_rate": 2e-05,
+      "loss": 0.0447638,
+      "step": 14805
+    },
+    {
+      "epoch": 29.612,
+      "grad_norm": 3.014904260635376,
+      "learning_rate": 2e-05,
+      "loss": 0.04129062,
+      "step": 14806
+    },
+    {
+      "epoch": 29.614,
+      "grad_norm": 1.2650154829025269,
+      "learning_rate": 2e-05,
+      "loss": 0.03646675,
+      "step": 14807
+    },
+    {
+      "epoch": 29.616,
+      "grad_norm": 1.6480337381362915,
+      "learning_rate": 2e-05,
+      "loss": 0.03977471,
+      "step": 14808
+    },
+    {
+      "epoch": 29.618,
+      "grad_norm": 1.184668779373169,
+      "learning_rate": 2e-05,
+      "loss": 0.05217928,
+      "step": 14809
+    },
+    {
+      "epoch": 29.62,
+      "grad_norm": 1.2116751670837402,
+      "learning_rate": 2e-05,
+      "loss": 0.03607161,
+      "step": 14810
+    },
+    {
+      "epoch": 29.622,
+      "grad_norm": 1.508227825164795,
+      "learning_rate": 2e-05,
+      "loss": 0.04605871,
+      "step": 14811
+    },
+    {
+      "epoch": 29.624,
+      "grad_norm": 1.2774351835250854,
+      "learning_rate": 2e-05,
+      "loss": 0.04796334,
+      "step": 14812
+    },
+    {
+      "epoch": 29.626,
+      "grad_norm": 1.0448981523513794,
+      "learning_rate": 2e-05,
+      "loss": 0.03082025,
+      "step": 14813
+    },
+    {
+      "epoch": 29.628,
+      "grad_norm": 1.2237423658370972,
+      "learning_rate": 2e-05,
+      "loss": 0.04101267,
+      "step": 14814
+    },
+    {
+      "epoch": 29.63,
+      "grad_norm": 1.5655183792114258,
+      "learning_rate": 2e-05,
+      "loss": 0.04582419,
+      "step": 14815
+    },
+    {
+      "epoch": 29.632,
+      "grad_norm": 2.1846954822540283,
+      "learning_rate": 2e-05,
+      "loss": 0.05257653,
+      "step": 14816
+    },
+    {
+      "epoch": 29.634,
+      "grad_norm": 1.238049030303955,
+      "learning_rate": 2e-05,
+      "loss": 0.03405455,
+      "step": 14817
+    },
+    {
+      "epoch": 29.636,
+      "grad_norm": 1.6897296905517578,
+      "learning_rate": 2e-05,
+      "loss": 0.01916685,
+      "step": 14818
+    },
+    {
+      "epoch": 29.638,
+      "grad_norm": 0.9575730562210083,
+      "learning_rate": 2e-05,
+      "loss": 0.03765364,
+      "step": 14819
+    },
+    {
+      "epoch": 29.64,
+      "grad_norm": 2.1810710430145264,
+      "learning_rate": 2e-05,
+      "loss": 0.05722524,
+      "step": 14820
+    },
+    {
+      "epoch": 29.642,
+      "grad_norm": 0.9834027886390686,
+      "learning_rate": 2e-05,
+      "loss": 0.04266867,
+      "step": 14821
+    },
+    {
+      "epoch": 29.644,
+      "grad_norm": 1.2944148778915405,
+      "learning_rate": 2e-05,
+      "loss": 0.05544176,
+      "step": 14822
+    },
+    {
+      "epoch": 29.646,
+      "grad_norm": 1.9650918245315552,
+      "learning_rate": 2e-05,
+      "loss": 0.02375646,
+      "step": 14823
+    },
+    {
+      "epoch": 29.648,
+      "grad_norm": 0.9951708912849426,
+      "learning_rate": 2e-05,
+      "loss": 0.02884598,
+      "step": 14824
+    },
+    {
+      "epoch": 29.65,
+      "grad_norm": 1.0523395538330078,
+      "learning_rate": 2e-05,
+      "loss": 0.03761973,
+      "step": 14825
+    },
+    {
+      "epoch": 29.652,
+      "grad_norm": 1.0838866233825684,
+      "learning_rate": 2e-05,
+      "loss": 0.03728612,
+      "step": 14826
+    },
+    {
+      "epoch": 29.654,
+      "grad_norm": 1.0605547428131104,
+      "learning_rate": 2e-05,
+      "loss": 0.03837313,
+      "step": 14827
+    },
+    {
+      "epoch": 29.656,
+      "grad_norm": 1.1938754320144653,
+      "learning_rate": 2e-05,
+      "loss": 0.05729951,
+      "step": 14828
+    },
+    {
+      "epoch": 29.658,
+      "grad_norm": 2.144594430923462,
+      "learning_rate": 2e-05,
+      "loss": 0.04445962,
+      "step": 14829
+    },
+    {
+      "epoch": 29.66,
+      "grad_norm": 1.8156507015228271,
+      "learning_rate": 2e-05,
+      "loss": 0.03740517,
+      "step": 14830
+    },
+    {
+      "epoch": 29.662,
+      "grad_norm": 1.244408130645752,
+      "learning_rate": 2e-05,
+      "loss": 0.04781131,
+      "step": 14831
+    },
+    {
+      "epoch": 29.664,
+      "grad_norm": 1.2653872966766357,
+      "learning_rate": 2e-05,
+      "loss": 0.0585079,
+      "step": 14832
+    },
+    {
+      "epoch": 29.666,
+      "grad_norm": 1.3794729709625244,
+      "learning_rate": 2e-05,
+      "loss": 0.03325066,
+      "step": 14833
+    },
+    {
+      "epoch": 29.668,
+      "grad_norm": 1.5158426761627197,
+      "learning_rate": 2e-05,
+      "loss": 0.04454071,
+      "step": 14834
+    },
+    {
+      "epoch": 29.67,
+      "grad_norm": 0.9856611490249634,
+      "learning_rate": 2e-05,
+      "loss": 0.03513188,
+      "step": 14835
+    },
+    {
+      "epoch": 29.672,
+      "grad_norm": 1.0112903118133545,
+      "learning_rate": 2e-05,
+      "loss": 0.03400346,
+      "step": 14836
+    },
+    {
+      "epoch": 29.674,
+      "grad_norm": 2.110278606414795,
+      "learning_rate": 2e-05,
+      "loss": 0.06449761,
+      "step": 14837
+    },
+    {
+      "epoch": 29.676,
+      "grad_norm": 1.4624208211898804,
+      "learning_rate": 2e-05,
+      "loss": 0.05108786,
+      "step": 14838
+    },
+    {
+      "epoch": 29.678,
+      "grad_norm": 1.6022717952728271,
+      "learning_rate": 2e-05,
+      "loss": 0.04408568,
+      "step": 14839
+    },
+    {
+      "epoch": 29.68,
+      "grad_norm": 1.1528983116149902,
+      "learning_rate": 2e-05,
+      "loss": 0.04578086,
+      "step": 14840
+    },
+    {
+      "epoch": 29.682,
+      "grad_norm": 1.0715017318725586,
+      "learning_rate": 2e-05,
+      "loss": 0.04369386,
+      "step": 14841
+    },
+    {
+      "epoch": 29.684,
+      "grad_norm": 1.3998444080352783,
+      "learning_rate": 2e-05,
+      "loss": 0.04720485,
+      "step": 14842
+    },
+    {
+      "epoch": 29.686,
+      "grad_norm": 1.2437803745269775,
+      "learning_rate": 2e-05,
+      "loss": 0.04152617,
+      "step": 14843
+    },
+    {
+      "epoch": 29.688,
+      "grad_norm": 1.2567319869995117,
+      "learning_rate": 2e-05,
+      "loss": 0.03946881,
+      "step": 14844
+    },
+    {
+      "epoch": 29.69,
+      "grad_norm": 1.7165799140930176,
+      "learning_rate": 2e-05,
+      "loss": 0.03655791,
+      "step": 14845
+    },
+    {
+      "epoch": 29.692,
+      "grad_norm": 0.9617201685905457,
+      "learning_rate": 2e-05,
+      "loss": 0.03272104,
+      "step": 14846
+    },
+    {
+      "epoch": 29.694,
+      "grad_norm": 1.2149977684020996,
+      "learning_rate": 2e-05,
+      "loss": 0.04295772,
+      "step": 14847
+    },
+    {
+      "epoch": 29.696,
+      "grad_norm": 1.5055172443389893,
+      "learning_rate": 2e-05,
+      "loss": 0.04826812,
+      "step": 14848
+    },
+    {
+      "epoch": 29.698,
+      "grad_norm": 1.1299159526824951,
+      "learning_rate": 2e-05,
+      "loss": 0.04282779,
+      "step": 14849
+    },
+    {
+      "epoch": 29.7,
+      "grad_norm": 1.5118989944458008,
+      "learning_rate": 2e-05,
+      "loss": 0.04391037,
+      "step": 14850
+    },
+    {
+      "epoch": 29.701999999999998,
+      "grad_norm": 2.338747501373291,
+      "learning_rate": 2e-05,
+      "loss": 0.05273699,
+      "step": 14851
+    },
+    {
+      "epoch": 29.704,
+      "grad_norm": 1.2019315958023071,
+      "learning_rate": 2e-05,
+      "loss": 0.04617017,
+      "step": 14852
+    },
+    {
+      "epoch": 29.706,
+      "grad_norm": 1.618383765220642,
+      "learning_rate": 2e-05,
+      "loss": 0.06011579,
+      "step": 14853
+    },
+    {
+      "epoch": 29.708,
+      "grad_norm": 1.8187501430511475,
+      "learning_rate": 2e-05,
+      "loss": 0.03678617,
+      "step": 14854
+    },
+    {
+      "epoch": 29.71,
+      "grad_norm": 1.143147587776184,
+      "learning_rate": 2e-05,
+      "loss": 0.04157409,
+      "step": 14855
+    },
+    {
+      "epoch": 29.712,
+      "grad_norm": 1.034891128540039,
+      "learning_rate": 2e-05,
+      "loss": 0.04029275,
+      "step": 14856
+    },
+    {
+      "epoch": 29.714,
+      "grad_norm": 1.3997042179107666,
+      "learning_rate": 2e-05,
+      "loss": 0.0383551,
+      "step": 14857
+    },
+    {
+      "epoch": 29.716,
+      "grad_norm": 1.0622479915618896,
+      "learning_rate": 2e-05,
+      "loss": 0.04367114,
+      "step": 14858
+    },
+    {
+      "epoch": 29.718,
+      "grad_norm": 1.824373722076416,
+      "learning_rate": 2e-05,
+      "loss": 0.04785391,
+      "step": 14859
+    },
+    {
+      "epoch": 29.72,
+      "grad_norm": 0.980838418006897,
+      "learning_rate": 2e-05,
+      "loss": 0.03948919,
+      "step": 14860
+    },
+    {
+      "epoch": 29.722,
+      "grad_norm": 1.3569544553756714,
+      "learning_rate": 2e-05,
+      "loss": 0.05038366,
+      "step": 14861
+    },
+    {
+      "epoch": 29.724,
+      "grad_norm": 1.304292917251587,
+      "learning_rate": 2e-05,
+      "loss": 0.03696183,
+      "step": 14862
+    },
+    {
+      "epoch": 29.726,
+      "grad_norm": 3.2446513175964355,
+      "learning_rate": 2e-05,
+      "loss": 0.05859087,
+      "step": 14863
+    },
+    {
+      "epoch": 29.728,
+      "grad_norm": 1.621229648590088,
+      "learning_rate": 2e-05,
+      "loss": 0.02783653,
+      "step": 14864
+    },
+    {
+      "epoch": 29.73,
+      "grad_norm": 1.2278355360031128,
+      "learning_rate": 2e-05,
+      "loss": 0.05444655,
+      "step": 14865
+    },
+    {
+      "epoch": 29.732,
+      "grad_norm": 1.2174080610275269,
+      "learning_rate": 2e-05,
+      "loss": 0.05196868,
+      "step": 14866
+    },
+    {
+      "epoch": 29.734,
+      "grad_norm": 1.6062108278274536,
+      "learning_rate": 2e-05,
+      "loss": 0.04332883,
+      "step": 14867
+    },
+    {
+      "epoch": 29.736,
+      "grad_norm": 1.161248803138733,
+      "learning_rate": 2e-05,
+      "loss": 0.03697128,
+      "step": 14868
+    },
+    {
+      "epoch": 29.738,
+      "grad_norm": 1.2859642505645752,
+      "learning_rate": 2e-05,
+      "loss": 0.05505038,
+      "step": 14869
+    },
+    {
+      "epoch": 29.74,
+      "grad_norm": 1.4202659130096436,
+      "learning_rate": 2e-05,
+      "loss": 0.0477881,
+      "step": 14870
+    },
+    {
+      "epoch": 29.742,
+      "grad_norm": 1.2499854564666748,
+      "learning_rate": 2e-05,
+      "loss": 0.05374869,
+      "step": 14871
+    },
+    {
+      "epoch": 29.744,
+      "grad_norm": 0.955565333366394,
+      "learning_rate": 2e-05,
+      "loss": 0.02939542,
+      "step": 14872
+    },
+    {
+      "epoch": 29.746,
+      "grad_norm": 1.3554538488388062,
+      "learning_rate": 2e-05,
+      "loss": 0.04614054,
+      "step": 14873
+    },
+    {
+      "epoch": 29.748,
+      "grad_norm": 1.089884638786316,
+      "learning_rate": 2e-05,
+      "loss": 0.03633293,
+      "step": 14874
+    },
+    {
+      "epoch": 29.75,
+      "grad_norm": 1.1395641565322876,
+      "learning_rate": 2e-05,
+      "loss": 0.04034775,
+      "step": 14875
+    },
+    {
+      "epoch": 29.752,
+      "grad_norm": 1.3981385231018066,
+      "learning_rate": 2e-05,
+      "loss": 0.04628996,
+      "step": 14876
+    },
+    {
+      "epoch": 29.754,
+      "grad_norm": 1.7084894180297852,
+      "learning_rate": 2e-05,
+      "loss": 0.04334879,
+      "step": 14877
+    },
+    {
+      "epoch": 29.756,
+      "grad_norm": 2.442875385284424,
+      "learning_rate": 2e-05,
+      "loss": 0.05451498,
+      "step": 14878
+    },
+    {
+      "epoch": 29.758,
+      "grad_norm": 1.4088022708892822,
+      "learning_rate": 2e-05,
+      "loss": 0.05375902,
+      "step": 14879
+    },
+    {
+      "epoch": 29.76,
+      "grad_norm": 1.7990258932113647,
+      "learning_rate": 2e-05,
+      "loss": 0.04616355,
+      "step": 14880
+    },
+    {
+      "epoch": 29.762,
+      "grad_norm": 2.0218417644500732,
+      "learning_rate": 2e-05,
+      "loss": 0.03352433,
+      "step": 14881
+    },
+    {
+      "epoch": 29.764,
+      "grad_norm": 1.8284379243850708,
+      "learning_rate": 2e-05,
+      "loss": 0.04442716,
+      "step": 14882
+    },
+    {
+      "epoch": 29.766,
+      "grad_norm": 1.8842039108276367,
+      "learning_rate": 2e-05,
+      "loss": 0.03985734,
+      "step": 14883
+    },
+    {
+      "epoch": 29.768,
+      "grad_norm": 1.5910454988479614,
+      "learning_rate": 2e-05,
+      "loss": 0.0417336,
+      "step": 14884
+    },
+    {
+      "epoch": 29.77,
+      "grad_norm": 5.6194987297058105,
+      "learning_rate": 2e-05,
+      "loss": 0.06214419,
+      "step": 14885
+    },
+    {
+      "epoch": 29.772,
+      "grad_norm": 1.221018671989441,
+      "learning_rate": 2e-05,
+      "loss": 0.05359631,
+      "step": 14886
+    },
+    {
+      "epoch": 29.774,
+      "grad_norm": 1.4850738048553467,
+      "learning_rate": 2e-05,
+      "loss": 0.04186449,
+      "step": 14887
+    },
+    {
+      "epoch": 29.776,
+      "grad_norm": 2.0336577892303467,
+      "learning_rate": 2e-05,
+      "loss": 0.03588335,
+      "step": 14888
+    },
+    {
+      "epoch": 29.778,
+      "grad_norm": 1.0733425617218018,
+      "learning_rate": 2e-05,
+      "loss": 0.04026532,
+      "step": 14889
+    },
+    {
+      "epoch": 29.78,
+      "grad_norm": 1.098156452178955,
+      "learning_rate": 2e-05,
+      "loss": 0.04264045,
+      "step": 14890
+    },
+    {
+      "epoch": 29.782,
+      "grad_norm": 1.4435909986495972,
+      "learning_rate": 2e-05,
+      "loss": 0.03248915,
+      "step": 14891
+    },
+    {
+      "epoch": 29.784,
+      "grad_norm": 3.9169139862060547,
+      "learning_rate": 2e-05,
+      "loss": 0.05105972,
+      "step": 14892
+    },
+    {
+      "epoch": 29.786,
+      "grad_norm": 1.969159722328186,
+      "learning_rate": 2e-05,
+      "loss": 0.05551322,
+      "step": 14893
+    },
+    {
+      "epoch": 29.788,
+      "grad_norm": 1.2635542154312134,
+      "learning_rate": 2e-05,
+      "loss": 0.04648653,
+      "step": 14894
+    },
+    {
+      "epoch": 29.79,
+      "grad_norm": 2.3477652072906494,
+      "learning_rate": 2e-05,
+      "loss": 0.06277344,
+      "step": 14895
+    },
+    {
+      "epoch": 29.792,
+      "grad_norm": 1.0149587392807007,
+      "learning_rate": 2e-05,
+      "loss": 0.03186602,
+      "step": 14896
+    },
+    {
+      "epoch": 29.794,
+      "grad_norm": 1.1373287439346313,
+      "learning_rate": 2e-05,
+      "loss": 0.03521444,
+      "step": 14897
+    },
+    {
+      "epoch": 29.796,
+      "grad_norm": 1.9701231718063354,
+      "learning_rate": 2e-05,
+      "loss": 0.05275909,
+      "step": 14898
+    },
+    {
+      "epoch": 29.798000000000002,
+      "grad_norm": 1.1814894676208496,
+      "learning_rate": 2e-05,
+      "loss": 0.0368986,
+      "step": 14899
+    },
+    {
+      "epoch": 29.8,
+      "grad_norm": 1.1683019399642944,
+      "learning_rate": 2e-05,
+      "loss": 0.04178146,
+      "step": 14900
+    },
+    {
+      "epoch": 29.802,
+      "grad_norm": 1.2779021263122559,
+      "learning_rate": 2e-05,
+      "loss": 0.03463981,
+      "step": 14901
+    },
+    {
+      "epoch": 29.804,
+      "grad_norm": 1.0893248319625854,
+      "learning_rate": 2e-05,
+      "loss": 0.03910547,
+      "step": 14902
+    },
+    {
+      "epoch": 29.806,
+      "grad_norm": 1.4731847047805786,
+      "learning_rate": 2e-05,
+      "loss": 0.04429988,
+      "step": 14903
+    },
+    {
+      "epoch": 29.808,
+      "grad_norm": 1.3464443683624268,
+      "learning_rate": 2e-05,
+      "loss": 0.04595333,
+      "step": 14904
+    },
+    {
+      "epoch": 29.81,
+      "grad_norm": 1.2633438110351562,
+      "learning_rate": 2e-05,
+      "loss": 0.03501693,
+      "step": 14905
+    },
+    {
+      "epoch": 29.812,
+      "grad_norm": 1.1309336423873901,
+      "learning_rate": 2e-05,
+      "loss": 0.04051533,
+      "step": 14906
+    },
+    {
+      "epoch": 29.814,
+      "grad_norm": 1.007876992225647,
+      "learning_rate": 2e-05,
+      "loss": 0.03409252,
+      "step": 14907
+    },
+    {
+      "epoch": 29.816,
+      "grad_norm": 1.6379777193069458,
+      "learning_rate": 2e-05,
+      "loss": 0.05974958,
+      "step": 14908
+    },
+    {
+      "epoch": 29.818,
+      "grad_norm": 1.3505051136016846,
+      "learning_rate": 2e-05,
+      "loss": 0.05870251,
+      "step": 14909
+    },
+    {
+      "epoch": 29.82,
+      "grad_norm": 1.0954749584197998,
+      "learning_rate": 2e-05,
+      "loss": 0.03586072,
+      "step": 14910
+    },
+    {
+      "epoch": 29.822,
+      "grad_norm": 0.8500576615333557,
+      "learning_rate": 2e-05,
+      "loss": 0.02894812,
+      "step": 14911
+    },
+    {
+      "epoch": 29.824,
+      "grad_norm": 0.9059650897979736,
+      "learning_rate": 2e-05,
+      "loss": 0.03347109,
+      "step": 14912
+    },
+    {
+      "epoch": 29.826,
+      "grad_norm": 2.9803061485290527,
+      "learning_rate": 2e-05,
+      "loss": 0.05787177,
+      "step": 14913
+    },
+    {
+      "epoch": 29.828,
+      "grad_norm": 1.2712470293045044,
+      "learning_rate": 2e-05,
+      "loss": 0.03648849,
+      "step": 14914
+    },
+    {
+      "epoch": 29.83,
+      "grad_norm": 1.3802344799041748,
+      "learning_rate": 2e-05,
+      "loss": 0.04678344,
+      "step": 14915
+    },
+    {
+      "epoch": 29.832,
+      "grad_norm": 0.9832471609115601,
+      "learning_rate": 2e-05,
+      "loss": 0.03517216,
+      "step": 14916
+    },
+    {
+      "epoch": 29.834,
+      "grad_norm": 1.1615490913391113,
+      "learning_rate": 2e-05,
+      "loss": 0.04364904,
+      "step": 14917
+    },
+    {
+      "epoch": 29.836,
+      "grad_norm": 1.3471040725708008,
+      "learning_rate": 2e-05,
+      "loss": 0.0301399,
+      "step": 14918
+    },
+    {
+      "epoch": 29.838,
+      "grad_norm": 1.2438772916793823,
+      "learning_rate": 2e-05,
+      "loss": 0.0445204,
+      "step": 14919
+    },
+    {
+      "epoch": 29.84,
+      "grad_norm": 1.1942741870880127,
+      "learning_rate": 2e-05,
+      "loss": 0.03932862,
+      "step": 14920
+    },
+    {
+      "epoch": 29.842,
+      "grad_norm": 1.0131380558013916,
+      "learning_rate": 2e-05,
+      "loss": 0.04467336,
+      "step": 14921
+    },
+    {
+      "epoch": 29.844,
+      "grad_norm": 1.1072560548782349,
+      "learning_rate": 2e-05,
+      "loss": 0.03384002,
+      "step": 14922
+    },
+    {
+      "epoch": 29.846,
+      "grad_norm": 1.2888267040252686,
+      "learning_rate": 2e-05,
+      "loss": 0.03786197,
+      "step": 14923
+    },
+    {
+      "epoch": 29.848,
+      "grad_norm": 1.1689302921295166,
+      "learning_rate": 2e-05,
+      "loss": 0.05054848,
+      "step": 14924
+    },
+    {
+      "epoch": 29.85,
+      "grad_norm": 1.3626478910446167,
+      "learning_rate": 2e-05,
+      "loss": 0.04308386,
+      "step": 14925
+    },
+    {
+      "epoch": 29.852,
+      "grad_norm": 1.0472064018249512,
+      "learning_rate": 2e-05,
+      "loss": 0.033345,
+      "step": 14926
+    },
+    {
+      "epoch": 29.854,
+      "grad_norm": 1.0115253925323486,
+      "learning_rate": 2e-05,
+      "loss": 0.03919667,
+      "step": 14927
+    },
+    {
+      "epoch": 29.856,
+      "grad_norm": 1.8129942417144775,
+      "learning_rate": 2e-05,
+      "loss": 0.04338409,
+      "step": 14928
+    },
+    {
+      "epoch": 29.858,
+      "grad_norm": 1.012441873550415,
+      "learning_rate": 2e-05,
+      "loss": 0.03030182,
+      "step": 14929
+    },
+    {
+      "epoch": 29.86,
+      "grad_norm": 1.0548601150512695,
+      "learning_rate": 2e-05,
+      "loss": 0.03725137,
+      "step": 14930
+    },
+    {
+      "epoch": 29.862,
+      "grad_norm": 1.027254581451416,
+      "learning_rate": 2e-05,
+      "loss": 0.04611764,
+      "step": 14931
+    },
+    {
+      "epoch": 29.864,
+      "grad_norm": 1.1189601421356201,
+      "learning_rate": 2e-05,
+      "loss": 0.03380036,
+      "step": 14932
+    },
+    {
+      "epoch": 29.866,
+      "grad_norm": 1.33846116065979,
+      "learning_rate": 2e-05,
+      "loss": 0.06069037,
+      "step": 14933
+    },
+    {
+      "epoch": 29.868,
+      "grad_norm": 1.096492052078247,
+      "learning_rate": 2e-05,
+      "loss": 0.04014472,
+      "step": 14934
+    },
+    {
+      "epoch": 29.87,
+      "grad_norm": 1.174947738647461,
+      "learning_rate": 2e-05,
+      "loss": 0.02727459,
+      "step": 14935
+    },
+    {
+      "epoch": 29.872,
+      "grad_norm": 1.087896466255188,
+      "learning_rate": 2e-05,
+      "loss": 0.04823601,
+      "step": 14936
+    },
+    {
+      "epoch": 29.874,
+      "grad_norm": 1.0803396701812744,
+      "learning_rate": 2e-05,
+      "loss": 0.04184086,
+      "step": 14937
+    },
+    {
+      "epoch": 29.876,
+      "grad_norm": 1.0180484056472778,
+      "learning_rate": 2e-05,
+      "loss": 0.02937475,
+      "step": 14938
+    },
+    {
+      "epoch": 29.878,
+      "grad_norm": 1.099775791168213,
+      "learning_rate": 2e-05,
+      "loss": 0.04277577,
+      "step": 14939
+    },
+    {
+      "epoch": 29.88,
+      "grad_norm": 1.3840006589889526,
+      "learning_rate": 2e-05,
+      "loss": 0.04482201,
+      "step": 14940
+    },
+    {
+      "epoch": 29.882,
+      "grad_norm": 1.119278907775879,
+      "learning_rate": 2e-05,
+      "loss": 0.04266882,
+      "step": 14941
+    },
+    {
+      "epoch": 29.884,
+      "grad_norm": 1.1960852146148682,
+      "learning_rate": 2e-05,
+      "loss": 0.03658342,
+      "step": 14942
+    },
+    {
+      "epoch": 29.886,
+      "grad_norm": 1.3054516315460205,
+      "learning_rate": 2e-05,
+      "loss": 0.04391373,
+      "step": 14943
+    },
+    {
+      "epoch": 29.888,
+      "grad_norm": 2.7474465370178223,
+      "learning_rate": 2e-05,
+      "loss": 0.04351743,
+      "step": 14944
+    },
+    {
+      "epoch": 29.89,
+      "grad_norm": 1.2123734951019287,
+      "learning_rate": 2e-05,
+      "loss": 0.03769352,
+      "step": 14945
+    },
+    {
+      "epoch": 29.892,
+      "grad_norm": 1.325861930847168,
+      "learning_rate": 2e-05,
+      "loss": 0.05362713,
+      "step": 14946
+    },
+    {
+      "epoch": 29.894,
+      "grad_norm": 1.4618078470230103,
+      "learning_rate": 2e-05,
+      "loss": 0.04493023,
+      "step": 14947
+    },
+    {
+      "epoch": 29.896,
+      "grad_norm": 1.041840672492981,
+      "learning_rate": 2e-05,
+      "loss": 0.03452935,
+      "step": 14948
+    },
+    {
+      "epoch": 29.898,
+      "grad_norm": 1.356163740158081,
+      "learning_rate": 2e-05,
+      "loss": 0.05633589,
+      "step": 14949
+    },
+    {
+      "epoch": 29.9,
+      "grad_norm": 1.409123420715332,
+      "learning_rate": 2e-05,
+      "loss": 0.04535218,
+      "step": 14950
+    },
+    {
+      "epoch": 29.902,
+      "grad_norm": 1.5181453227996826,
+      "learning_rate": 2e-05,
+      "loss": 0.03477838,
+      "step": 14951
+    },
+    {
+      "epoch": 29.904,
+      "grad_norm": 1.1335440874099731,
+      "learning_rate": 2e-05,
+      "loss": 0.05675086,
+      "step": 14952
+    },
+    {
+      "epoch": 29.906,
+      "grad_norm": 2.706404209136963,
+      "learning_rate": 2e-05,
+      "loss": 0.04841974,
+      "step": 14953
+    },
+    {
+      "epoch": 29.908,
+      "grad_norm": 1.2822026014328003,
+      "learning_rate": 2e-05,
+      "loss": 0.0460251,
+      "step": 14954
+    },
+    {
+      "epoch": 29.91,
+      "grad_norm": 1.2433853149414062,
+      "learning_rate": 2e-05,
+      "loss": 0.04511236,
+      "step": 14955
+    },
+    {
+      "epoch": 29.912,
+      "grad_norm": 1.2542903423309326,
+      "learning_rate": 2e-05,
+      "loss": 0.04461327,
+      "step": 14956
+    },
+    {
+      "epoch": 29.914,
+      "grad_norm": 1.5348360538482666,
+      "learning_rate": 2e-05,
+      "loss": 0.06257273,
+      "step": 14957
+    },
+    {
+      "epoch": 29.916,
+      "grad_norm": 1.2867058515548706,
+      "learning_rate": 2e-05,
+      "loss": 0.03613833,
+      "step": 14958
+    },
+    {
+      "epoch": 29.918,
+      "grad_norm": 1.8032394647598267,
+      "learning_rate": 2e-05,
+      "loss": 0.06370901,
+      "step": 14959
+    },
+    {
+      "epoch": 29.92,
+      "grad_norm": 2.311445951461792,
+      "learning_rate": 2e-05,
+      "loss": 0.03839797,
+      "step": 14960
+    },
+    {
+      "epoch": 29.922,
+      "grad_norm": 1.3040590286254883,
+      "learning_rate": 2e-05,
+      "loss": 0.02956476,
+      "step": 14961
+    },
+    {
+      "epoch": 29.924,
+      "grad_norm": 1.2339740991592407,
+      "learning_rate": 2e-05,
+      "loss": 0.03441235,
+      "step": 14962
+    },
+    {
+      "epoch": 29.926,
+      "grad_norm": 1.8125553131103516,
+      "learning_rate": 2e-05,
+      "loss": 0.04188786,
+      "step": 14963
+    },
+    {
+      "epoch": 29.928,
+      "grad_norm": 1.2054352760314941,
+      "learning_rate": 2e-05,
+      "loss": 0.02116534,
+      "step": 14964
+    },
+    {
+      "epoch": 29.93,
+      "grad_norm": 1.1890215873718262,
+      "learning_rate": 2e-05,
+      "loss": 0.04022015,
+      "step": 14965
+    },
+    {
+      "epoch": 29.932,
+      "grad_norm": 1.709370493888855,
+      "learning_rate": 2e-05,
+      "loss": 0.04972868,
+      "step": 14966
+    },
+    {
+      "epoch": 29.934,
+      "grad_norm": 1.2925797700881958,
+      "learning_rate": 2e-05,
+      "loss": 0.0480578,
+      "step": 14967
+    },
+    {
+      "epoch": 29.936,
+      "grad_norm": 1.549068808555603,
+      "learning_rate": 2e-05,
+      "loss": 0.05139837,
+      "step": 14968
+    },
+    {
+      "epoch": 29.938,
+      "grad_norm": 1.3146717548370361,
+      "learning_rate": 2e-05,
+      "loss": 0.03840823,
+      "step": 14969
+    },
+    {
+      "epoch": 29.94,
+      "grad_norm": 1.0093461275100708,
+      "learning_rate": 2e-05,
+      "loss": 0.03652712,
+      "step": 14970
+    },
+    {
+      "epoch": 29.942,
+      "grad_norm": 3.679654121398926,
+      "learning_rate": 2e-05,
+      "loss": 0.06581252,
+      "step": 14971
+    },
+    {
+      "epoch": 29.944,
+      "grad_norm": 1.36691415309906,
+      "learning_rate": 2e-05,
+      "loss": 0.04719871,
+      "step": 14972
+    },
+    {
+      "epoch": 29.946,
+      "grad_norm": 1.2385770082473755,
+      "learning_rate": 2e-05,
+      "loss": 0.04421876,
+      "step": 14973
+    },
+    {
+      "epoch": 29.948,
+      "grad_norm": 1.2146538496017456,
+      "learning_rate": 2e-05,
+      "loss": 0.03504267,
+      "step": 14974
+    },
+    {
+      "epoch": 29.95,
+      "grad_norm": 1.7719730138778687,
+      "learning_rate": 2e-05,
+      "loss": 0.04804779,
+      "step": 14975
+    },
+    {
+      "epoch": 29.951999999999998,
+      "grad_norm": 1.596618890762329,
+      "learning_rate": 2e-05,
+      "loss": 0.06619457,
+      "step": 14976
+    },
+    {
+      "epoch": 29.954,
+      "grad_norm": 0.9984351396560669,
+      "learning_rate": 2e-05,
+      "loss": 0.02848834,
+      "step": 14977
+    },
+    {
+      "epoch": 29.956,
+      "grad_norm": 1.1433227062225342,
+      "learning_rate": 2e-05,
+      "loss": 0.0418468,
+      "step": 14978
+    },
+    {
+      "epoch": 29.958,
+      "grad_norm": 1.499070644378662,
+      "learning_rate": 2e-05,
+      "loss": 0.05216137,
+      "step": 14979
+    },
+    {
+      "epoch": 29.96,
+      "grad_norm": 1.0211125612258911,
+      "learning_rate": 2e-05,
+      "loss": 0.03615819,
+      "step": 14980
+    },
+    {
+      "epoch": 29.962,
+      "grad_norm": 3.567038059234619,
+      "learning_rate": 2e-05,
+      "loss": 0.05577373,
+      "step": 14981
+    },
+    {
+      "epoch": 29.964,
+      "grad_norm": 1.994671106338501,
+      "learning_rate": 2e-05,
+      "loss": 0.04374481,
+      "step": 14982
+    },
+    {
+      "epoch": 29.966,
+      "grad_norm": 1.1184746026992798,
+      "learning_rate": 2e-05,
+      "loss": 0.04952614,
+      "step": 14983
+    },
+    {
+      "epoch": 29.968,
+      "grad_norm": 1.5312135219573975,
+      "learning_rate": 2e-05,
+      "loss": 0.04638387,
+      "step": 14984
+    },
+    {
+      "epoch": 29.97,
+      "grad_norm": 1.1531622409820557,
+      "learning_rate": 2e-05,
+      "loss": 0.04479223,
+      "step": 14985
+    },
+    {
+      "epoch": 29.972,
+      "grad_norm": 1.111600637435913,
+      "learning_rate": 2e-05,
+      "loss": 0.03504475,
+      "step": 14986
+    },
+    {
+      "epoch": 29.974,
+      "grad_norm": 1.4171292781829834,
+      "learning_rate": 2e-05,
+      "loss": 0.03749974,
+      "step": 14987
+    },
+    {
+      "epoch": 29.976,
+      "grad_norm": 0.8947274684906006,
+      "learning_rate": 2e-05,
+      "loss": 0.02938852,
+      "step": 14988
+    },
+    {
+      "epoch": 29.978,
+      "grad_norm": 1.0345733165740967,
+      "learning_rate": 2e-05,
+      "loss": 0.04883969,
+      "step": 14989
+    },
+    {
+      "epoch": 29.98,
+      "grad_norm": 1.0411769151687622,
+      "learning_rate": 2e-05,
+      "loss": 0.04379828,
+      "step": 14990
+    },
+    {
+      "epoch": 29.982,
+      "grad_norm": 1.4568036794662476,
+      "learning_rate": 2e-05,
+      "loss": 0.04191837,
+      "step": 14991
+    },
+    {
+      "epoch": 29.984,
+      "grad_norm": 1.2673814296722412,
+      "learning_rate": 2e-05,
+      "loss": 0.04016875,
+      "step": 14992
+    },
+    {
+      "epoch": 29.986,
+      "grad_norm": 0.9755784869194031,
+      "learning_rate": 2e-05,
+      "loss": 0.03788521,
+      "step": 14993
+    },
+    {
+      "epoch": 29.988,
+      "grad_norm": 1.2146108150482178,
+      "learning_rate": 2e-05,
+      "loss": 0.0330191,
+      "step": 14994
+    },
+    {
+      "epoch": 29.99,
+      "grad_norm": 1.6079703569412231,
+      "learning_rate": 2e-05,
+      "loss": 0.05264592,
+      "step": 14995
+    },
+    {
+      "epoch": 29.992,
+      "grad_norm": 1.3531326055526733,
+      "learning_rate": 2e-05,
+      "loss": 0.05724774,
+      "step": 14996
+    },
+    {
+      "epoch": 29.994,
+      "grad_norm": 1.4556341171264648,
+      "learning_rate": 2e-05,
+      "loss": 0.06128872,
+      "step": 14997
+    },
+    {
+      "epoch": 29.996,
+      "grad_norm": 1.7786065340042114,
+      "learning_rate": 2e-05,
+      "loss": 0.06668452,
+      "step": 14998
+    },
+    {
+      "epoch": 29.998,
+      "grad_norm": 1.122294306755066,
+      "learning_rate": 2e-05,
+      "loss": 0.04705849,
+      "step": 14999
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 1.9331656694412231,
+      "learning_rate": 2e-05,
+      "loss": 0.0316266,
+      "step": 15000
+    },
+    {
+      "epoch": 30.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.998,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9820359281437125,
+        "Equal_1": 0.996,
+        "Equal_2": 0.9820359281437125,
+        "Equal_3": 0.9780439121756487,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9960079840319361,
+        "Parallel_1": 0.9919839679358717,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 1.0,
+        "Perpendicular_2": 0.988,
+        "Perpendicular_3": 0.8937875751503006,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 0.998,
+        "PointLiesOnCircle_3": 0.9916666666666667,
+        "PointLiesOnLine_1": 0.9939879759519038,
+        "PointLiesOnLine_2": 0.9939879759519038,
+        "PointLiesOnLine_3": 0.9900199600798403
+      },
+      "eval_runtime": 320.3735,
+      "eval_samples_per_second": 32.774,
+      "eval_steps_per_second": 0.655,
+      "step": 15000
+    },
+    {
+      "epoch": 30.002,
+      "grad_norm": 1.4232646226882935,
+      "learning_rate": 2e-05,
+      "loss": 0.04287136,
+      "step": 15001
+    },
+    {
+      "epoch": 30.004,
+      "grad_norm": 1.111973524093628,
+      "learning_rate": 2e-05,
+      "loss": 0.04549501,
+      "step": 15002
+    },
+    {
+      "epoch": 30.006,
+      "grad_norm": 1.4063668251037598,
+      "learning_rate": 2e-05,
+      "loss": 0.04815795,
+      "step": 15003
+    },
+    {
+      "epoch": 30.008,
+      "grad_norm": 1.0374184846878052,
+      "learning_rate": 2e-05,
+      "loss": 0.03240955,
+      "step": 15004
+    },
+    {
+      "epoch": 30.01,
+      "grad_norm": 2.511753797531128,
+      "learning_rate": 2e-05,
+      "loss": 0.05327015,
+      "step": 15005
+    },
+    {
+      "epoch": 30.012,
+      "grad_norm": 1.5259581804275513,
+      "learning_rate": 2e-05,
+      "loss": 0.04226237,
+      "step": 15006
+    },
+    {
+      "epoch": 30.014,
+      "grad_norm": 2.2420709133148193,
+      "learning_rate": 2e-05,
+      "loss": 0.05237993,
+      "step": 15007
+    },
+    {
+      "epoch": 30.016,
+      "grad_norm": 1.380776047706604,
+      "learning_rate": 2e-05,
+      "loss": 0.04655606,
+      "step": 15008
+    },
+    {
+      "epoch": 30.018,
+      "grad_norm": 1.1870553493499756,
+      "learning_rate": 2e-05,
+      "loss": 0.03608441,
+      "step": 15009
+    },
+    {
+      "epoch": 30.02,
+      "grad_norm": 1.869041085243225,
+      "learning_rate": 2e-05,
+      "loss": 0.05452606,
+      "step": 15010
+    },
+    {
+      "epoch": 30.022,
+      "grad_norm": 1.5327285528182983,
+      "learning_rate": 2e-05,
+      "loss": 0.07322414,
+      "step": 15011
+    },
+    {
+      "epoch": 30.024,
+      "grad_norm": 1.1738957166671753,
+      "learning_rate": 2e-05,
+      "loss": 0.04997585,
+      "step": 15012
+    },
+    {
+      "epoch": 30.026,
+      "grad_norm": 1.7185429334640503,
+      "learning_rate": 2e-05,
+      "loss": 0.0466965,
+      "step": 15013
+    },
+    {
+      "epoch": 30.028,
+      "grad_norm": 1.3917158842086792,
+      "learning_rate": 2e-05,
+      "loss": 0.04702974,
+      "step": 15014
+    },
+    {
+      "epoch": 30.03,
+      "grad_norm": 1.652574896812439,
+      "learning_rate": 2e-05,
+      "loss": 0.03737196,
+      "step": 15015
+    },
+    {
+      "epoch": 30.032,
+      "grad_norm": 1.2738018035888672,
+      "learning_rate": 2e-05,
+      "loss": 0.04564106,
+      "step": 15016
+    },
+    {
+      "epoch": 30.034,
+      "grad_norm": 1.3657252788543701,
+      "learning_rate": 2e-05,
+      "loss": 0.0442913,
+      "step": 15017
+    },
+    {
+      "epoch": 30.036,
+      "grad_norm": 1.1747430562973022,
+      "learning_rate": 2e-05,
+      "loss": 0.02756719,
+      "step": 15018
+    },
+    {
+      "epoch": 30.038,
+      "grad_norm": 1.0821881294250488,
+      "learning_rate": 2e-05,
+      "loss": 0.0346719,
+      "step": 15019
+    },
+    {
+      "epoch": 30.04,
+      "grad_norm": 1.230070948600769,
+      "learning_rate": 2e-05,
+      "loss": 0.03856944,
+      "step": 15020
+    },
+    {
+      "epoch": 30.042,
+      "grad_norm": 1.540521502494812,
+      "learning_rate": 2e-05,
+      "loss": 0.04057005,
+      "step": 15021
+    },
+    {
+      "epoch": 30.044,
+      "grad_norm": 2.8572275638580322,
+      "learning_rate": 2e-05,
+      "loss": 0.04592321,
+      "step": 15022
+    },
+    {
+      "epoch": 30.046,
+      "grad_norm": 0.957338809967041,
+      "learning_rate": 2e-05,
+      "loss": 0.03693869,
+      "step": 15023
+    },
+    {
+      "epoch": 30.048,
+      "grad_norm": 1.0699844360351562,
+      "learning_rate": 2e-05,
+      "loss": 0.03753718,
+      "step": 15024
+    },
+    {
+      "epoch": 30.05,
+      "grad_norm": 2.457496404647827,
+      "learning_rate": 2e-05,
+      "loss": 0.04840863,
+      "step": 15025
+    },
+    {
+      "epoch": 30.052,
+      "grad_norm": 1.2310270071029663,
+      "learning_rate": 2e-05,
+      "loss": 0.04352465,
+      "step": 15026
+    },
+    {
+      "epoch": 30.054,
+      "grad_norm": 1.2693308591842651,
+      "learning_rate": 2e-05,
+      "loss": 0.04948821,
+      "step": 15027
+    },
+    {
+      "epoch": 30.056,
+      "grad_norm": 1.4048819541931152,
+      "learning_rate": 2e-05,
+      "loss": 0.0399585,
+      "step": 15028
+    },
+    {
+      "epoch": 30.058,
+      "grad_norm": 2.0493316650390625,
+      "learning_rate": 2e-05,
+      "loss": 0.05211222,
+      "step": 15029
+    },
+    {
+      "epoch": 30.06,
+      "grad_norm": 3.7002038955688477,
+      "learning_rate": 2e-05,
+      "loss": 0.05875326,
+      "step": 15030
+    },
+    {
+      "epoch": 30.062,
+      "grad_norm": 1.7289574146270752,
+      "learning_rate": 2e-05,
+      "loss": 0.04627757,
+      "step": 15031
+    },
+    {
+      "epoch": 30.064,
+      "grad_norm": 1.0221132040023804,
+      "learning_rate": 2e-05,
+      "loss": 0.03416388,
+      "step": 15032
+    },
+    {
+      "epoch": 30.066,
+      "grad_norm": 1.4639520645141602,
+      "learning_rate": 2e-05,
+      "loss": 0.04067467,
+      "step": 15033
+    },
+    {
+      "epoch": 30.068,
+      "grad_norm": 3.5001463890075684,
+      "learning_rate": 2e-05,
+      "loss": 0.04698951,
+      "step": 15034
+    },
+    {
+      "epoch": 30.07,
+      "grad_norm": 2.3693861961364746,
+      "learning_rate": 2e-05,
+      "loss": 0.04970219,
+      "step": 15035
+    },
+    {
+      "epoch": 30.072,
+      "grad_norm": 1.3238275051116943,
+      "learning_rate": 2e-05,
+      "loss": 0.05959756,
+      "step": 15036
+    },
+    {
+      "epoch": 30.074,
+      "grad_norm": 1.5840702056884766,
+      "learning_rate": 2e-05,
+      "loss": 0.04837484,
+      "step": 15037
+    },
+    {
+      "epoch": 30.076,
+      "grad_norm": 0.9318821430206299,
+      "learning_rate": 2e-05,
+      "loss": 0.03147685,
+      "step": 15038
+    },
+    {
+      "epoch": 30.078,
+      "grad_norm": 1.0989964008331299,
+      "learning_rate": 2e-05,
+      "loss": 0.03799355,
+      "step": 15039
+    },
+    {
+      "epoch": 30.08,
+      "grad_norm": 1.4380488395690918,
+      "learning_rate": 2e-05,
+      "loss": 0.04152,
+      "step": 15040
+    },
+    {
+      "epoch": 30.082,
+      "grad_norm": 1.2408324480056763,
+      "learning_rate": 2e-05,
+      "loss": 0.04909746,
+      "step": 15041
+    },
+    {
+      "epoch": 30.084,
+      "grad_norm": 1.1136013269424438,
+      "learning_rate": 2e-05,
+      "loss": 0.03348158,
+      "step": 15042
+    },
+    {
+      "epoch": 30.086,
+      "grad_norm": 1.2548508644104004,
+      "learning_rate": 2e-05,
+      "loss": 0.0451351,
+      "step": 15043
+    },
+    {
+      "epoch": 30.088,
+      "grad_norm": 1.0726171731948853,
+      "learning_rate": 2e-05,
+      "loss": 0.04324069,
+      "step": 15044
+    },
+    {
+      "epoch": 30.09,
+      "grad_norm": 1.0724010467529297,
+      "learning_rate": 2e-05,
+      "loss": 0.0484529,
+      "step": 15045
+    },
+    {
+      "epoch": 30.092,
+      "grad_norm": 1.228132963180542,
+      "learning_rate": 2e-05,
+      "loss": 0.03065848,
+      "step": 15046
+    },
+    {
+      "epoch": 30.094,
+      "grad_norm": 1.099954605102539,
+      "learning_rate": 2e-05,
+      "loss": 0.04499891,
+      "step": 15047
+    },
+    {
+      "epoch": 30.096,
+      "grad_norm": 1.0415964126586914,
+      "learning_rate": 2e-05,
+      "loss": 0.03373061,
+      "step": 15048
+    },
+    {
+      "epoch": 30.098,
+      "grad_norm": 1.237966775894165,
+      "learning_rate": 2e-05,
+      "loss": 0.04612883,
+      "step": 15049
+    },
+    {
+      "epoch": 30.1,
+      "grad_norm": 0.8815692067146301,
+      "learning_rate": 2e-05,
+      "loss": 0.02756138,
+      "step": 15050
+    },
+    {
+      "epoch": 30.102,
+      "grad_norm": 1.3700236082077026,
+      "learning_rate": 2e-05,
+      "loss": 0.06041703,
+      "step": 15051
+    },
+    {
+      "epoch": 30.104,
+      "grad_norm": 1.2398154735565186,
+      "learning_rate": 2e-05,
+      "loss": 0.03576667,
+      "step": 15052
+    },
+    {
+      "epoch": 30.106,
+      "grad_norm": 1.1791821718215942,
+      "learning_rate": 2e-05,
+      "loss": 0.05038423,
+      "step": 15053
+    },
+    {
+      "epoch": 30.108,
+      "grad_norm": 2.0784451961517334,
+      "learning_rate": 2e-05,
+      "loss": 0.05495892,
+      "step": 15054
+    },
+    {
+      "epoch": 30.11,
+      "grad_norm": 2.1504173278808594,
+      "learning_rate": 2e-05,
+      "loss": 0.03890841,
+      "step": 15055
+    },
+    {
+      "epoch": 30.112,
+      "grad_norm": 3.112828254699707,
+      "learning_rate": 2e-05,
+      "loss": 0.05653836,
+      "step": 15056
+    },
+    {
+      "epoch": 30.114,
+      "grad_norm": 1.1712428331375122,
+      "learning_rate": 2e-05,
+      "loss": 0.05188262,
+      "step": 15057
+    },
+    {
+      "epoch": 30.116,
+      "grad_norm": 1.5940415859222412,
+      "learning_rate": 2e-05,
+      "loss": 0.05263701,
+      "step": 15058
+    },
+    {
+      "epoch": 30.118,
+      "grad_norm": 1.2304260730743408,
+      "learning_rate": 2e-05,
+      "loss": 0.04616508,
+      "step": 15059
+    },
+    {
+      "epoch": 30.12,
+      "grad_norm": 3.7707254886627197,
+      "learning_rate": 2e-05,
+      "loss": 0.05711006,
+      "step": 15060
+    },
+    {
+      "epoch": 30.122,
+      "grad_norm": 1.8504637479782104,
+      "learning_rate": 2e-05,
+      "loss": 0.04005916,
+      "step": 15061
+    },
+    {
+      "epoch": 30.124,
+      "grad_norm": 1.3818079233169556,
+      "learning_rate": 2e-05,
+      "loss": 0.0316132,
+      "step": 15062
+    },
+    {
+      "epoch": 30.126,
+      "grad_norm": 1.582114577293396,
+      "learning_rate": 2e-05,
+      "loss": 0.05220442,
+      "step": 15063
+    },
+    {
+      "epoch": 30.128,
+      "grad_norm": 1.2166645526885986,
+      "learning_rate": 2e-05,
+      "loss": 0.03328218,
+      "step": 15064
+    },
+    {
+      "epoch": 30.13,
+      "grad_norm": 1.2545546293258667,
+      "learning_rate": 2e-05,
+      "loss": 0.03880144,
+      "step": 15065
+    },
+    {
+      "epoch": 30.132,
+      "grad_norm": 1.1685189008712769,
+      "learning_rate": 2e-05,
+      "loss": 0.0422597,
+      "step": 15066
+    },
+    {
+      "epoch": 30.134,
+      "grad_norm": 1.2280715703964233,
+      "learning_rate": 2e-05,
+      "loss": 0.03680715,
+      "step": 15067
+    },
+    {
+      "epoch": 30.136,
+      "grad_norm": 1.2604179382324219,
+      "learning_rate": 2e-05,
+      "loss": 0.0577843,
+      "step": 15068
+    },
+    {
+      "epoch": 30.138,
+      "grad_norm": 1.333001971244812,
+      "learning_rate": 2e-05,
+      "loss": 0.03825287,
+      "step": 15069
+    },
+    {
+      "epoch": 30.14,
+      "grad_norm": 1.515370488166809,
+      "learning_rate": 2e-05,
+      "loss": 0.03778506,
+      "step": 15070
+    },
+    {
+      "epoch": 30.142,
+      "grad_norm": 1.265008568763733,
+      "learning_rate": 2e-05,
+      "loss": 0.03106946,
+      "step": 15071
+    },
+    {
+      "epoch": 30.144,
+      "grad_norm": 2.2077817916870117,
+      "learning_rate": 2e-05,
+      "loss": 0.03389125,
+      "step": 15072
+    },
+    {
+      "epoch": 30.146,
+      "grad_norm": 1.4839131832122803,
+      "learning_rate": 2e-05,
+      "loss": 0.0673124,
+      "step": 15073
+    },
+    {
+      "epoch": 30.148,
+      "grad_norm": 1.1836014986038208,
+      "learning_rate": 2e-05,
+      "loss": 0.0335351,
+      "step": 15074
+    },
+    {
+      "epoch": 30.15,
+      "grad_norm": 1.3027890920639038,
+      "learning_rate": 2e-05,
+      "loss": 0.04611604,
+      "step": 15075
+    },
+    {
+      "epoch": 30.152,
+      "grad_norm": 1.2303056716918945,
+      "learning_rate": 2e-05,
+      "loss": 0.04055518,
+      "step": 15076
+    },
+    {
+      "epoch": 30.154,
+      "grad_norm": 1.6350833177566528,
+      "learning_rate": 2e-05,
+      "loss": 0.0367165,
+      "step": 15077
+    },
+    {
+      "epoch": 30.156,
+      "grad_norm": 0.9786274433135986,
+      "learning_rate": 2e-05,
+      "loss": 0.03549477,
+      "step": 15078
+    },
+    {
+      "epoch": 30.158,
+      "grad_norm": 1.5880675315856934,
+      "learning_rate": 2e-05,
+      "loss": 0.04486579,
+      "step": 15079
+    },
+    {
+      "epoch": 30.16,
+      "grad_norm": 1.2543953657150269,
+      "learning_rate": 2e-05,
+      "loss": 0.04821232,
+      "step": 15080
+    },
+    {
+      "epoch": 30.162,
+      "grad_norm": 2.140390396118164,
+      "learning_rate": 2e-05,
+      "loss": 0.04136558,
+      "step": 15081
+    },
+    {
+      "epoch": 30.164,
+      "grad_norm": 1.0265830755233765,
+      "learning_rate": 2e-05,
+      "loss": 0.03865357,
+      "step": 15082
+    },
+    {
+      "epoch": 30.166,
+      "grad_norm": 1.3790358304977417,
+      "learning_rate": 2e-05,
+      "loss": 0.06286014,
+      "step": 15083
+    },
+    {
+      "epoch": 30.168,
+      "grad_norm": 2.187826633453369,
+      "learning_rate": 2e-05,
+      "loss": 0.06071652,
+      "step": 15084
+    },
+    {
+      "epoch": 30.17,
+      "grad_norm": 1.165565848350525,
+      "learning_rate": 2e-05,
+      "loss": 0.04597174,
+      "step": 15085
+    },
+    {
+      "epoch": 30.172,
+      "grad_norm": 1.5375369787216187,
+      "learning_rate": 2e-05,
+      "loss": 0.03429234,
+      "step": 15086
+    },
+    {
+      "epoch": 30.174,
+      "grad_norm": 2.7333414554595947,
+      "learning_rate": 2e-05,
+      "loss": 0.05888695,
+      "step": 15087
+    },
+    {
+      "epoch": 30.176,
+      "grad_norm": 1.0312190055847168,
+      "learning_rate": 2e-05,
+      "loss": 0.0352921,
+      "step": 15088
+    },
+    {
+      "epoch": 30.178,
+      "grad_norm": 4.89245080947876,
+      "learning_rate": 2e-05,
+      "loss": 0.0566744,
+      "step": 15089
+    },
+    {
+      "epoch": 30.18,
+      "grad_norm": 1.7868283987045288,
+      "learning_rate": 2e-05,
+      "loss": 0.05283628,
+      "step": 15090
+    },
+    {
+      "epoch": 30.182,
+      "grad_norm": 2.932136297225952,
+      "learning_rate": 2e-05,
+      "loss": 0.04512861,
+      "step": 15091
+    },
+    {
+      "epoch": 30.184,
+      "grad_norm": 1.8231194019317627,
+      "learning_rate": 2e-05,
+      "loss": 0.04028752,
+      "step": 15092
+    },
+    {
+      "epoch": 30.186,
+      "grad_norm": 1.5239124298095703,
+      "learning_rate": 2e-05,
+      "loss": 0.03271829,
+      "step": 15093
+    },
+    {
+      "epoch": 30.188,
+      "grad_norm": 1.2162010669708252,
+      "learning_rate": 2e-05,
+      "loss": 0.03735817,
+      "step": 15094
+    },
+    {
+      "epoch": 30.19,
+      "grad_norm": 1.240540623664856,
+      "learning_rate": 2e-05,
+      "loss": 0.04491368,
+      "step": 15095
+    },
+    {
+      "epoch": 30.192,
+      "grad_norm": 1.1648021936416626,
+      "learning_rate": 2e-05,
+      "loss": 0.04308267,
+      "step": 15096
+    },
+    {
+      "epoch": 30.194,
+      "grad_norm": 1.0710548162460327,
+      "learning_rate": 2e-05,
+      "loss": 0.04492667,
+      "step": 15097
+    },
+    {
+      "epoch": 30.196,
+      "grad_norm": 1.1542491912841797,
+      "learning_rate": 2e-05,
+      "loss": 0.0378947,
+      "step": 15098
+    },
+    {
+      "epoch": 30.198,
+      "grad_norm": 1.8039984703063965,
+      "learning_rate": 2e-05,
+      "loss": 0.05451577,
+      "step": 15099
+    },
+    {
+      "epoch": 30.2,
+      "grad_norm": 1.4430378675460815,
+      "learning_rate": 2e-05,
+      "loss": 0.05195063,
+      "step": 15100
+    },
+    {
+      "epoch": 30.202,
+      "grad_norm": 1.3164018392562866,
+      "learning_rate": 2e-05,
+      "loss": 0.03912523,
+      "step": 15101
+    },
+    {
+      "epoch": 30.204,
+      "grad_norm": 1.669808030128479,
+      "learning_rate": 2e-05,
+      "loss": 0.04372003,
+      "step": 15102
+    },
+    {
+      "epoch": 30.206,
+      "grad_norm": 1.5180182456970215,
+      "learning_rate": 2e-05,
+      "loss": 0.03866254,
+      "step": 15103
+    },
+    {
+      "epoch": 30.208,
+      "grad_norm": 1.6664462089538574,
+      "learning_rate": 2e-05,
+      "loss": 0.0531385,
+      "step": 15104
+    },
+    {
+      "epoch": 30.21,
+      "grad_norm": 1.4806171655654907,
+      "learning_rate": 2e-05,
+      "loss": 0.05282668,
+      "step": 15105
+    },
+    {
+      "epoch": 30.212,
+      "grad_norm": 1.128859281539917,
+      "learning_rate": 2e-05,
+      "loss": 0.03547608,
+      "step": 15106
+    },
+    {
+      "epoch": 30.214,
+      "grad_norm": 1.117397427558899,
+      "learning_rate": 2e-05,
+      "loss": 0.03773435,
+      "step": 15107
+    },
+    {
+      "epoch": 30.216,
+      "grad_norm": 1.2309256792068481,
+      "learning_rate": 2e-05,
+      "loss": 0.04536531,
+      "step": 15108
+    },
+    {
+      "epoch": 30.218,
+      "grad_norm": 1.3051683902740479,
+      "learning_rate": 2e-05,
+      "loss": 0.04438914,
+      "step": 15109
+    },
+    {
+      "epoch": 30.22,
+      "grad_norm": 1.4824970960617065,
+      "learning_rate": 2e-05,
+      "loss": 0.05524845,
+      "step": 15110
+    },
+    {
+      "epoch": 30.222,
+      "grad_norm": 1.3169000148773193,
+      "learning_rate": 2e-05,
+      "loss": 0.04102128,
+      "step": 15111
+    },
+    {
+      "epoch": 30.224,
+      "grad_norm": 1.2333604097366333,
+      "learning_rate": 2e-05,
+      "loss": 0.06657355,
+      "step": 15112
+    },
+    {
+      "epoch": 30.226,
+      "grad_norm": 0.9591508507728577,
+      "learning_rate": 2e-05,
+      "loss": 0.02981432,
+      "step": 15113
+    },
+    {
+      "epoch": 30.228,
+      "grad_norm": 1.213059425354004,
+      "learning_rate": 2e-05,
+      "loss": 0.0342233,
+      "step": 15114
+    },
+    {
+      "epoch": 30.23,
+      "grad_norm": 1.1020963191986084,
+      "learning_rate": 2e-05,
+      "loss": 0.02755513,
+      "step": 15115
+    },
+    {
+      "epoch": 30.232,
+      "grad_norm": 1.0923129320144653,
+      "learning_rate": 2e-05,
+      "loss": 0.04014743,
+      "step": 15116
+    },
+    {
+      "epoch": 30.234,
+      "grad_norm": 1.1201938390731812,
+      "learning_rate": 2e-05,
+      "loss": 0.05072346,
+      "step": 15117
+    },
+    {
+      "epoch": 30.236,
+      "grad_norm": 3.7093849182128906,
+      "learning_rate": 2e-05,
+      "loss": 0.06261256,
+      "step": 15118
+    },
+    {
+      "epoch": 30.238,
+      "grad_norm": 1.1667890548706055,
+      "learning_rate": 2e-05,
+      "loss": 0.04364653,
+      "step": 15119
+    },
+    {
+      "epoch": 30.24,
+      "grad_norm": 1.1062977313995361,
+      "learning_rate": 2e-05,
+      "loss": 0.05012048,
+      "step": 15120
+    },
+    {
+      "epoch": 30.242,
+      "grad_norm": 1.3190714120864868,
+      "learning_rate": 2e-05,
+      "loss": 0.05197113,
+      "step": 15121
+    },
+    {
+      "epoch": 30.244,
+      "grad_norm": 1.458884596824646,
+      "learning_rate": 2e-05,
+      "loss": 0.04996847,
+      "step": 15122
+    },
+    {
+      "epoch": 30.246,
+      "grad_norm": 1.366876244544983,
+      "learning_rate": 2e-05,
+      "loss": 0.04006827,
+      "step": 15123
+    },
+    {
+      "epoch": 30.248,
+      "grad_norm": 1.0993216037750244,
+      "learning_rate": 2e-05,
+      "loss": 0.03217662,
+      "step": 15124
+    },
+    {
+      "epoch": 30.25,
+      "grad_norm": 0.9922396540641785,
+      "learning_rate": 2e-05,
+      "loss": 0.04653015,
+      "step": 15125
+    },
+    {
+      "epoch": 30.252,
+      "grad_norm": 1.0631383657455444,
+      "learning_rate": 2e-05,
+      "loss": 0.03681334,
+      "step": 15126
+    },
+    {
+      "epoch": 30.254,
+      "grad_norm": 1.0309144258499146,
+      "learning_rate": 2e-05,
+      "loss": 0.04099189,
+      "step": 15127
+    },
+    {
+      "epoch": 30.256,
+      "grad_norm": 1.1436161994934082,
+      "learning_rate": 2e-05,
+      "loss": 0.04641513,
+      "step": 15128
+    },
+    {
+      "epoch": 30.258,
+      "grad_norm": 1.134342074394226,
+      "learning_rate": 2e-05,
+      "loss": 0.03863002,
+      "step": 15129
+    },
+    {
+      "epoch": 30.26,
+      "grad_norm": 1.5008431673049927,
+      "learning_rate": 2e-05,
+      "loss": 0.04123659,
+      "step": 15130
+    },
+    {
+      "epoch": 30.262,
+      "grad_norm": 1.044177770614624,
+      "learning_rate": 2e-05,
+      "loss": 0.03907245,
+      "step": 15131
+    },
+    {
+      "epoch": 30.264,
+      "grad_norm": 0.9937227964401245,
+      "learning_rate": 2e-05,
+      "loss": 0.03749677,
+      "step": 15132
+    },
+    {
+      "epoch": 30.266,
+      "grad_norm": 3.122323989868164,
+      "learning_rate": 2e-05,
+      "loss": 0.05537373,
+      "step": 15133
+    },
+    {
+      "epoch": 30.268,
+      "grad_norm": 1.1054846048355103,
+      "learning_rate": 2e-05,
+      "loss": 0.04427914,
+      "step": 15134
+    },
+    {
+      "epoch": 30.27,
+      "grad_norm": 1.256409764289856,
+      "learning_rate": 2e-05,
+      "loss": 0.05179529,
+      "step": 15135
+    },
+    {
+      "epoch": 30.272,
+      "grad_norm": 1.4733872413635254,
+      "learning_rate": 2e-05,
+      "loss": 0.04531301,
+      "step": 15136
+    },
+    {
+      "epoch": 30.274,
+      "grad_norm": 1.2605996131896973,
+      "learning_rate": 2e-05,
+      "loss": 0.0307104,
+      "step": 15137
+    },
+    {
+      "epoch": 30.276,
+      "grad_norm": 1.0388493537902832,
+      "learning_rate": 2e-05,
+      "loss": 0.04817986,
+      "step": 15138
+    },
+    {
+      "epoch": 30.278,
+      "grad_norm": 1.6238524913787842,
+      "learning_rate": 2e-05,
+      "loss": 0.05085561,
+      "step": 15139
+    },
+    {
+      "epoch": 30.28,
+      "grad_norm": 1.096824288368225,
+      "learning_rate": 2e-05,
+      "loss": 0.04942642,
+      "step": 15140
+    },
+    {
+      "epoch": 30.282,
+      "grad_norm": 1.2300045490264893,
+      "learning_rate": 2e-05,
+      "loss": 0.0502276,
+      "step": 15141
+    },
+    {
+      "epoch": 30.284,
+      "grad_norm": 1.2435781955718994,
+      "learning_rate": 2e-05,
+      "loss": 0.04534131,
+      "step": 15142
+    },
+    {
+      "epoch": 30.286,
+      "grad_norm": 1.386216163635254,
+      "learning_rate": 2e-05,
+      "loss": 0.04588864,
+      "step": 15143
+    },
+    {
+      "epoch": 30.288,
+      "grad_norm": 1.2593432664871216,
+      "learning_rate": 2e-05,
+      "loss": 0.04650512,
+      "step": 15144
+    },
+    {
+      "epoch": 30.29,
+      "grad_norm": 1.6124224662780762,
+      "learning_rate": 2e-05,
+      "loss": 0.04883378,
+      "step": 15145
+    },
+    {
+      "epoch": 30.292,
+      "grad_norm": 2.2996997833251953,
+      "learning_rate": 2e-05,
+      "loss": 0.04866761,
+      "step": 15146
+    },
+    {
+      "epoch": 30.294,
+      "grad_norm": 1.7783503532409668,
+      "learning_rate": 2e-05,
+      "loss": 0.06645172,
+      "step": 15147
+    },
+    {
+      "epoch": 30.296,
+      "grad_norm": 1.2972900867462158,
+      "learning_rate": 2e-05,
+      "loss": 0.05338973,
+      "step": 15148
+    },
+    {
+      "epoch": 30.298,
+      "grad_norm": 2.96677827835083,
+      "learning_rate": 2e-05,
+      "loss": 0.05938792,
+      "step": 15149
+    },
+    {
+      "epoch": 30.3,
+      "grad_norm": 1.1180599927902222,
+      "learning_rate": 2e-05,
+      "loss": 0.03344629,
+      "step": 15150
+    },
+    {
+      "epoch": 30.302,
+      "grad_norm": 1.7869832515716553,
+      "learning_rate": 2e-05,
+      "loss": 0.05831323,
+      "step": 15151
+    },
+    {
+      "epoch": 30.304,
+      "grad_norm": 1.225154161453247,
+      "learning_rate": 2e-05,
+      "loss": 0.05661523,
+      "step": 15152
+    },
+    {
+      "epoch": 30.306,
+      "grad_norm": 2.486720561981201,
+      "learning_rate": 2e-05,
+      "loss": 0.05136902,
+      "step": 15153
+    },
+    {
+      "epoch": 30.308,
+      "grad_norm": 1.1094615459442139,
+      "learning_rate": 2e-05,
+      "loss": 0.04413821,
+      "step": 15154
+    },
+    {
+      "epoch": 30.31,
+      "grad_norm": 2.087968349456787,
+      "learning_rate": 2e-05,
+      "loss": 0.05737863,
+      "step": 15155
+    },
+    {
+      "epoch": 30.312,
+      "grad_norm": 1.0131031274795532,
+      "learning_rate": 2e-05,
+      "loss": 0.04438207,
+      "step": 15156
+    },
+    {
+      "epoch": 30.314,
+      "grad_norm": 1.278663992881775,
+      "learning_rate": 2e-05,
+      "loss": 0.04451298,
+      "step": 15157
+    },
+    {
+      "epoch": 30.316,
+      "grad_norm": 1.0880504846572876,
+      "learning_rate": 2e-05,
+      "loss": 0.04140671,
+      "step": 15158
+    },
+    {
+      "epoch": 30.318,
+      "grad_norm": 1.6311753988265991,
+      "learning_rate": 2e-05,
+      "loss": 0.04358608,
+      "step": 15159
+    },
+    {
+      "epoch": 30.32,
+      "grad_norm": 1.3990247249603271,
+      "learning_rate": 2e-05,
+      "loss": 0.05173497,
+      "step": 15160
+    },
+    {
+      "epoch": 30.322,
+      "grad_norm": 1.4861396551132202,
+      "learning_rate": 2e-05,
+      "loss": 0.03629741,
+      "step": 15161
+    },
+    {
+      "epoch": 30.324,
+      "grad_norm": 1.2476096153259277,
+      "learning_rate": 2e-05,
+      "loss": 0.03799465,
+      "step": 15162
+    },
+    {
+      "epoch": 30.326,
+      "grad_norm": 1.2157728672027588,
+      "learning_rate": 2e-05,
+      "loss": 0.06220795,
+      "step": 15163
+    },
+    {
+      "epoch": 30.328,
+      "grad_norm": 1.5057649612426758,
+      "learning_rate": 2e-05,
+      "loss": 0.04527867,
+      "step": 15164
+    },
+    {
+      "epoch": 30.33,
+      "grad_norm": 1.7932603359222412,
+      "learning_rate": 2e-05,
+      "loss": 0.05007517,
+      "step": 15165
+    },
+    {
+      "epoch": 30.332,
+      "grad_norm": 1.5305033922195435,
+      "learning_rate": 2e-05,
+      "loss": 0.03679098,
+      "step": 15166
+    },
+    {
+      "epoch": 30.334,
+      "grad_norm": 1.2285985946655273,
+      "learning_rate": 2e-05,
+      "loss": 0.0510888,
+      "step": 15167
+    },
+    {
+      "epoch": 30.336,
+      "grad_norm": 1.0248448848724365,
+      "learning_rate": 2e-05,
+      "loss": 0.04463315,
+      "step": 15168
+    },
+    {
+      "epoch": 30.338,
+      "grad_norm": 1.1598585844039917,
+      "learning_rate": 2e-05,
+      "loss": 0.04191558,
+      "step": 15169
+    },
+    {
+      "epoch": 30.34,
+      "grad_norm": 1.415334939956665,
+      "learning_rate": 2e-05,
+      "loss": 0.04389124,
+      "step": 15170
+    },
+    {
+      "epoch": 30.342,
+      "grad_norm": 1.105470895767212,
+      "learning_rate": 2e-05,
+      "loss": 0.04412735,
+      "step": 15171
+    },
+    {
+      "epoch": 30.344,
+      "grad_norm": 2.4653127193450928,
+      "learning_rate": 2e-05,
+      "loss": 0.04922763,
+      "step": 15172
+    },
+    {
+      "epoch": 30.346,
+      "grad_norm": 0.9219580292701721,
+      "learning_rate": 2e-05,
+      "loss": 0.02929104,
+      "step": 15173
+    },
+    {
+      "epoch": 30.348,
+      "grad_norm": 1.086243987083435,
+      "learning_rate": 2e-05,
+      "loss": 0.04635917,
+      "step": 15174
+    },
+    {
+      "epoch": 30.35,
+      "grad_norm": 0.9130674004554749,
+      "learning_rate": 2e-05,
+      "loss": 0.031245,
+      "step": 15175
+    },
+    {
+      "epoch": 30.352,
+      "grad_norm": 0.9588483572006226,
+      "learning_rate": 2e-05,
+      "loss": 0.03446349,
+      "step": 15176
+    },
+    {
+      "epoch": 30.354,
+      "grad_norm": 1.1762733459472656,
+      "learning_rate": 2e-05,
+      "loss": 0.04649212,
+      "step": 15177
+    },
+    {
+      "epoch": 30.356,
+      "grad_norm": 1.3569097518920898,
+      "learning_rate": 2e-05,
+      "loss": 0.05233772,
+      "step": 15178
+    },
+    {
+      "epoch": 30.358,
+      "grad_norm": 1.1183359622955322,
+      "learning_rate": 2e-05,
+      "loss": 0.02675498,
+      "step": 15179
+    },
+    {
+      "epoch": 30.36,
+      "grad_norm": 1.455556869506836,
+      "learning_rate": 2e-05,
+      "loss": 0.05393392,
+      "step": 15180
+    },
+    {
+      "epoch": 30.362,
+      "grad_norm": 1.3345918655395508,
+      "learning_rate": 2e-05,
+      "loss": 0.04711598,
+      "step": 15181
+    },
+    {
+      "epoch": 30.364,
+      "grad_norm": 1.4557232856750488,
+      "learning_rate": 2e-05,
+      "loss": 0.05467194,
+      "step": 15182
+    },
+    {
+      "epoch": 30.366,
+      "grad_norm": 1.2925028800964355,
+      "learning_rate": 2e-05,
+      "loss": 0.04946,
+      "step": 15183
+    },
+    {
+      "epoch": 30.368,
+      "grad_norm": 1.2487906217575073,
+      "learning_rate": 2e-05,
+      "loss": 0.04537457,
+      "step": 15184
+    },
+    {
+      "epoch": 30.37,
+      "grad_norm": 1.4377646446228027,
+      "learning_rate": 2e-05,
+      "loss": 0.05969262,
+      "step": 15185
+    },
+    {
+      "epoch": 30.372,
+      "grad_norm": 1.0396651029586792,
+      "learning_rate": 2e-05,
+      "loss": 0.03547468,
+      "step": 15186
+    },
+    {
+      "epoch": 30.374,
+      "grad_norm": 1.273114562034607,
+      "learning_rate": 2e-05,
+      "loss": 0.04310979,
+      "step": 15187
+    },
+    {
+      "epoch": 30.376,
+      "grad_norm": 0.977163553237915,
+      "learning_rate": 2e-05,
+      "loss": 0.03705889,
+      "step": 15188
+    },
+    {
+      "epoch": 30.378,
+      "grad_norm": 2.641759157180786,
+      "learning_rate": 2e-05,
+      "loss": 0.05824747,
+      "step": 15189
+    },
+    {
+      "epoch": 30.38,
+      "grad_norm": 2.5414299964904785,
+      "learning_rate": 2e-05,
+      "loss": 0.04271651,
+      "step": 15190
+    },
+    {
+      "epoch": 30.382,
+      "grad_norm": 1.6944334506988525,
+      "learning_rate": 2e-05,
+      "loss": 0.0411639,
+      "step": 15191
+    },
+    {
+      "epoch": 30.384,
+      "grad_norm": 1.1185425519943237,
+      "learning_rate": 2e-05,
+      "loss": 0.03539154,
+      "step": 15192
+    },
+    {
+      "epoch": 30.386,
+      "grad_norm": 2.133718252182007,
+      "learning_rate": 2e-05,
+      "loss": 0.03833131,
+      "step": 15193
+    },
+    {
+      "epoch": 30.388,
+      "grad_norm": 1.6777894496917725,
+      "learning_rate": 2e-05,
+      "loss": 0.04095278,
+      "step": 15194
+    },
+    {
+      "epoch": 30.39,
+      "grad_norm": 2.4663455486297607,
+      "learning_rate": 2e-05,
+      "loss": 0.03233901,
+      "step": 15195
+    },
+    {
+      "epoch": 30.392,
+      "grad_norm": 1.5365792512893677,
+      "learning_rate": 2e-05,
+      "loss": 0.05167238,
+      "step": 15196
+    },
+    {
+      "epoch": 30.394,
+      "grad_norm": 1.1537460088729858,
+      "learning_rate": 2e-05,
+      "loss": 0.03557686,
+      "step": 15197
+    },
+    {
+      "epoch": 30.396,
+      "grad_norm": 0.9411190152168274,
+      "learning_rate": 2e-05,
+      "loss": 0.02952364,
+      "step": 15198
+    },
+    {
+      "epoch": 30.398,
+      "grad_norm": 1.2708302736282349,
+      "learning_rate": 2e-05,
+      "loss": 0.05088828,
+      "step": 15199
+    },
+    {
+      "epoch": 30.4,
+      "grad_norm": 2.0306994915008545,
+      "learning_rate": 2e-05,
+      "loss": 0.04890837,
+      "step": 15200
+    },
+    {
+      "epoch": 30.402,
+      "grad_norm": 1.7756644487380981,
+      "learning_rate": 2e-05,
+      "loss": 0.04039244,
+      "step": 15201
+    },
+    {
+      "epoch": 30.404,
+      "grad_norm": 1.4186558723449707,
+      "learning_rate": 2e-05,
+      "loss": 0.05239846,
+      "step": 15202
+    },
+    {
+      "epoch": 30.406,
+      "grad_norm": 1.0431525707244873,
+      "learning_rate": 2e-05,
+      "loss": 0.03030356,
+      "step": 15203
+    },
+    {
+      "epoch": 30.408,
+      "grad_norm": 2.5661303997039795,
+      "learning_rate": 2e-05,
+      "loss": 0.06191105,
+      "step": 15204
+    },
+    {
+      "epoch": 30.41,
+      "grad_norm": 1.5152759552001953,
+      "learning_rate": 2e-05,
+      "loss": 0.0515747,
+      "step": 15205
+    },
+    {
+      "epoch": 30.412,
+      "grad_norm": 1.1445180177688599,
+      "learning_rate": 2e-05,
+      "loss": 0.04189182,
+      "step": 15206
+    },
+    {
+      "epoch": 30.414,
+      "grad_norm": 1.1220874786376953,
+      "learning_rate": 2e-05,
+      "loss": 0.04300126,
+      "step": 15207
+    },
+    {
+      "epoch": 30.416,
+      "grad_norm": 1.1378350257873535,
+      "learning_rate": 2e-05,
+      "loss": 0.03469217,
+      "step": 15208
+    },
+    {
+      "epoch": 30.418,
+      "grad_norm": 1.337202548980713,
+      "learning_rate": 2e-05,
+      "loss": 0.03686753,
+      "step": 15209
+    },
+    {
+      "epoch": 30.42,
+      "grad_norm": 0.9815614223480225,
+      "learning_rate": 2e-05,
+      "loss": 0.03325251,
+      "step": 15210
+    },
+    {
+      "epoch": 30.422,
+      "grad_norm": 0.8998761773109436,
+      "learning_rate": 2e-05,
+      "loss": 0.03149875,
+      "step": 15211
+    },
+    {
+      "epoch": 30.424,
+      "grad_norm": 1.0622773170471191,
+      "learning_rate": 2e-05,
+      "loss": 0.04623077,
+      "step": 15212
+    },
+    {
+      "epoch": 30.426,
+      "grad_norm": 2.55696177482605,
+      "learning_rate": 2e-05,
+      "loss": 0.05916242,
+      "step": 15213
+    },
+    {
+      "epoch": 30.428,
+      "grad_norm": 1.0451836585998535,
+      "learning_rate": 2e-05,
+      "loss": 0.03193265,
+      "step": 15214
+    },
+    {
+      "epoch": 30.43,
+      "grad_norm": 1.0104472637176514,
+      "learning_rate": 2e-05,
+      "loss": 0.04023196,
+      "step": 15215
+    },
+    {
+      "epoch": 30.432,
+      "grad_norm": 1.5846662521362305,
+      "learning_rate": 2e-05,
+      "loss": 0.04536471,
+      "step": 15216
+    },
+    {
+      "epoch": 30.434,
+      "grad_norm": 1.2485827207565308,
+      "learning_rate": 2e-05,
+      "loss": 0.05067682,
+      "step": 15217
+    },
+    {
+      "epoch": 30.436,
+      "grad_norm": 2.142920732498169,
+      "learning_rate": 2e-05,
+      "loss": 0.04851421,
+      "step": 15218
+    },
+    {
+      "epoch": 30.438,
+      "grad_norm": 2.670254945755005,
+      "learning_rate": 2e-05,
+      "loss": 0.03117582,
+      "step": 15219
+    },
+    {
+      "epoch": 30.44,
+      "grad_norm": 2.0113894939422607,
+      "learning_rate": 2e-05,
+      "loss": 0.04463819,
+      "step": 15220
+    },
+    {
+      "epoch": 30.442,
+      "grad_norm": 1.347212791442871,
+      "learning_rate": 2e-05,
+      "loss": 0.03855886,
+      "step": 15221
+    },
+    {
+      "epoch": 30.444,
+      "grad_norm": 1.3231719732284546,
+      "learning_rate": 2e-05,
+      "loss": 0.04579235,
+      "step": 15222
+    },
+    {
+      "epoch": 30.446,
+      "grad_norm": 1.3619619607925415,
+      "learning_rate": 2e-05,
+      "loss": 0.05623931,
+      "step": 15223
+    },
+    {
+      "epoch": 30.448,
+      "grad_norm": 1.569140076637268,
+      "learning_rate": 2e-05,
+      "loss": 0.05256468,
+      "step": 15224
+    },
+    {
+      "epoch": 30.45,
+      "grad_norm": 1.5590763092041016,
+      "learning_rate": 2e-05,
+      "loss": 0.04745616,
+      "step": 15225
+    },
+    {
+      "epoch": 30.452,
+      "grad_norm": 1.8052546977996826,
+      "learning_rate": 2e-05,
+      "loss": 0.03997829,
+      "step": 15226
+    },
+    {
+      "epoch": 30.454,
+      "grad_norm": 1.4032646417617798,
+      "learning_rate": 2e-05,
+      "loss": 0.03643226,
+      "step": 15227
+    },
+    {
+      "epoch": 30.456,
+      "grad_norm": 1.6574431657791138,
+      "learning_rate": 2e-05,
+      "loss": 0.04806088,
+      "step": 15228
+    },
+    {
+      "epoch": 30.458,
+      "grad_norm": 1.1438560485839844,
+      "learning_rate": 2e-05,
+      "loss": 0.04555596,
+      "step": 15229
+    },
+    {
+      "epoch": 30.46,
+      "grad_norm": 1.6412994861602783,
+      "learning_rate": 2e-05,
+      "loss": 0.05041576,
+      "step": 15230
+    },
+    {
+      "epoch": 30.462,
+      "grad_norm": 1.7027461528778076,
+      "learning_rate": 2e-05,
+      "loss": 0.04613438,
+      "step": 15231
+    },
+    {
+      "epoch": 30.464,
+      "grad_norm": 1.8641504049301147,
+      "learning_rate": 2e-05,
+      "loss": 0.05400945,
+      "step": 15232
+    },
+    {
+      "epoch": 30.466,
+      "grad_norm": 1.0201693773269653,
+      "learning_rate": 2e-05,
+      "loss": 0.03487154,
+      "step": 15233
+    },
+    {
+      "epoch": 30.468,
+      "grad_norm": 3.329221248626709,
+      "learning_rate": 2e-05,
+      "loss": 0.0438869,
+      "step": 15234
+    },
+    {
+      "epoch": 30.47,
+      "grad_norm": 2.2483973503112793,
+      "learning_rate": 2e-05,
+      "loss": 0.05888098,
+      "step": 15235
+    },
+    {
+      "epoch": 30.472,
+      "grad_norm": 1.1053797006607056,
+      "learning_rate": 2e-05,
+      "loss": 0.02992825,
+      "step": 15236
+    },
+    {
+      "epoch": 30.474,
+      "grad_norm": 1.3227661848068237,
+      "learning_rate": 2e-05,
+      "loss": 0.0443338,
+      "step": 15237
+    },
+    {
+      "epoch": 30.476,
+      "grad_norm": 1.077270269393921,
+      "learning_rate": 2e-05,
+      "loss": 0.03753895,
+      "step": 15238
+    },
+    {
+      "epoch": 30.478,
+      "grad_norm": 1.0654442310333252,
+      "learning_rate": 2e-05,
+      "loss": 0.04142526,
+      "step": 15239
+    },
+    {
+      "epoch": 30.48,
+      "grad_norm": 1.0802520513534546,
+      "learning_rate": 2e-05,
+      "loss": 0.04611638,
+      "step": 15240
+    },
+    {
+      "epoch": 30.482,
+      "grad_norm": 1.2629036903381348,
+      "learning_rate": 2e-05,
+      "loss": 0.06196934,
+      "step": 15241
+    },
+    {
+      "epoch": 30.484,
+      "grad_norm": 1.756333589553833,
+      "learning_rate": 2e-05,
+      "loss": 0.04763553,
+      "step": 15242
+    },
+    {
+      "epoch": 30.486,
+      "grad_norm": 1.0675008296966553,
+      "learning_rate": 2e-05,
+      "loss": 0.03430801,
+      "step": 15243
+    },
+    {
+      "epoch": 30.488,
+      "grad_norm": 5.384986400604248,
+      "learning_rate": 2e-05,
+      "loss": 0.03727914,
+      "step": 15244
+    },
+    {
+      "epoch": 30.49,
+      "grad_norm": 1.504477620124817,
+      "learning_rate": 2e-05,
+      "loss": 0.05895821,
+      "step": 15245
+    },
+    {
+      "epoch": 30.492,
+      "grad_norm": 1.3034749031066895,
+      "learning_rate": 2e-05,
+      "loss": 0.04594578,
+      "step": 15246
+    },
+    {
+      "epoch": 30.494,
+      "grad_norm": 1.38698148727417,
+      "learning_rate": 2e-05,
+      "loss": 0.05439015,
+      "step": 15247
+    },
+    {
+      "epoch": 30.496,
+      "grad_norm": 1.0708712339401245,
+      "learning_rate": 2e-05,
+      "loss": 0.03840674,
+      "step": 15248
+    },
+    {
+      "epoch": 30.498,
+      "grad_norm": 1.4558812379837036,
+      "learning_rate": 2e-05,
+      "loss": 0.02542023,
+      "step": 15249
+    },
+    {
+      "epoch": 30.5,
+      "grad_norm": 1.7556490898132324,
+      "learning_rate": 2e-05,
+      "loss": 0.0409022,
+      "step": 15250
+    },
+    {
+      "epoch": 30.502,
+      "grad_norm": 1.2140228748321533,
+      "learning_rate": 2e-05,
+      "loss": 0.03575923,
+      "step": 15251
+    },
+    {
+      "epoch": 30.504,
+      "grad_norm": 0.9344707131385803,
+      "learning_rate": 2e-05,
+      "loss": 0.03167684,
+      "step": 15252
+    },
+    {
+      "epoch": 30.506,
+      "grad_norm": 1.3499372005462646,
+      "learning_rate": 2e-05,
+      "loss": 0.03214483,
+      "step": 15253
+    },
+    {
+      "epoch": 30.508,
+      "grad_norm": 1.1017028093338013,
+      "learning_rate": 2e-05,
+      "loss": 0.02525232,
+      "step": 15254
+    },
+    {
+      "epoch": 30.51,
+      "grad_norm": 1.4885841608047485,
+      "learning_rate": 2e-05,
+      "loss": 0.06516996,
+      "step": 15255
+    },
+    {
+      "epoch": 30.512,
+      "grad_norm": 1.4931526184082031,
+      "learning_rate": 2e-05,
+      "loss": 0.06183759,
+      "step": 15256
+    },
+    {
+      "epoch": 30.514,
+      "grad_norm": 0.9158805012702942,
+      "learning_rate": 2e-05,
+      "loss": 0.03264806,
+      "step": 15257
+    },
+    {
+      "epoch": 30.516,
+      "grad_norm": 1.7295507192611694,
+      "learning_rate": 2e-05,
+      "loss": 0.0437915,
+      "step": 15258
+    },
+    {
+      "epoch": 30.518,
+      "grad_norm": 1.2573812007904053,
+      "learning_rate": 2e-05,
+      "loss": 0.04718456,
+      "step": 15259
+    },
+    {
+      "epoch": 30.52,
+      "grad_norm": 1.111760139465332,
+      "learning_rate": 2e-05,
+      "loss": 0.03778823,
+      "step": 15260
+    },
+    {
+      "epoch": 30.522,
+      "grad_norm": 1.3841087818145752,
+      "learning_rate": 2e-05,
+      "loss": 0.05139244,
+      "step": 15261
+    },
+    {
+      "epoch": 30.524,
+      "grad_norm": 1.2217211723327637,
+      "learning_rate": 2e-05,
+      "loss": 0.05407543,
+      "step": 15262
+    },
+    {
+      "epoch": 30.526,
+      "grad_norm": 1.9204769134521484,
+      "learning_rate": 2e-05,
+      "loss": 0.05501651,
+      "step": 15263
+    },
+    {
+      "epoch": 30.528,
+      "grad_norm": 1.0646864175796509,
+      "learning_rate": 2e-05,
+      "loss": 0.04272152,
+      "step": 15264
+    },
+    {
+      "epoch": 30.53,
+      "grad_norm": 1.138542652130127,
+      "learning_rate": 2e-05,
+      "loss": 0.04351484,
+      "step": 15265
+    },
+    {
+      "epoch": 30.532,
+      "grad_norm": 1.1147321462631226,
+      "learning_rate": 2e-05,
+      "loss": 0.03230875,
+      "step": 15266
+    },
+    {
+      "epoch": 30.534,
+      "grad_norm": 1.3464140892028809,
+      "learning_rate": 2e-05,
+      "loss": 0.05336832,
+      "step": 15267
+    },
+    {
+      "epoch": 30.536,
+      "grad_norm": 1.3282221555709839,
+      "learning_rate": 2e-05,
+      "loss": 0.06214736,
+      "step": 15268
+    },
+    {
+      "epoch": 30.538,
+      "grad_norm": 1.1814717054367065,
+      "learning_rate": 2e-05,
+      "loss": 0.03814064,
+      "step": 15269
+    },
+    {
+      "epoch": 30.54,
+      "grad_norm": 1.167283535003662,
+      "learning_rate": 2e-05,
+      "loss": 0.04938902,
+      "step": 15270
+    },
+    {
+      "epoch": 30.542,
+      "grad_norm": 1.5443036556243896,
+      "learning_rate": 2e-05,
+      "loss": 0.04786984,
+      "step": 15271
+    },
+    {
+      "epoch": 30.544,
+      "grad_norm": 1.666343331336975,
+      "learning_rate": 2e-05,
+      "loss": 0.05091058,
+      "step": 15272
+    },
+    {
+      "epoch": 30.546,
+      "grad_norm": 1.4363032579421997,
+      "learning_rate": 2e-05,
+      "loss": 0.04783217,
+      "step": 15273
+    },
+    {
+      "epoch": 30.548000000000002,
+      "grad_norm": 1.348711371421814,
+      "learning_rate": 2e-05,
+      "loss": 0.05052876,
+      "step": 15274
+    },
+    {
+      "epoch": 30.55,
+      "grad_norm": 1.2497211694717407,
+      "learning_rate": 2e-05,
+      "loss": 0.03438094,
+      "step": 15275
+    },
+    {
+      "epoch": 30.552,
+      "grad_norm": 1.3024879693984985,
+      "learning_rate": 2e-05,
+      "loss": 0.04458308,
+      "step": 15276
+    },
+    {
+      "epoch": 30.554,
+      "grad_norm": 1.4879077672958374,
+      "learning_rate": 2e-05,
+      "loss": 0.03748794,
+      "step": 15277
+    },
+    {
+      "epoch": 30.556,
+      "grad_norm": 1.565683126449585,
+      "learning_rate": 2e-05,
+      "loss": 0.03531917,
+      "step": 15278
+    },
+    {
+      "epoch": 30.558,
+      "grad_norm": 1.226104974746704,
+      "learning_rate": 2e-05,
+      "loss": 0.04298712,
+      "step": 15279
+    },
+    {
+      "epoch": 30.56,
+      "grad_norm": 1.5387117862701416,
+      "learning_rate": 2e-05,
+      "loss": 0.04004962,
+      "step": 15280
+    },
+    {
+      "epoch": 30.562,
+      "grad_norm": 1.0032356977462769,
+      "learning_rate": 2e-05,
+      "loss": 0.03250517,
+      "step": 15281
+    },
+    {
+      "epoch": 30.564,
+      "grad_norm": 1.0234278440475464,
+      "learning_rate": 2e-05,
+      "loss": 0.04238243,
+      "step": 15282
+    },
+    {
+      "epoch": 30.566,
+      "grad_norm": 1.151049256324768,
+      "learning_rate": 2e-05,
+      "loss": 0.04576783,
+      "step": 15283
+    },
+    {
+      "epoch": 30.568,
+      "grad_norm": 2.4301161766052246,
+      "learning_rate": 2e-05,
+      "loss": 0.04964666,
+      "step": 15284
+    },
+    {
+      "epoch": 30.57,
+      "grad_norm": 1.798929214477539,
+      "learning_rate": 2e-05,
+      "loss": 0.04637855,
+      "step": 15285
+    },
+    {
+      "epoch": 30.572,
+      "grad_norm": 1.205482006072998,
+      "learning_rate": 2e-05,
+      "loss": 0.04132317,
+      "step": 15286
+    },
+    {
+      "epoch": 30.574,
+      "grad_norm": 1.730236530303955,
+      "learning_rate": 2e-05,
+      "loss": 0.04239135,
+      "step": 15287
+    },
+    {
+      "epoch": 30.576,
+      "grad_norm": 1.3888801336288452,
+      "learning_rate": 2e-05,
+      "loss": 0.04025697,
+      "step": 15288
+    },
+    {
+      "epoch": 30.578,
+      "grad_norm": 1.2676804065704346,
+      "learning_rate": 2e-05,
+      "loss": 0.05485225,
+      "step": 15289
+    },
+    {
+      "epoch": 30.58,
+      "grad_norm": 1.2869915962219238,
+      "learning_rate": 2e-05,
+      "loss": 0.04968848,
+      "step": 15290
+    },
+    {
+      "epoch": 30.582,
+      "grad_norm": 1.0613614320755005,
+      "learning_rate": 2e-05,
+      "loss": 0.03498927,
+      "step": 15291
+    },
+    {
+      "epoch": 30.584,
+      "grad_norm": 1.000826120376587,
+      "learning_rate": 2e-05,
+      "loss": 0.03525396,
+      "step": 15292
+    },
+    {
+      "epoch": 30.586,
+      "grad_norm": 1.5731277465820312,
+      "learning_rate": 2e-05,
+      "loss": 0.05101633,
+      "step": 15293
+    },
+    {
+      "epoch": 30.588,
+      "grad_norm": 1.1912803649902344,
+      "learning_rate": 2e-05,
+      "loss": 0.04636155,
+      "step": 15294
+    },
+    {
+      "epoch": 30.59,
+      "grad_norm": 1.1083427667617798,
+      "learning_rate": 2e-05,
+      "loss": 0.04011728,
+      "step": 15295
+    },
+    {
+      "epoch": 30.592,
+      "grad_norm": 0.8691169023513794,
+      "learning_rate": 2e-05,
+      "loss": 0.02712849,
+      "step": 15296
+    },
+    {
+      "epoch": 30.594,
+      "grad_norm": 1.5057884454727173,
+      "learning_rate": 2e-05,
+      "loss": 0.05116381,
+      "step": 15297
+    },
+    {
+      "epoch": 30.596,
+      "grad_norm": 1.8393473625183105,
+      "learning_rate": 2e-05,
+      "loss": 0.05369818,
+      "step": 15298
+    },
+    {
+      "epoch": 30.598,
+      "grad_norm": 1.4146645069122314,
+      "learning_rate": 2e-05,
+      "loss": 0.04586646,
+      "step": 15299
+    },
+    {
+      "epoch": 30.6,
+      "grad_norm": 1.0296281576156616,
+      "learning_rate": 2e-05,
+      "loss": 0.02862342,
+      "step": 15300
+    },
+    {
+      "epoch": 30.602,
+      "grad_norm": 1.4884545803070068,
+      "learning_rate": 2e-05,
+      "loss": 0.03266924,
+      "step": 15301
+    },
+    {
+      "epoch": 30.604,
+      "grad_norm": 1.6892744302749634,
+      "learning_rate": 2e-05,
+      "loss": 0.03972404,
+      "step": 15302
+    },
+    {
+      "epoch": 30.606,
+      "grad_norm": 1.5136027336120605,
+      "learning_rate": 2e-05,
+      "loss": 0.02378543,
+      "step": 15303
+    },
+    {
+      "epoch": 30.608,
+      "grad_norm": 1.1681137084960938,
+      "learning_rate": 2e-05,
+      "loss": 0.0225147,
+      "step": 15304
+    },
+    {
+      "epoch": 30.61,
+      "grad_norm": 1.0413873195648193,
+      "learning_rate": 2e-05,
+      "loss": 0.03786266,
+      "step": 15305
+    },
+    {
+      "epoch": 30.612,
+      "grad_norm": 1.3476159572601318,
+      "learning_rate": 2e-05,
+      "loss": 0.0479761,
+      "step": 15306
+    },
+    {
+      "epoch": 30.614,
+      "grad_norm": 1.4096819162368774,
+      "learning_rate": 2e-05,
+      "loss": 0.04315577,
+      "step": 15307
+    },
+    {
+      "epoch": 30.616,
+      "grad_norm": 1.1657134294509888,
+      "learning_rate": 2e-05,
+      "loss": 0.03629996,
+      "step": 15308
+    },
+    {
+      "epoch": 30.618,
+      "grad_norm": 1.0118683576583862,
+      "learning_rate": 2e-05,
+      "loss": 0.03799508,
+      "step": 15309
+    },
+    {
+      "epoch": 30.62,
+      "grad_norm": 1.3761227130889893,
+      "learning_rate": 2e-05,
+      "loss": 0.04091384,
+      "step": 15310
+    },
+    {
+      "epoch": 30.622,
+      "grad_norm": 1.3315359354019165,
+      "learning_rate": 2e-05,
+      "loss": 0.05451095,
+      "step": 15311
+    },
+    {
+      "epoch": 30.624,
+      "grad_norm": 1.0484113693237305,
+      "learning_rate": 2e-05,
+      "loss": 0.04014854,
+      "step": 15312
+    },
+    {
+      "epoch": 30.626,
+      "grad_norm": 1.0485554933547974,
+      "learning_rate": 2e-05,
+      "loss": 0.03734627,
+      "step": 15313
+    },
+    {
+      "epoch": 30.628,
+      "grad_norm": 1.7621971368789673,
+      "learning_rate": 2e-05,
+      "loss": 0.06587996,
+      "step": 15314
+    },
+    {
+      "epoch": 30.63,
+      "grad_norm": 1.1702368259429932,
+      "learning_rate": 2e-05,
+      "loss": 0.03241003,
+      "step": 15315
+    },
+    {
+      "epoch": 30.632,
+      "grad_norm": 0.9568459987640381,
+      "learning_rate": 2e-05,
+      "loss": 0.03606553,
+      "step": 15316
+    },
+    {
+      "epoch": 30.634,
+      "grad_norm": 1.383159875869751,
+      "learning_rate": 2e-05,
+      "loss": 0.05273655,
+      "step": 15317
+    },
+    {
+      "epoch": 30.636,
+      "grad_norm": 1.2365126609802246,
+      "learning_rate": 2e-05,
+      "loss": 0.03494096,
+      "step": 15318
+    },
+    {
+      "epoch": 30.638,
+      "grad_norm": 1.046418309211731,
+      "learning_rate": 2e-05,
+      "loss": 0.03035589,
+      "step": 15319
+    },
+    {
+      "epoch": 30.64,
+      "grad_norm": 1.7063677310943604,
+      "learning_rate": 2e-05,
+      "loss": 0.04815275,
+      "step": 15320
+    },
+    {
+      "epoch": 30.642,
+      "grad_norm": 0.9858196973800659,
+      "learning_rate": 2e-05,
+      "loss": 0.0392507,
+      "step": 15321
+    },
+    {
+      "epoch": 30.644,
+      "grad_norm": 2.5675480365753174,
+      "learning_rate": 2e-05,
+      "loss": 0.06889924,
+      "step": 15322
+    },
+    {
+      "epoch": 30.646,
+      "grad_norm": 1.2175625562667847,
+      "learning_rate": 2e-05,
+      "loss": 0.04431954,
+      "step": 15323
+    },
+    {
+      "epoch": 30.648,
+      "grad_norm": 1.0321712493896484,
+      "learning_rate": 2e-05,
+      "loss": 0.03850232,
+      "step": 15324
+    },
+    {
+      "epoch": 30.65,
+      "grad_norm": 1.097364068031311,
+      "learning_rate": 2e-05,
+      "loss": 0.04117165,
+      "step": 15325
+    },
+    {
+      "epoch": 30.652,
+      "grad_norm": 1.0748999118804932,
+      "learning_rate": 2e-05,
+      "loss": 0.03478822,
+      "step": 15326
+    },
+    {
+      "epoch": 30.654,
+      "grad_norm": 3.8500382900238037,
+      "learning_rate": 2e-05,
+      "loss": 0.03891071,
+      "step": 15327
+    },
+    {
+      "epoch": 30.656,
+      "grad_norm": 1.0356502532958984,
+      "learning_rate": 2e-05,
+      "loss": 0.04003483,
+      "step": 15328
+    },
+    {
+      "epoch": 30.658,
+      "grad_norm": 1.2226482629776,
+      "learning_rate": 2e-05,
+      "loss": 0.04361072,
+      "step": 15329
+    },
+    {
+      "epoch": 30.66,
+      "grad_norm": 3.1957709789276123,
+      "learning_rate": 2e-05,
+      "loss": 0.04747942,
+      "step": 15330
+    },
+    {
+      "epoch": 30.662,
+      "grad_norm": 1.8828949928283691,
+      "learning_rate": 2e-05,
+      "loss": 0.05194776,
+      "step": 15331
+    },
+    {
+      "epoch": 30.664,
+      "grad_norm": 1.5473778247833252,
+      "learning_rate": 2e-05,
+      "loss": 0.05460526,
+      "step": 15332
+    },
+    {
+      "epoch": 30.666,
+      "grad_norm": 1.3898783922195435,
+      "learning_rate": 2e-05,
+      "loss": 0.05540655,
+      "step": 15333
+    },
+    {
+      "epoch": 30.668,
+      "grad_norm": 0.9823468923568726,
+      "learning_rate": 2e-05,
+      "loss": 0.02938699,
+      "step": 15334
+    },
+    {
+      "epoch": 30.67,
+      "grad_norm": 2.376797676086426,
+      "learning_rate": 2e-05,
+      "loss": 0.03299019,
+      "step": 15335
+    },
+    {
+      "epoch": 30.672,
+      "grad_norm": 1.2253479957580566,
+      "learning_rate": 2e-05,
+      "loss": 0.03981312,
+      "step": 15336
+    },
+    {
+      "epoch": 30.674,
+      "grad_norm": 1.2062132358551025,
+      "learning_rate": 2e-05,
+      "loss": 0.03478318,
+      "step": 15337
+    },
+    {
+      "epoch": 30.676,
+      "grad_norm": 0.9848882555961609,
+      "learning_rate": 2e-05,
+      "loss": 0.03393116,
+      "step": 15338
+    },
+    {
+      "epoch": 30.678,
+      "grad_norm": 1.059923768043518,
+      "learning_rate": 2e-05,
+      "loss": 0.03325968,
+      "step": 15339
+    },
+    {
+      "epoch": 30.68,
+      "grad_norm": 1.6200644969940186,
+      "learning_rate": 2e-05,
+      "loss": 0.03570171,
+      "step": 15340
+    },
+    {
+      "epoch": 30.682,
+      "grad_norm": 1.2216757535934448,
+      "learning_rate": 2e-05,
+      "loss": 0.04476602,
+      "step": 15341
+    },
+    {
+      "epoch": 30.684,
+      "grad_norm": 2.0105185508728027,
+      "learning_rate": 2e-05,
+      "loss": 0.04382947,
+      "step": 15342
+    },
+    {
+      "epoch": 30.686,
+      "grad_norm": 1.8764110803604126,
+      "learning_rate": 2e-05,
+      "loss": 0.05112007,
+      "step": 15343
+    },
+    {
+      "epoch": 30.688,
+      "grad_norm": 1.9776445627212524,
+      "learning_rate": 2e-05,
+      "loss": 0.05051792,
+      "step": 15344
+    },
+    {
+      "epoch": 30.69,
+      "grad_norm": 1.2636760473251343,
+      "learning_rate": 2e-05,
+      "loss": 0.03654606,
+      "step": 15345
+    },
+    {
+      "epoch": 30.692,
+      "grad_norm": 1.58008873462677,
+      "learning_rate": 2e-05,
+      "loss": 0.05276753,
+      "step": 15346
+    },
+    {
+      "epoch": 30.694,
+      "grad_norm": 1.1315622329711914,
+      "learning_rate": 2e-05,
+      "loss": 0.04263277,
+      "step": 15347
+    },
+    {
+      "epoch": 30.696,
+      "grad_norm": 0.9804048538208008,
+      "learning_rate": 2e-05,
+      "loss": 0.03775415,
+      "step": 15348
+    },
+    {
+      "epoch": 30.698,
+      "grad_norm": 1.381704568862915,
+      "learning_rate": 2e-05,
+      "loss": 0.03425687,
+      "step": 15349
+    },
+    {
+      "epoch": 30.7,
+      "grad_norm": 0.9259615540504456,
+      "learning_rate": 2e-05,
+      "loss": 0.02952418,
+      "step": 15350
+    },
+    {
+      "epoch": 30.701999999999998,
+      "grad_norm": 1.5775662660598755,
+      "learning_rate": 2e-05,
+      "loss": 0.05005368,
+      "step": 15351
+    },
+    {
+      "epoch": 30.704,
+      "grad_norm": 1.3794008493423462,
+      "learning_rate": 2e-05,
+      "loss": 0.0453692,
+      "step": 15352
+    },
+    {
+      "epoch": 30.706,
+      "grad_norm": 1.189494013786316,
+      "learning_rate": 2e-05,
+      "loss": 0.04057778,
+      "step": 15353
+    },
+    {
+      "epoch": 30.708,
+      "grad_norm": 0.9436183571815491,
+      "learning_rate": 2e-05,
+      "loss": 0.02952637,
+      "step": 15354
+    },
+    {
+      "epoch": 30.71,
+      "grad_norm": 1.8463188409805298,
+      "learning_rate": 2e-05,
+      "loss": 0.05099573,
+      "step": 15355
+    },
+    {
+      "epoch": 30.712,
+      "grad_norm": 1.0513994693756104,
+      "learning_rate": 2e-05,
+      "loss": 0.03942338,
+      "step": 15356
+    },
+    {
+      "epoch": 30.714,
+      "grad_norm": 1.306980013847351,
+      "learning_rate": 2e-05,
+      "loss": 0.04591175,
+      "step": 15357
+    },
+    {
+      "epoch": 30.716,
+      "grad_norm": 1.1594239473342896,
+      "learning_rate": 2e-05,
+      "loss": 0.03216893,
+      "step": 15358
+    },
+    {
+      "epoch": 30.718,
+      "grad_norm": 1.6709867715835571,
+      "learning_rate": 2e-05,
+      "loss": 0.04788967,
+      "step": 15359
+    },
+    {
+      "epoch": 30.72,
+      "grad_norm": 1.4016337394714355,
+      "learning_rate": 2e-05,
+      "loss": 0.04345062,
+      "step": 15360
+    },
+    {
+      "epoch": 30.722,
+      "grad_norm": 1.594774603843689,
+      "learning_rate": 2e-05,
+      "loss": 0.04838118,
+      "step": 15361
+    },
+    {
+      "epoch": 30.724,
+      "grad_norm": 1.3870517015457153,
+      "learning_rate": 2e-05,
+      "loss": 0.04288274,
+      "step": 15362
+    },
+    {
+      "epoch": 30.726,
+      "grad_norm": 1.472120761871338,
+      "learning_rate": 2e-05,
+      "loss": 0.05352866,
+      "step": 15363
+    },
+    {
+      "epoch": 30.728,
+      "grad_norm": 1.9960273504257202,
+      "learning_rate": 2e-05,
+      "loss": 0.0549426,
+      "step": 15364
+    },
+    {
+      "epoch": 30.73,
+      "grad_norm": 1.4268264770507812,
+      "learning_rate": 2e-05,
+      "loss": 0.04215187,
+      "step": 15365
+    },
+    {
+      "epoch": 30.732,
+      "grad_norm": 1.359204649925232,
+      "learning_rate": 2e-05,
+      "loss": 0.03503023,
+      "step": 15366
+    },
+    {
+      "epoch": 30.734,
+      "grad_norm": 1.4836820363998413,
+      "learning_rate": 2e-05,
+      "loss": 0.04756457,
+      "step": 15367
+    },
+    {
+      "epoch": 30.736,
+      "grad_norm": 1.3566566705703735,
+      "learning_rate": 2e-05,
+      "loss": 0.04132615,
+      "step": 15368
+    },
+    {
+      "epoch": 30.738,
+      "grad_norm": 1.0889496803283691,
+      "learning_rate": 2e-05,
+      "loss": 0.03694168,
+      "step": 15369
+    },
+    {
+      "epoch": 30.74,
+      "grad_norm": 1.9702776670455933,
+      "learning_rate": 2e-05,
+      "loss": 0.04029308,
+      "step": 15370
+    },
+    {
+      "epoch": 30.742,
+      "grad_norm": 1.7063497304916382,
+      "learning_rate": 2e-05,
+      "loss": 0.04850524,
+      "step": 15371
+    },
+    {
+      "epoch": 30.744,
+      "grad_norm": 1.1768672466278076,
+      "learning_rate": 2e-05,
+      "loss": 0.04457154,
+      "step": 15372
+    },
+    {
+      "epoch": 30.746,
+      "grad_norm": 1.506422758102417,
+      "learning_rate": 2e-05,
+      "loss": 0.05803453,
+      "step": 15373
+    },
+    {
+      "epoch": 30.748,
+      "grad_norm": 1.2613164186477661,
+      "learning_rate": 2e-05,
+      "loss": 0.0555273,
+      "step": 15374
+    },
+    {
+      "epoch": 30.75,
+      "grad_norm": 1.302585482597351,
+      "learning_rate": 2e-05,
+      "loss": 0.03637126,
+      "step": 15375
+    },
+    {
+      "epoch": 30.752,
+      "grad_norm": 1.2151802778244019,
+      "learning_rate": 2e-05,
+      "loss": 0.04023366,
+      "step": 15376
+    },
+    {
+      "epoch": 30.754,
+      "grad_norm": 1.4890795946121216,
+      "learning_rate": 2e-05,
+      "loss": 0.04400902,
+      "step": 15377
+    },
+    {
+      "epoch": 30.756,
+      "grad_norm": 1.8800874948501587,
+      "learning_rate": 2e-05,
+      "loss": 0.04399022,
+      "step": 15378
+    },
+    {
+      "epoch": 30.758,
+      "grad_norm": 1.0342926979064941,
+      "learning_rate": 2e-05,
+      "loss": 0.04493139,
+      "step": 15379
+    },
+    {
+      "epoch": 30.76,
+      "grad_norm": 2.302096366882324,
+      "learning_rate": 2e-05,
+      "loss": 0.0552254,
+      "step": 15380
+    },
+    {
+      "epoch": 30.762,
+      "grad_norm": 0.9423291087150574,
+      "learning_rate": 2e-05,
+      "loss": 0.03075608,
+      "step": 15381
+    },
+    {
+      "epoch": 30.764,
+      "grad_norm": 1.7802855968475342,
+      "learning_rate": 2e-05,
+      "loss": 0.0369587,
+      "step": 15382
+    },
+    {
+      "epoch": 30.766,
+      "grad_norm": 1.1834336519241333,
+      "learning_rate": 2e-05,
+      "loss": 0.04712176,
+      "step": 15383
+    },
+    {
+      "epoch": 30.768,
+      "grad_norm": 1.2314602136611938,
+      "learning_rate": 2e-05,
+      "loss": 0.04508484,
+      "step": 15384
+    },
+    {
+      "epoch": 30.77,
+      "grad_norm": 1.5756516456604004,
+      "learning_rate": 2e-05,
+      "loss": 0.04346953,
+      "step": 15385
+    },
+    {
+      "epoch": 30.772,
+      "grad_norm": 3.2680647373199463,
+      "learning_rate": 2e-05,
+      "loss": 0.03386783,
+      "step": 15386
+    },
+    {
+      "epoch": 30.774,
+      "grad_norm": 1.0917185544967651,
+      "learning_rate": 2e-05,
+      "loss": 0.04377945,
+      "step": 15387
+    },
+    {
+      "epoch": 30.776,
+      "grad_norm": 1.5100263357162476,
+      "learning_rate": 2e-05,
+      "loss": 0.0553541,
+      "step": 15388
+    },
+    {
+      "epoch": 30.778,
+      "grad_norm": 1.24834406375885,
+      "learning_rate": 2e-05,
+      "loss": 0.03370894,
+      "step": 15389
+    },
+    {
+      "epoch": 30.78,
+      "grad_norm": 1.9447808265686035,
+      "learning_rate": 2e-05,
+      "loss": 0.05794008,
+      "step": 15390
+    },
+    {
+      "epoch": 30.782,
+      "grad_norm": 1.1977484226226807,
+      "learning_rate": 2e-05,
+      "loss": 0.04914448,
+      "step": 15391
+    },
+    {
+      "epoch": 30.784,
+      "grad_norm": 2.5817272663116455,
+      "learning_rate": 2e-05,
+      "loss": 0.04607561,
+      "step": 15392
+    },
+    {
+      "epoch": 30.786,
+      "grad_norm": 0.9399213790893555,
+      "learning_rate": 2e-05,
+      "loss": 0.03578041,
+      "step": 15393
+    },
+    {
+      "epoch": 30.788,
+      "grad_norm": 1.2921048402786255,
+      "learning_rate": 2e-05,
+      "loss": 0.03382245,
+      "step": 15394
+    },
+    {
+      "epoch": 30.79,
+      "grad_norm": 1.1487122774124146,
+      "learning_rate": 2e-05,
+      "loss": 0.04007035,
+      "step": 15395
+    },
+    {
+      "epoch": 30.792,
+      "grad_norm": 1.0508596897125244,
+      "learning_rate": 2e-05,
+      "loss": 0.03421424,
+      "step": 15396
+    },
+    {
+      "epoch": 30.794,
+      "grad_norm": 0.864069938659668,
+      "learning_rate": 2e-05,
+      "loss": 0.02603904,
+      "step": 15397
+    },
+    {
+      "epoch": 30.796,
+      "grad_norm": 1.4718056917190552,
+      "learning_rate": 2e-05,
+      "loss": 0.038355,
+      "step": 15398
+    },
+    {
+      "epoch": 30.798000000000002,
+      "grad_norm": 1.2328131198883057,
+      "learning_rate": 2e-05,
+      "loss": 0.04491291,
+      "step": 15399
+    },
+    {
+      "epoch": 30.8,
+      "grad_norm": 1.186107873916626,
+      "learning_rate": 2e-05,
+      "loss": 0.05053133,
+      "step": 15400
+    },
+    {
+      "epoch": 30.802,
+      "grad_norm": 1.0685789585113525,
+      "learning_rate": 2e-05,
+      "loss": 0.04427924,
+      "step": 15401
+    },
+    {
+      "epoch": 30.804,
+      "grad_norm": 1.1698602437973022,
+      "learning_rate": 2e-05,
+      "loss": 0.03796928,
+      "step": 15402
+    },
+    {
+      "epoch": 30.806,
+      "grad_norm": 1.2792614698410034,
+      "learning_rate": 2e-05,
+      "loss": 0.0548584,
+      "step": 15403
+    },
+    {
+      "epoch": 30.808,
+      "grad_norm": 1.533371090888977,
+      "learning_rate": 2e-05,
+      "loss": 0.04547575,
+      "step": 15404
+    },
+    {
+      "epoch": 30.81,
+      "grad_norm": 1.5394388437271118,
+      "learning_rate": 2e-05,
+      "loss": 0.05188145,
+      "step": 15405
+    },
+    {
+      "epoch": 30.812,
+      "grad_norm": 1.0687896013259888,
+      "learning_rate": 2e-05,
+      "loss": 0.04358595,
+      "step": 15406
+    },
+    {
+      "epoch": 30.814,
+      "grad_norm": 1.0643762350082397,
+      "learning_rate": 2e-05,
+      "loss": 0.03647963,
+      "step": 15407
+    },
+    {
+      "epoch": 30.816,
+      "grad_norm": 1.1489731073379517,
+      "learning_rate": 2e-05,
+      "loss": 0.05470124,
+      "step": 15408
+    },
+    {
+      "epoch": 30.818,
+      "grad_norm": 1.8806666135787964,
+      "learning_rate": 2e-05,
+      "loss": 0.05590469,
+      "step": 15409
+    },
+    {
+      "epoch": 30.82,
+      "grad_norm": 1.0280253887176514,
+      "learning_rate": 2e-05,
+      "loss": 0.04052454,
+      "step": 15410
+    },
+    {
+      "epoch": 30.822,
+      "grad_norm": 1.6329935789108276,
+      "learning_rate": 2e-05,
+      "loss": 0.04793561,
+      "step": 15411
+    },
+    {
+      "epoch": 30.824,
+      "grad_norm": 1.4511553049087524,
+      "learning_rate": 2e-05,
+      "loss": 0.03960864,
+      "step": 15412
+    },
+    {
+      "epoch": 30.826,
+      "grad_norm": 1.6550666093826294,
+      "learning_rate": 2e-05,
+      "loss": 0.04335929,
+      "step": 15413
+    },
+    {
+      "epoch": 30.828,
+      "grad_norm": 1.3845441341400146,
+      "learning_rate": 2e-05,
+      "loss": 0.03868318,
+      "step": 15414
+    },
+    {
+      "epoch": 30.83,
+      "grad_norm": 1.1391074657440186,
+      "learning_rate": 2e-05,
+      "loss": 0.0478869,
+      "step": 15415
+    },
+    {
+      "epoch": 30.832,
+      "grad_norm": 1.278219223022461,
+      "learning_rate": 2e-05,
+      "loss": 0.04307889,
+      "step": 15416
+    },
+    {
+      "epoch": 30.834,
+      "grad_norm": 1.268723726272583,
+      "learning_rate": 2e-05,
+      "loss": 0.03802036,
+      "step": 15417
+    },
+    {
+      "epoch": 30.836,
+      "grad_norm": 1.159140706062317,
+      "learning_rate": 2e-05,
+      "loss": 0.03451194,
+      "step": 15418
+    },
+    {
+      "epoch": 30.838,
+      "grad_norm": 1.1894630193710327,
+      "learning_rate": 2e-05,
+      "loss": 0.0473679,
+      "step": 15419
+    },
+    {
+      "epoch": 30.84,
+      "grad_norm": 1.262392520904541,
+      "learning_rate": 2e-05,
+      "loss": 0.03899877,
+      "step": 15420
+    },
+    {
+      "epoch": 30.842,
+      "grad_norm": 1.446376085281372,
+      "learning_rate": 2e-05,
+      "loss": 0.03620021,
+      "step": 15421
+    },
+    {
+      "epoch": 30.844,
+      "grad_norm": 1.4437825679779053,
+      "learning_rate": 2e-05,
+      "loss": 0.04326116,
+      "step": 15422
+    },
+    {
+      "epoch": 30.846,
+      "grad_norm": 1.1866358518600464,
+      "learning_rate": 2e-05,
+      "loss": 0.03965632,
+      "step": 15423
+    },
+    {
+      "epoch": 30.848,
+      "grad_norm": 1.379371166229248,
+      "learning_rate": 2e-05,
+      "loss": 0.04506073,
+      "step": 15424
+    },
+    {
+      "epoch": 30.85,
+      "grad_norm": 2.1914074420928955,
+      "learning_rate": 2e-05,
+      "loss": 0.05980007,
+      "step": 15425
+    },
+    {
+      "epoch": 30.852,
+      "grad_norm": 1.2148512601852417,
+      "learning_rate": 2e-05,
+      "loss": 0.04434382,
+      "step": 15426
+    },
+    {
+      "epoch": 30.854,
+      "grad_norm": 1.0785166025161743,
+      "learning_rate": 2e-05,
+      "loss": 0.03507778,
+      "step": 15427
+    },
+    {
+      "epoch": 30.856,
+      "grad_norm": 1.0878171920776367,
+      "learning_rate": 2e-05,
+      "loss": 0.03332923,
+      "step": 15428
+    },
+    {
+      "epoch": 30.858,
+      "grad_norm": 1.4415518045425415,
+      "learning_rate": 2e-05,
+      "loss": 0.05356176,
+      "step": 15429
+    },
+    {
+      "epoch": 30.86,
+      "grad_norm": 1.4083247184753418,
+      "learning_rate": 2e-05,
+      "loss": 0.05534295,
+      "step": 15430
+    },
+    {
+      "epoch": 30.862,
+      "grad_norm": 1.1404030323028564,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 15431
+    },
+    {
+      "epoch": 30.864,
+      "grad_norm": 1.4626610279083252,
+      "learning_rate": 2e-05,
+      "loss": 0.03837757,
+      "step": 15432
+    },
+    {
+      "epoch": 30.866,
+      "grad_norm": 1.0799211263656616,
+      "learning_rate": 2e-05,
+      "loss": 0.04735259,
+      "step": 15433
+    },
+    {
+      "epoch": 30.868,
+      "grad_norm": 1.3441239595413208,
+      "learning_rate": 2e-05,
+      "loss": 0.0439587,
+      "step": 15434
+    },
+    {
+      "epoch": 30.87,
+      "grad_norm": 1.341903567314148,
+      "learning_rate": 2e-05,
+      "loss": 0.04261306,
+      "step": 15435
+    },
+    {
+      "epoch": 30.872,
+      "grad_norm": 2.014629602432251,
+      "learning_rate": 2e-05,
+      "loss": 0.04418968,
+      "step": 15436
+    },
+    {
+      "epoch": 30.874,
+      "grad_norm": 1.4627952575683594,
+      "learning_rate": 2e-05,
+      "loss": 0.04787583,
+      "step": 15437
+    },
+    {
+      "epoch": 30.876,
+      "grad_norm": 1.0740972757339478,
+      "learning_rate": 2e-05,
+      "loss": 0.03408062,
+      "step": 15438
+    },
+    {
+      "epoch": 30.878,
+      "grad_norm": 1.458024024963379,
+      "learning_rate": 2e-05,
+      "loss": 0.05833004,
+      "step": 15439
+    },
+    {
+      "epoch": 30.88,
+      "grad_norm": 1.0955424308776855,
+      "learning_rate": 2e-05,
+      "loss": 0.04713885,
+      "step": 15440
+    },
+    {
+      "epoch": 30.882,
+      "grad_norm": 1.673150658607483,
+      "learning_rate": 2e-05,
+      "loss": 0.05364336,
+      "step": 15441
+    },
+    {
+      "epoch": 30.884,
+      "grad_norm": 1.1266365051269531,
+      "learning_rate": 2e-05,
+      "loss": 0.04248474,
+      "step": 15442
+    },
+    {
+      "epoch": 30.886,
+      "grad_norm": 1.5739974975585938,
+      "learning_rate": 2e-05,
+      "loss": 0.03701893,
+      "step": 15443
+    },
+    {
+      "epoch": 30.888,
+      "grad_norm": 1.4596102237701416,
+      "learning_rate": 2e-05,
+      "loss": 0.047738,
+      "step": 15444
+    },
+    {
+      "epoch": 30.89,
+      "grad_norm": 1.0531946420669556,
+      "learning_rate": 2e-05,
+      "loss": 0.04053745,
+      "step": 15445
+    },
+    {
+      "epoch": 30.892,
+      "grad_norm": 0.9864213466644287,
+      "learning_rate": 2e-05,
+      "loss": 0.03761032,
+      "step": 15446
+    },
+    {
+      "epoch": 30.894,
+      "grad_norm": 2.1464390754699707,
+      "learning_rate": 2e-05,
+      "loss": 0.05027218,
+      "step": 15447
+    },
+    {
+      "epoch": 30.896,
+      "grad_norm": 1.2808313369750977,
+      "learning_rate": 2e-05,
+      "loss": 0.04857701,
+      "step": 15448
+    },
+    {
+      "epoch": 30.898,
+      "grad_norm": 1.7590348720550537,
+      "learning_rate": 2e-05,
+      "loss": 0.05534477,
+      "step": 15449
+    },
+    {
+      "epoch": 30.9,
+      "grad_norm": 1.0882501602172852,
+      "learning_rate": 2e-05,
+      "loss": 0.04137859,
+      "step": 15450
+    },
+    {
+      "epoch": 30.902,
+      "grad_norm": 0.946066677570343,
+      "learning_rate": 2e-05,
+      "loss": 0.03223543,
+      "step": 15451
+    },
+    {
+      "epoch": 30.904,
+      "grad_norm": 0.9430605173110962,
+      "learning_rate": 2e-05,
+      "loss": 0.0311843,
+      "step": 15452
+    },
+    {
+      "epoch": 30.906,
+      "grad_norm": 1.3041696548461914,
+      "learning_rate": 2e-05,
+      "loss": 0.05397451,
+      "step": 15453
+    },
+    {
+      "epoch": 30.908,
+      "grad_norm": 1.3684122562408447,
+      "learning_rate": 2e-05,
+      "loss": 0.05539594,
+      "step": 15454
+    },
+    {
+      "epoch": 30.91,
+      "grad_norm": 2.258601665496826,
+      "learning_rate": 2e-05,
+      "loss": 0.05372835,
+      "step": 15455
+    },
+    {
+      "epoch": 30.912,
+      "grad_norm": 2.0608880519866943,
+      "learning_rate": 2e-05,
+      "loss": 0.03330045,
+      "step": 15456
+    },
+    {
+      "epoch": 30.914,
+      "grad_norm": 1.1213666200637817,
+      "learning_rate": 2e-05,
+      "loss": 0.04753671,
+      "step": 15457
+    },
+    {
+      "epoch": 30.916,
+      "grad_norm": 0.942790150642395,
+      "learning_rate": 2e-05,
+      "loss": 0.03259289,
+      "step": 15458
+    },
+    {
+      "epoch": 30.918,
+      "grad_norm": 1.2493982315063477,
+      "learning_rate": 2e-05,
+      "loss": 0.04098676,
+      "step": 15459
+    },
+    {
+      "epoch": 30.92,
+      "grad_norm": 1.1307854652404785,
+      "learning_rate": 2e-05,
+      "loss": 0.04249147,
+      "step": 15460
+    },
+    {
+      "epoch": 30.922,
+      "grad_norm": 1.5322867631912231,
+      "learning_rate": 2e-05,
+      "loss": 0.03617536,
+      "step": 15461
+    },
+    {
+      "epoch": 30.924,
+      "grad_norm": 1.1813342571258545,
+      "learning_rate": 2e-05,
+      "loss": 0.05639866,
+      "step": 15462
+    },
+    {
+      "epoch": 30.926,
+      "grad_norm": 0.8674895763397217,
+      "learning_rate": 2e-05,
+      "loss": 0.03328463,
+      "step": 15463
+    },
+    {
+      "epoch": 30.928,
+      "grad_norm": 1.5556342601776123,
+      "learning_rate": 2e-05,
+      "loss": 0.05676714,
+      "step": 15464
+    },
+    {
+      "epoch": 30.93,
+      "grad_norm": 1.0045814514160156,
+      "learning_rate": 2e-05,
+      "loss": 0.03216821,
+      "step": 15465
+    },
+    {
+      "epoch": 30.932,
+      "grad_norm": 1.0363186597824097,
+      "learning_rate": 2e-05,
+      "loss": 0.04014078,
+      "step": 15466
+    },
+    {
+      "epoch": 30.934,
+      "grad_norm": 1.386189341545105,
+      "learning_rate": 2e-05,
+      "loss": 0.05402887,
+      "step": 15467
+    },
+    {
+      "epoch": 30.936,
+      "grad_norm": 1.137304663658142,
+      "learning_rate": 2e-05,
+      "loss": 0.03464006,
+      "step": 15468
+    },
+    {
+      "epoch": 30.938,
+      "grad_norm": 1.3049153089523315,
+      "learning_rate": 2e-05,
+      "loss": 0.04342077,
+      "step": 15469
+    },
+    {
+      "epoch": 30.94,
+      "grad_norm": 1.1227730512619019,
+      "learning_rate": 2e-05,
+      "loss": 0.04238366,
+      "step": 15470
+    },
+    {
+      "epoch": 30.942,
+      "grad_norm": 1.2162871360778809,
+      "learning_rate": 2e-05,
+      "loss": 0.06160378,
+      "step": 15471
+    },
+    {
+      "epoch": 30.944,
+      "grad_norm": 0.8703532218933105,
+      "learning_rate": 2e-05,
+      "loss": 0.02413613,
+      "step": 15472
+    },
+    {
+      "epoch": 30.946,
+      "grad_norm": 1.9870779514312744,
+      "learning_rate": 2e-05,
+      "loss": 0.03381436,
+      "step": 15473
+    },
+    {
+      "epoch": 30.948,
+      "grad_norm": 1.9186214208602905,
+      "learning_rate": 2e-05,
+      "loss": 0.05125932,
+      "step": 15474
+    },
+    {
+      "epoch": 30.95,
+      "grad_norm": 1.7328497171401978,
+      "learning_rate": 2e-05,
+      "loss": 0.0443011,
+      "step": 15475
+    },
+    {
+      "epoch": 30.951999999999998,
+      "grad_norm": 0.9649899005889893,
+      "learning_rate": 2e-05,
+      "loss": 0.03079892,
+      "step": 15476
+    },
+    {
+      "epoch": 30.954,
+      "grad_norm": 0.9868589043617249,
+      "learning_rate": 2e-05,
+      "loss": 0.02492999,
+      "step": 15477
+    },
+    {
+      "epoch": 30.956,
+      "grad_norm": 1.2447783946990967,
+      "learning_rate": 2e-05,
+      "loss": 0.05273892,
+      "step": 15478
+    },
+    {
+      "epoch": 30.958,
+      "grad_norm": 1.2232327461242676,
+      "learning_rate": 2e-05,
+      "loss": 0.04913422,
+      "step": 15479
+    },
+    {
+      "epoch": 30.96,
+      "grad_norm": 1.6843918561935425,
+      "learning_rate": 2e-05,
+      "loss": 0.05371039,
+      "step": 15480
+    },
+    {
+      "epoch": 30.962,
+      "grad_norm": 0.8725624680519104,
+      "learning_rate": 2e-05,
+      "loss": 0.02640246,
+      "step": 15481
+    },
+    {
+      "epoch": 30.964,
+      "grad_norm": 1.2975506782531738,
+      "learning_rate": 2e-05,
+      "loss": 0.05456071,
+      "step": 15482
+    },
+    {
+      "epoch": 30.966,
+      "grad_norm": 1.7329095602035522,
+      "learning_rate": 2e-05,
+      "loss": 0.04484017,
+      "step": 15483
+    },
+    {
+      "epoch": 30.968,
+      "grad_norm": 1.4026141166687012,
+      "learning_rate": 2e-05,
+      "loss": 0.05021109,
+      "step": 15484
+    },
+    {
+      "epoch": 30.97,
+      "grad_norm": 1.046205997467041,
+      "learning_rate": 2e-05,
+      "loss": 0.03913869,
+      "step": 15485
+    },
+    {
+      "epoch": 30.972,
+      "grad_norm": 3.3396363258361816,
+      "learning_rate": 2e-05,
+      "loss": 0.04359189,
+      "step": 15486
+    },
+    {
+      "epoch": 30.974,
+      "grad_norm": 1.369861125946045,
+      "learning_rate": 2e-05,
+      "loss": 0.04546359,
+      "step": 15487
+    },
+    {
+      "epoch": 30.976,
+      "grad_norm": 1.4066284894943237,
+      "learning_rate": 2e-05,
+      "loss": 0.03458145,
+      "step": 15488
+    },
+    {
+      "epoch": 30.978,
+      "grad_norm": 1.479104995727539,
+      "learning_rate": 2e-05,
+      "loss": 0.04172707,
+      "step": 15489
+    },
+    {
+      "epoch": 30.98,
+      "grad_norm": 1.0808502435684204,
+      "learning_rate": 2e-05,
+      "loss": 0.0374383,
+      "step": 15490
+    },
+    {
+      "epoch": 30.982,
+      "grad_norm": 1.153609275817871,
+      "learning_rate": 2e-05,
+      "loss": 0.04683513,
+      "step": 15491
+    },
+    {
+      "epoch": 30.984,
+      "grad_norm": 1.3860514163970947,
+      "learning_rate": 2e-05,
+      "loss": 0.04752363,
+      "step": 15492
+    },
+    {
+      "epoch": 30.986,
+      "grad_norm": 1.005586862564087,
+      "learning_rate": 2e-05,
+      "loss": 0.03461302,
+      "step": 15493
+    },
+    {
+      "epoch": 30.988,
+      "grad_norm": 2.63100266456604,
+      "learning_rate": 2e-05,
+      "loss": 0.05241647,
+      "step": 15494
+    },
+    {
+      "epoch": 30.99,
+      "grad_norm": 1.4951649904251099,
+      "learning_rate": 2e-05,
+      "loss": 0.03878006,
+      "step": 15495
+    },
+    {
+      "epoch": 30.992,
+      "grad_norm": 1.208921194076538,
+      "learning_rate": 2e-05,
+      "loss": 0.04567692,
+      "step": 15496
+    },
+    {
+      "epoch": 30.994,
+      "grad_norm": 1.2812520265579224,
+      "learning_rate": 2e-05,
+      "loss": 0.03168442,
+      "step": 15497
+    },
+    {
+      "epoch": 30.996,
+      "grad_norm": 1.3722220659255981,
+      "learning_rate": 2e-05,
+      "loss": 0.04763746,
+      "step": 15498
+    },
+    {
+      "epoch": 30.998,
+      "grad_norm": 1.4294661283493042,
+      "learning_rate": 2e-05,
+      "loss": 0.05221847,
+      "step": 15499
+    },
+    {
+      "epoch": 31.0,
+      "grad_norm": 1.2685956954956055,
+      "learning_rate": 2e-05,
+      "loss": 0.05318395,
+      "step": 15500
+    },
+    {
+      "epoch": 31.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.994,
+        "AngleClassification_2": 0.998,
+        "AngleClassification_3": 0.9780439121756487,
+        "Equal_1": 1.0,
+        "Equal_2": 0.9880239520958084,
+        "Equal_3": 0.9860279441117764,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.998003992015968,
+        "Parallel_1": 0.9899799599198397,
+        "Parallel_2": 0.9919839679358717,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.992,
+        "Perpendicular_3": 0.8957915831663327,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 0.9976666666666667,
+        "PointLiesOnCircle_3": 0.9936,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9880239520958084
+      },
+      "eval_runtime": 319.7306,
+      "eval_samples_per_second": 32.84,
+      "eval_steps_per_second": 0.657,
+      "step": 15500
+    },
+    {
+      "epoch": 31.002,
+      "grad_norm": 1.2148770093917847,
+      "learning_rate": 2e-05,
+      "loss": 0.04367323,
+      "step": 15501
+    },
+    {
+      "epoch": 31.004,
+      "grad_norm": 1.2555484771728516,
+      "learning_rate": 2e-05,
+      "loss": 0.04165998,
+      "step": 15502
+    },
+    {
+      "epoch": 31.006,
+      "grad_norm": 1.7054599523544312,
+      "learning_rate": 2e-05,
+      "loss": 0.04306837,
+      "step": 15503
+    },
+    {
+      "epoch": 31.008,
+      "grad_norm": 1.1247222423553467,
+      "learning_rate": 2e-05,
+      "loss": 0.04493858,
+      "step": 15504
+    },
+    {
+      "epoch": 31.01,
+      "grad_norm": 1.21216881275177,
+      "learning_rate": 2e-05,
+      "loss": 0.04912875,
+      "step": 15505
+    },
+    {
+      "epoch": 31.012,
+      "grad_norm": 1.0440160036087036,
+      "learning_rate": 2e-05,
+      "loss": 0.03388001,
+      "step": 15506
+    },
+    {
+      "epoch": 31.014,
+      "grad_norm": 2.1243808269500732,
+      "learning_rate": 2e-05,
+      "loss": 0.04466413,
+      "step": 15507
+    },
+    {
+      "epoch": 31.016,
+      "grad_norm": 1.024837851524353,
+      "learning_rate": 2e-05,
+      "loss": 0.03612889,
+      "step": 15508
+    },
+    {
+      "epoch": 31.018,
+      "grad_norm": 1.3261427879333496,
+      "learning_rate": 2e-05,
+      "loss": 0.04339204,
+      "step": 15509
+    },
+    {
+      "epoch": 31.02,
+      "grad_norm": 1.1927242279052734,
+      "learning_rate": 2e-05,
+      "loss": 0.04285469,
+      "step": 15510
+    },
+    {
+      "epoch": 31.022,
+      "grad_norm": 1.6944687366485596,
+      "learning_rate": 2e-05,
+      "loss": 0.06444372,
+      "step": 15511
+    },
+    {
+      "epoch": 31.024,
+      "grad_norm": 1.8218998908996582,
+      "learning_rate": 2e-05,
+      "loss": 0.04665653,
+      "step": 15512
+    },
+    {
+      "epoch": 31.026,
+      "grad_norm": 0.961631715297699,
+      "learning_rate": 2e-05,
+      "loss": 0.02488809,
+      "step": 15513
+    },
+    {
+      "epoch": 31.028,
+      "grad_norm": 1.3068662881851196,
+      "learning_rate": 2e-05,
+      "loss": 0.04824544,
+      "step": 15514
+    },
+    {
+      "epoch": 31.03,
+      "grad_norm": 1.7906019687652588,
+      "learning_rate": 2e-05,
+      "loss": 0.03604405,
+      "step": 15515
+    },
+    {
+      "epoch": 31.032,
+      "grad_norm": 0.9942117929458618,
+      "learning_rate": 2e-05,
+      "loss": 0.04298211,
+      "step": 15516
+    },
+    {
+      "epoch": 31.034,
+      "grad_norm": 1.2944165468215942,
+      "learning_rate": 2e-05,
+      "loss": 0.04983748,
+      "step": 15517
+    },
+    {
+      "epoch": 31.036,
+      "grad_norm": 1.3310195207595825,
+      "learning_rate": 2e-05,
+      "loss": 0.04212454,
+      "step": 15518
+    },
+    {
+      "epoch": 31.038,
+      "grad_norm": 1.0146511793136597,
+      "learning_rate": 2e-05,
+      "loss": 0.03470632,
+      "step": 15519
+    },
+    {
+      "epoch": 31.04,
+      "grad_norm": 1.0683945417404175,
+      "learning_rate": 2e-05,
+      "loss": 0.03555411,
+      "step": 15520
+    },
+    {
+      "epoch": 31.042,
+      "grad_norm": 0.9860395789146423,
+      "learning_rate": 2e-05,
+      "loss": 0.03534026,
+      "step": 15521
+    },
+    {
+      "epoch": 31.044,
+      "grad_norm": 1.2321219444274902,
+      "learning_rate": 2e-05,
+      "loss": 0.04802042,
+      "step": 15522
+    },
+    {
+      "epoch": 31.046,
+      "grad_norm": 0.9068757891654968,
+      "learning_rate": 2e-05,
+      "loss": 0.03739195,
+      "step": 15523
+    },
+    {
+      "epoch": 31.048,
+      "grad_norm": 1.0807992219924927,
+      "learning_rate": 2e-05,
+      "loss": 0.03552995,
+      "step": 15524
+    },
+    {
+      "epoch": 31.05,
+      "grad_norm": 1.2086913585662842,
+      "learning_rate": 2e-05,
+      "loss": 0.03402312,
+      "step": 15525
+    },
+    {
+      "epoch": 31.052,
+      "grad_norm": 1.1114925146102905,
+      "learning_rate": 2e-05,
+      "loss": 0.04508002,
+      "step": 15526
+    },
+    {
+      "epoch": 31.054,
+      "grad_norm": 1.053460717201233,
+      "learning_rate": 2e-05,
+      "loss": 0.03332749,
+      "step": 15527
+    },
+    {
+      "epoch": 31.056,
+      "grad_norm": 1.5291039943695068,
+      "learning_rate": 2e-05,
+      "loss": 0.07005154,
+      "step": 15528
+    },
+    {
+      "epoch": 31.058,
+      "grad_norm": 1.3969502449035645,
+      "learning_rate": 2e-05,
+      "loss": 0.03875972,
+      "step": 15529
+    },
+    {
+      "epoch": 31.06,
+      "grad_norm": 1.258709192276001,
+      "learning_rate": 2e-05,
+      "loss": 0.0648275,
+      "step": 15530
+    },
+    {
+      "epoch": 31.062,
+      "grad_norm": 2.1577484607696533,
+      "learning_rate": 2e-05,
+      "loss": 0.04727962,
+      "step": 15531
+    },
+    {
+      "epoch": 31.064,
+      "grad_norm": 1.453369379043579,
+      "learning_rate": 2e-05,
+      "loss": 0.03450019,
+      "step": 15532
+    },
+    {
+      "epoch": 31.066,
+      "grad_norm": 1.3698201179504395,
+      "learning_rate": 2e-05,
+      "loss": 0.0564566,
+      "step": 15533
+    },
+    {
+      "epoch": 31.068,
+      "grad_norm": 1.4147790670394897,
+      "learning_rate": 2e-05,
+      "loss": 0.06183496,
+      "step": 15534
+    },
+    {
+      "epoch": 31.07,
+      "grad_norm": 0.8963314294815063,
+      "learning_rate": 2e-05,
+      "loss": 0.0314716,
+      "step": 15535
+    },
+    {
+      "epoch": 31.072,
+      "grad_norm": 1.7133666276931763,
+      "learning_rate": 2e-05,
+      "loss": 0.03604105,
+      "step": 15536
+    },
+    {
+      "epoch": 31.074,
+      "grad_norm": 0.9274123907089233,
+      "learning_rate": 2e-05,
+      "loss": 0.03644059,
+      "step": 15537
+    },
+    {
+      "epoch": 31.076,
+      "grad_norm": 0.9932688474655151,
+      "learning_rate": 2e-05,
+      "loss": 0.03320706,
+      "step": 15538
+    },
+    {
+      "epoch": 31.078,
+      "grad_norm": 1.139175534248352,
+      "learning_rate": 2e-05,
+      "loss": 0.04096094,
+      "step": 15539
+    },
+    {
+      "epoch": 31.08,
+      "grad_norm": 2.054170846939087,
+      "learning_rate": 2e-05,
+      "loss": 0.03290722,
+      "step": 15540
+    },
+    {
+      "epoch": 31.082,
+      "grad_norm": 1.0771855115890503,
+      "learning_rate": 2e-05,
+      "loss": 0.04088487,
+      "step": 15541
+    },
+    {
+      "epoch": 31.084,
+      "grad_norm": 4.079372406005859,
+      "learning_rate": 2e-05,
+      "loss": 0.06007933,
+      "step": 15542
+    },
+    {
+      "epoch": 31.086,
+      "grad_norm": 4.406796932220459,
+      "learning_rate": 2e-05,
+      "loss": 0.0371773,
+      "step": 15543
+    },
+    {
+      "epoch": 31.088,
+      "grad_norm": 1.232641339302063,
+      "learning_rate": 2e-05,
+      "loss": 0.04272714,
+      "step": 15544
+    },
+    {
+      "epoch": 31.09,
+      "grad_norm": 1.339889407157898,
+      "learning_rate": 2e-05,
+      "loss": 0.03682033,
+      "step": 15545
+    },
+    {
+      "epoch": 31.092,
+      "grad_norm": 1.6685773134231567,
+      "learning_rate": 2e-05,
+      "loss": 0.04497613,
+      "step": 15546
+    },
+    {
+      "epoch": 31.094,
+      "grad_norm": 1.1808186769485474,
+      "learning_rate": 2e-05,
+      "loss": 0.04455563,
+      "step": 15547
+    },
+    {
+      "epoch": 31.096,
+      "grad_norm": 1.0447876453399658,
+      "learning_rate": 2e-05,
+      "loss": 0.03678391,
+      "step": 15548
+    },
+    {
+      "epoch": 31.098,
+      "grad_norm": 1.3275498151779175,
+      "learning_rate": 2e-05,
+      "loss": 0.05487559,
+      "step": 15549
+    },
+    {
+      "epoch": 31.1,
+      "grad_norm": 1.0034887790679932,
+      "learning_rate": 2e-05,
+      "loss": 0.03993573,
+      "step": 15550
+    },
+    {
+      "epoch": 31.102,
+      "grad_norm": 1.2984068393707275,
+      "learning_rate": 2e-05,
+      "loss": 0.03917623,
+      "step": 15551
+    },
+    {
+      "epoch": 31.104,
+      "grad_norm": 1.0606168508529663,
+      "learning_rate": 2e-05,
+      "loss": 0.03269617,
+      "step": 15552
+    },
+    {
+      "epoch": 31.106,
+      "grad_norm": 1.4763702154159546,
+      "learning_rate": 2e-05,
+      "loss": 0.05329809,
+      "step": 15553
+    },
+    {
+      "epoch": 31.108,
+      "grad_norm": 1.3631935119628906,
+      "learning_rate": 2e-05,
+      "loss": 0.03446741,
+      "step": 15554
+    },
+    {
+      "epoch": 31.11,
+      "grad_norm": 1.3570765256881714,
+      "learning_rate": 2e-05,
+      "loss": 0.04257222,
+      "step": 15555
+    },
+    {
+      "epoch": 31.112,
+      "grad_norm": 1.1822189092636108,
+      "learning_rate": 2e-05,
+      "loss": 0.04207947,
+      "step": 15556
+    },
+    {
+      "epoch": 31.114,
+      "grad_norm": 1.150335431098938,
+      "learning_rate": 2e-05,
+      "loss": 0.04551108,
+      "step": 15557
+    },
+    {
+      "epoch": 31.116,
+      "grad_norm": 1.3339632749557495,
+      "learning_rate": 2e-05,
+      "loss": 0.05508456,
+      "step": 15558
+    },
+    {
+      "epoch": 31.118,
+      "grad_norm": 2.0711183547973633,
+      "learning_rate": 2e-05,
+      "loss": 0.0530576,
+      "step": 15559
+    },
+    {
+      "epoch": 31.12,
+      "grad_norm": 1.3315399885177612,
+      "learning_rate": 2e-05,
+      "loss": 0.04907047,
+      "step": 15560
+    },
+    {
+      "epoch": 31.122,
+      "grad_norm": 1.204527497291565,
+      "learning_rate": 2e-05,
+      "loss": 0.04079658,
+      "step": 15561
+    },
+    {
+      "epoch": 31.124,
+      "grad_norm": 2.65455961227417,
+      "learning_rate": 2e-05,
+      "loss": 0.05661328,
+      "step": 15562
+    },
+    {
+      "epoch": 31.126,
+      "grad_norm": 1.3059120178222656,
+      "learning_rate": 2e-05,
+      "loss": 0.03965393,
+      "step": 15563
+    },
+    {
+      "epoch": 31.128,
+      "grad_norm": 1.079121708869934,
+      "learning_rate": 2e-05,
+      "loss": 0.04260961,
+      "step": 15564
+    },
+    {
+      "epoch": 31.13,
+      "grad_norm": 2.0246469974517822,
+      "learning_rate": 2e-05,
+      "loss": 0.03333722,
+      "step": 15565
+    },
+    {
+      "epoch": 31.132,
+      "grad_norm": 1.0588722229003906,
+      "learning_rate": 2e-05,
+      "loss": 0.04730598,
+      "step": 15566
+    },
+    {
+      "epoch": 31.134,
+      "grad_norm": 1.0964921712875366,
+      "learning_rate": 2e-05,
+      "loss": 0.0434188,
+      "step": 15567
+    },
+    {
+      "epoch": 31.136,
+      "grad_norm": 1.1378031969070435,
+      "learning_rate": 2e-05,
+      "loss": 0.04260813,
+      "step": 15568
+    },
+    {
+      "epoch": 31.138,
+      "grad_norm": 3.1608223915100098,
+      "learning_rate": 2e-05,
+      "loss": 0.0434106,
+      "step": 15569
+    },
+    {
+      "epoch": 31.14,
+      "grad_norm": 1.2680219411849976,
+      "learning_rate": 2e-05,
+      "loss": 0.04677624,
+      "step": 15570
+    },
+    {
+      "epoch": 31.142,
+      "grad_norm": 1.1608775854110718,
+      "learning_rate": 2e-05,
+      "loss": 0.04532572,
+      "step": 15571
+    },
+    {
+      "epoch": 31.144,
+      "grad_norm": 1.3475708961486816,
+      "learning_rate": 2e-05,
+      "loss": 0.05460385,
+      "step": 15572
+    },
+    {
+      "epoch": 31.146,
+      "grad_norm": 1.1137781143188477,
+      "learning_rate": 2e-05,
+      "loss": 0.04142278,
+      "step": 15573
+    },
+    {
+      "epoch": 31.148,
+      "grad_norm": 1.082829475402832,
+      "learning_rate": 2e-05,
+      "loss": 0.04784383,
+      "step": 15574
+    },
+    {
+      "epoch": 31.15,
+      "grad_norm": 1.2136229276657104,
+      "learning_rate": 2e-05,
+      "loss": 0.03027528,
+      "step": 15575
+    },
+    {
+      "epoch": 31.152,
+      "grad_norm": 1.0264850854873657,
+      "learning_rate": 2e-05,
+      "loss": 0.04279248,
+      "step": 15576
+    },
+    {
+      "epoch": 31.154,
+      "grad_norm": 1.5535433292388916,
+      "learning_rate": 2e-05,
+      "loss": 0.05268865,
+      "step": 15577
+    },
+    {
+      "epoch": 31.156,
+      "grad_norm": 1.4060524702072144,
+      "learning_rate": 2e-05,
+      "loss": 0.0343322,
+      "step": 15578
+    },
+    {
+      "epoch": 31.158,
+      "grad_norm": 0.6968619227409363,
+      "learning_rate": 2e-05,
+      "loss": 0.01679928,
+      "step": 15579
+    },
+    {
+      "epoch": 31.16,
+      "grad_norm": 1.15040123462677,
+      "learning_rate": 2e-05,
+      "loss": 0.04167127,
+      "step": 15580
+    },
+    {
+      "epoch": 31.162,
+      "grad_norm": 1.524133563041687,
+      "learning_rate": 2e-05,
+      "loss": 0.03878558,
+      "step": 15581
+    },
+    {
+      "epoch": 31.164,
+      "grad_norm": 2.028996706008911,
+      "learning_rate": 2e-05,
+      "loss": 0.05743674,
+      "step": 15582
+    },
+    {
+      "epoch": 31.166,
+      "grad_norm": 1.5730241537094116,
+      "learning_rate": 2e-05,
+      "loss": 0.04803181,
+      "step": 15583
+    },
+    {
+      "epoch": 31.168,
+      "grad_norm": 1.2136539220809937,
+      "learning_rate": 2e-05,
+      "loss": 0.03908694,
+      "step": 15584
+    },
+    {
+      "epoch": 31.17,
+      "grad_norm": 2.451467514038086,
+      "learning_rate": 2e-05,
+      "loss": 0.05021393,
+      "step": 15585
+    },
+    {
+      "epoch": 31.172,
+      "grad_norm": 0.967592179775238,
+      "learning_rate": 2e-05,
+      "loss": 0.02962617,
+      "step": 15586
+    },
+    {
+      "epoch": 31.174,
+      "grad_norm": 1.205581784248352,
+      "learning_rate": 2e-05,
+      "loss": 0.04359256,
+      "step": 15587
+    },
+    {
+      "epoch": 31.176,
+      "grad_norm": 1.8534575700759888,
+      "learning_rate": 2e-05,
+      "loss": 0.05359968,
+      "step": 15588
+    },
+    {
+      "epoch": 31.178,
+      "grad_norm": 1.4265109300613403,
+      "learning_rate": 2e-05,
+      "loss": 0.0480662,
+      "step": 15589
+    },
+    {
+      "epoch": 31.18,
+      "grad_norm": 2.661834716796875,
+      "learning_rate": 2e-05,
+      "loss": 0.05001631,
+      "step": 15590
+    },
+    {
+      "epoch": 31.182,
+      "grad_norm": 2.1135103702545166,
+      "learning_rate": 2e-05,
+      "loss": 0.03104893,
+      "step": 15591
+    },
+    {
+      "epoch": 31.184,
+      "grad_norm": 1.0201138257980347,
+      "learning_rate": 2e-05,
+      "loss": 0.03780366,
+      "step": 15592
+    },
+    {
+      "epoch": 31.186,
+      "grad_norm": 1.1135469675064087,
+      "learning_rate": 2e-05,
+      "loss": 0.03954279,
+      "step": 15593
+    },
+    {
+      "epoch": 31.188,
+      "grad_norm": 1.0150232315063477,
+      "learning_rate": 2e-05,
+      "loss": 0.0309415,
+      "step": 15594
+    },
+    {
+      "epoch": 31.19,
+      "grad_norm": 1.0710225105285645,
+      "learning_rate": 2e-05,
+      "loss": 0.03865776,
+      "step": 15595
+    },
+    {
+      "epoch": 31.192,
+      "grad_norm": 1.226999044418335,
+      "learning_rate": 2e-05,
+      "loss": 0.04560003,
+      "step": 15596
+    },
+    {
+      "epoch": 31.194,
+      "grad_norm": 1.3833179473876953,
+      "learning_rate": 2e-05,
+      "loss": 0.05527067,
+      "step": 15597
+    },
+    {
+      "epoch": 31.196,
+      "grad_norm": 1.1716560125350952,
+      "learning_rate": 2e-05,
+      "loss": 0.03517221,
+      "step": 15598
+    },
+    {
+      "epoch": 31.198,
+      "grad_norm": 1.7883951663970947,
+      "learning_rate": 2e-05,
+      "loss": 0.04532555,
+      "step": 15599
+    },
+    {
+      "epoch": 31.2,
+      "grad_norm": 1.13947594165802,
+      "learning_rate": 2e-05,
+      "loss": 0.03804951,
+      "step": 15600
+    },
+    {
+      "epoch": 31.202,
+      "grad_norm": 1.1460981369018555,
+      "learning_rate": 2e-05,
+      "loss": 0.04975809,
+      "step": 15601
+    },
+    {
+      "epoch": 31.204,
+      "grad_norm": 2.1435766220092773,
+      "learning_rate": 2e-05,
+      "loss": 0.04334519,
+      "step": 15602
+    },
+    {
+      "epoch": 31.206,
+      "grad_norm": 2.5278499126434326,
+      "learning_rate": 2e-05,
+      "loss": 0.04724212,
+      "step": 15603
+    },
+    {
+      "epoch": 31.208,
+      "grad_norm": 3.0959975719451904,
+      "learning_rate": 2e-05,
+      "loss": 0.04795504,
+      "step": 15604
+    },
+    {
+      "epoch": 31.21,
+      "grad_norm": 1.3681577444076538,
+      "learning_rate": 2e-05,
+      "loss": 0.04222552,
+      "step": 15605
+    },
+    {
+      "epoch": 31.212,
+      "grad_norm": 2.0547592639923096,
+      "learning_rate": 2e-05,
+      "loss": 0.04761481,
+      "step": 15606
+    },
+    {
+      "epoch": 31.214,
+      "grad_norm": 1.327380657196045,
+      "learning_rate": 2e-05,
+      "loss": 0.05030813,
+      "step": 15607
+    },
+    {
+      "epoch": 31.216,
+      "grad_norm": 0.9326903223991394,
+      "learning_rate": 2e-05,
+      "loss": 0.02635142,
+      "step": 15608
+    },
+    {
+      "epoch": 31.218,
+      "grad_norm": 2.4155571460723877,
+      "learning_rate": 2e-05,
+      "loss": 0.04080499,
+      "step": 15609
+    },
+    {
+      "epoch": 31.22,
+      "grad_norm": 0.9260361194610596,
+      "learning_rate": 2e-05,
+      "loss": 0.03045768,
+      "step": 15610
+    },
+    {
+      "epoch": 31.222,
+      "grad_norm": 1.0131243467330933,
+      "learning_rate": 2e-05,
+      "loss": 0.03150607,
+      "step": 15611
+    },
+    {
+      "epoch": 31.224,
+      "grad_norm": 1.1739399433135986,
+      "learning_rate": 2e-05,
+      "loss": 0.03871925,
+      "step": 15612
+    },
+    {
+      "epoch": 31.226,
+      "grad_norm": 1.072050929069519,
+      "learning_rate": 2e-05,
+      "loss": 0.04814139,
+      "step": 15613
+    },
+    {
+      "epoch": 31.228,
+      "grad_norm": 2.805246591567993,
+      "learning_rate": 2e-05,
+      "loss": 0.0489997,
+      "step": 15614
+    },
+    {
+      "epoch": 31.23,
+      "grad_norm": 2.114262580871582,
+      "learning_rate": 2e-05,
+      "loss": 0.06214041,
+      "step": 15615
+    },
+    {
+      "epoch": 31.232,
+      "grad_norm": 1.275655746459961,
+      "learning_rate": 2e-05,
+      "loss": 0.03479639,
+      "step": 15616
+    },
+    {
+      "epoch": 31.234,
+      "grad_norm": 1.3052101135253906,
+      "learning_rate": 2e-05,
+      "loss": 0.04745651,
+      "step": 15617
+    },
+    {
+      "epoch": 31.236,
+      "grad_norm": 2.008206605911255,
+      "learning_rate": 2e-05,
+      "loss": 0.02935056,
+      "step": 15618
+    },
+    {
+      "epoch": 31.238,
+      "grad_norm": 1.4074678421020508,
+      "learning_rate": 2e-05,
+      "loss": 0.05427198,
+      "step": 15619
+    },
+    {
+      "epoch": 31.24,
+      "grad_norm": 1.0969489812850952,
+      "learning_rate": 2e-05,
+      "loss": 0.04732572,
+      "step": 15620
+    },
+    {
+      "epoch": 31.242,
+      "grad_norm": 1.1318331956863403,
+      "learning_rate": 2e-05,
+      "loss": 0.04238062,
+      "step": 15621
+    },
+    {
+      "epoch": 31.244,
+      "grad_norm": 1.1431554555892944,
+      "learning_rate": 2e-05,
+      "loss": 0.05024539,
+      "step": 15622
+    },
+    {
+      "epoch": 31.246,
+      "grad_norm": 1.2575552463531494,
+      "learning_rate": 2e-05,
+      "loss": 0.03069566,
+      "step": 15623
+    },
+    {
+      "epoch": 31.248,
+      "grad_norm": 1.2473465204238892,
+      "learning_rate": 2e-05,
+      "loss": 0.04475783,
+      "step": 15624
+    },
+    {
+      "epoch": 31.25,
+      "grad_norm": 1.2875679731369019,
+      "learning_rate": 2e-05,
+      "loss": 0.03152344,
+      "step": 15625
+    },
+    {
+      "epoch": 31.252,
+      "grad_norm": 1.624121069908142,
+      "learning_rate": 2e-05,
+      "loss": 0.06142828,
+      "step": 15626
+    },
+    {
+      "epoch": 31.254,
+      "grad_norm": 1.1332637071609497,
+      "learning_rate": 2e-05,
+      "loss": 0.03317603,
+      "step": 15627
+    },
+    {
+      "epoch": 31.256,
+      "grad_norm": 1.1676865816116333,
+      "learning_rate": 2e-05,
+      "loss": 0.04019779,
+      "step": 15628
+    },
+    {
+      "epoch": 31.258,
+      "grad_norm": 2.188838005065918,
+      "learning_rate": 2e-05,
+      "loss": 0.03890157,
+      "step": 15629
+    },
+    {
+      "epoch": 31.26,
+      "grad_norm": 1.019037127494812,
+      "learning_rate": 2e-05,
+      "loss": 0.03228922,
+      "step": 15630
+    },
+    {
+      "epoch": 31.262,
+      "grad_norm": 1.123388409614563,
+      "learning_rate": 2e-05,
+      "loss": 0.03372766,
+      "step": 15631
+    },
+    {
+      "epoch": 31.264,
+      "grad_norm": 3.509737491607666,
+      "learning_rate": 2e-05,
+      "loss": 0.04830921,
+      "step": 15632
+    },
+    {
+      "epoch": 31.266,
+      "grad_norm": 1.3756506443023682,
+      "learning_rate": 2e-05,
+      "loss": 0.0438694,
+      "step": 15633
+    },
+    {
+      "epoch": 31.268,
+      "grad_norm": 1.2744717597961426,
+      "learning_rate": 2e-05,
+      "loss": 0.03653997,
+      "step": 15634
+    },
+    {
+      "epoch": 31.27,
+      "grad_norm": 1.5209195613861084,
+      "learning_rate": 2e-05,
+      "loss": 0.04338511,
+      "step": 15635
+    },
+    {
+      "epoch": 31.272,
+      "grad_norm": 2.2282023429870605,
+      "learning_rate": 2e-05,
+      "loss": 0.04440949,
+      "step": 15636
+    },
+    {
+      "epoch": 31.274,
+      "grad_norm": 1.1879875659942627,
+      "learning_rate": 2e-05,
+      "loss": 0.02991764,
+      "step": 15637
+    },
+    {
+      "epoch": 31.276,
+      "grad_norm": 0.8613758683204651,
+      "learning_rate": 2e-05,
+      "loss": 0.02830381,
+      "step": 15638
+    },
+    {
+      "epoch": 31.278,
+      "grad_norm": 1.0136206150054932,
+      "learning_rate": 2e-05,
+      "loss": 0.03036819,
+      "step": 15639
+    },
+    {
+      "epoch": 31.28,
+      "grad_norm": 1.2615082263946533,
+      "learning_rate": 2e-05,
+      "loss": 0.04580786,
+      "step": 15640
+    },
+    {
+      "epoch": 31.282,
+      "grad_norm": 1.1112679243087769,
+      "learning_rate": 2e-05,
+      "loss": 0.04442069,
+      "step": 15641
+    },
+    {
+      "epoch": 31.284,
+      "grad_norm": 1.0508791208267212,
+      "learning_rate": 2e-05,
+      "loss": 0.04491913,
+      "step": 15642
+    },
+    {
+      "epoch": 31.286,
+      "grad_norm": 1.1440688371658325,
+      "learning_rate": 2e-05,
+      "loss": 0.04976808,
+      "step": 15643
+    },
+    {
+      "epoch": 31.288,
+      "grad_norm": 1.3392294645309448,
+      "learning_rate": 2e-05,
+      "loss": 0.04487741,
+      "step": 15644
+    },
+    {
+      "epoch": 31.29,
+      "grad_norm": 1.3118019104003906,
+      "learning_rate": 2e-05,
+      "loss": 0.04272094,
+      "step": 15645
+    },
+    {
+      "epoch": 31.292,
+      "grad_norm": 1.3110008239746094,
+      "learning_rate": 2e-05,
+      "loss": 0.04961982,
+      "step": 15646
+    },
+    {
+      "epoch": 31.294,
+      "grad_norm": 1.3569951057434082,
+      "learning_rate": 2e-05,
+      "loss": 0.05010619,
+      "step": 15647
+    },
+    {
+      "epoch": 31.296,
+      "grad_norm": 1.1815541982650757,
+      "learning_rate": 2e-05,
+      "loss": 0.03432131,
+      "step": 15648
+    },
+    {
+      "epoch": 31.298,
+      "grad_norm": 1.3439204692840576,
+      "learning_rate": 2e-05,
+      "loss": 0.06246966,
+      "step": 15649
+    },
+    {
+      "epoch": 31.3,
+      "grad_norm": 1.0219476222991943,
+      "learning_rate": 2e-05,
+      "loss": 0.02712579,
+      "step": 15650
+    },
+    {
+      "epoch": 31.302,
+      "grad_norm": 0.9462128281593323,
+      "learning_rate": 2e-05,
+      "loss": 0.03200362,
+      "step": 15651
+    },
+    {
+      "epoch": 31.304,
+      "grad_norm": 1.5896122455596924,
+      "learning_rate": 2e-05,
+      "loss": 0.05370201,
+      "step": 15652
+    },
+    {
+      "epoch": 31.306,
+      "grad_norm": 2.8889029026031494,
+      "learning_rate": 2e-05,
+      "loss": 0.03914107,
+      "step": 15653
+    },
+    {
+      "epoch": 31.308,
+      "grad_norm": 1.5959099531173706,
+      "learning_rate": 2e-05,
+      "loss": 0.05540708,
+      "step": 15654
+    },
+    {
+      "epoch": 31.31,
+      "grad_norm": 1.9818872213363647,
+      "learning_rate": 2e-05,
+      "loss": 0.05058894,
+      "step": 15655
+    },
+    {
+      "epoch": 31.312,
+      "grad_norm": 1.4036006927490234,
+      "learning_rate": 2e-05,
+      "loss": 0.05569137,
+      "step": 15656
+    },
+    {
+      "epoch": 31.314,
+      "grad_norm": 1.2897917032241821,
+      "learning_rate": 2e-05,
+      "loss": 0.04362039,
+      "step": 15657
+    },
+    {
+      "epoch": 31.316,
+      "grad_norm": 1.1039726734161377,
+      "learning_rate": 2e-05,
+      "loss": 0.03560334,
+      "step": 15658
+    },
+    {
+      "epoch": 31.318,
+      "grad_norm": 1.2764880657196045,
+      "learning_rate": 2e-05,
+      "loss": 0.04904519,
+      "step": 15659
+    },
+    {
+      "epoch": 31.32,
+      "grad_norm": 2.2581369876861572,
+      "learning_rate": 2e-05,
+      "loss": 0.06880252,
+      "step": 15660
+    },
+    {
+      "epoch": 31.322,
+      "grad_norm": 1.424851655960083,
+      "learning_rate": 2e-05,
+      "loss": 0.03886406,
+      "step": 15661
+    },
+    {
+      "epoch": 31.324,
+      "grad_norm": 0.9566469192504883,
+      "learning_rate": 2e-05,
+      "loss": 0.03768406,
+      "step": 15662
+    },
+    {
+      "epoch": 31.326,
+      "grad_norm": 1.1624170541763306,
+      "learning_rate": 2e-05,
+      "loss": 0.04413099,
+      "step": 15663
+    },
+    {
+      "epoch": 31.328,
+      "grad_norm": 1.2362680435180664,
+      "learning_rate": 2e-05,
+      "loss": 0.03142476,
+      "step": 15664
+    },
+    {
+      "epoch": 31.33,
+      "grad_norm": 1.912945032119751,
+      "learning_rate": 2e-05,
+      "loss": 0.05409247,
+      "step": 15665
+    },
+    {
+      "epoch": 31.332,
+      "grad_norm": 2.0025062561035156,
+      "learning_rate": 2e-05,
+      "loss": 0.05454734,
+      "step": 15666
+    },
+    {
+      "epoch": 31.334,
+      "grad_norm": 1.087705373764038,
+      "learning_rate": 2e-05,
+      "loss": 0.04066493,
+      "step": 15667
+    },
+    {
+      "epoch": 31.336,
+      "grad_norm": 1.9127389192581177,
+      "learning_rate": 2e-05,
+      "loss": 0.05669509,
+      "step": 15668
+    },
+    {
+      "epoch": 31.338,
+      "grad_norm": 1.421940803527832,
+      "learning_rate": 2e-05,
+      "loss": 0.05013736,
+      "step": 15669
+    },
+    {
+      "epoch": 31.34,
+      "grad_norm": 2.1149110794067383,
+      "learning_rate": 2e-05,
+      "loss": 0.05951951,
+      "step": 15670
+    },
+    {
+      "epoch": 31.342,
+      "grad_norm": 2.23716402053833,
+      "learning_rate": 2e-05,
+      "loss": 0.04989032,
+      "step": 15671
+    },
+    {
+      "epoch": 31.344,
+      "grad_norm": 1.1148024797439575,
+      "learning_rate": 2e-05,
+      "loss": 0.03459584,
+      "step": 15672
+    },
+    {
+      "epoch": 31.346,
+      "grad_norm": 1.1864360570907593,
+      "learning_rate": 2e-05,
+      "loss": 0.04163406,
+      "step": 15673
+    },
+    {
+      "epoch": 31.348,
+      "grad_norm": 1.1121900081634521,
+      "learning_rate": 2e-05,
+      "loss": 0.04536529,
+      "step": 15674
+    },
+    {
+      "epoch": 31.35,
+      "grad_norm": 1.37013840675354,
+      "learning_rate": 2e-05,
+      "loss": 0.03631928,
+      "step": 15675
+    },
+    {
+      "epoch": 31.352,
+      "grad_norm": 1.0761029720306396,
+      "learning_rate": 2e-05,
+      "loss": 0.04334534,
+      "step": 15676
+    },
+    {
+      "epoch": 31.354,
+      "grad_norm": 1.2131975889205933,
+      "learning_rate": 2e-05,
+      "loss": 0.03891152,
+      "step": 15677
+    },
+    {
+      "epoch": 31.356,
+      "grad_norm": 1.1239383220672607,
+      "learning_rate": 2e-05,
+      "loss": 0.03443652,
+      "step": 15678
+    },
+    {
+      "epoch": 31.358,
+      "grad_norm": 1.4421390295028687,
+      "learning_rate": 2e-05,
+      "loss": 0.03504834,
+      "step": 15679
+    },
+    {
+      "epoch": 31.36,
+      "grad_norm": 1.6474403142929077,
+      "learning_rate": 2e-05,
+      "loss": 0.06349871,
+      "step": 15680
+    },
+    {
+      "epoch": 31.362,
+      "grad_norm": 1.2030842304229736,
+      "learning_rate": 2e-05,
+      "loss": 0.03427202,
+      "step": 15681
+    },
+    {
+      "epoch": 31.364,
+      "grad_norm": 1.1239585876464844,
+      "learning_rate": 2e-05,
+      "loss": 0.03592511,
+      "step": 15682
+    },
+    {
+      "epoch": 31.366,
+      "grad_norm": 1.0693938732147217,
+      "learning_rate": 2e-05,
+      "loss": 0.03550804,
+      "step": 15683
+    },
+    {
+      "epoch": 31.368,
+      "grad_norm": 1.2990278005599976,
+      "learning_rate": 2e-05,
+      "loss": 0.04546881,
+      "step": 15684
+    },
+    {
+      "epoch": 31.37,
+      "grad_norm": 1.4115841388702393,
+      "learning_rate": 2e-05,
+      "loss": 0.04322497,
+      "step": 15685
+    },
+    {
+      "epoch": 31.372,
+      "grad_norm": 1.1592683792114258,
+      "learning_rate": 2e-05,
+      "loss": 0.04757714,
+      "step": 15686
+    },
+    {
+      "epoch": 31.374,
+      "grad_norm": 1.0848994255065918,
+      "learning_rate": 2e-05,
+      "loss": 0.04322045,
+      "step": 15687
+    },
+    {
+      "epoch": 31.376,
+      "grad_norm": 1.319326639175415,
+      "learning_rate": 2e-05,
+      "loss": 0.04263577,
+      "step": 15688
+    },
+    {
+      "epoch": 31.378,
+      "grad_norm": 1.180602788925171,
+      "learning_rate": 2e-05,
+      "loss": 0.04362357,
+      "step": 15689
+    },
+    {
+      "epoch": 31.38,
+      "grad_norm": 1.1107655763626099,
+      "learning_rate": 2e-05,
+      "loss": 0.05218142,
+      "step": 15690
+    },
+    {
+      "epoch": 31.382,
+      "grad_norm": 1.0870996713638306,
+      "learning_rate": 2e-05,
+      "loss": 0.03446791,
+      "step": 15691
+    },
+    {
+      "epoch": 31.384,
+      "grad_norm": 1.0287660360336304,
+      "learning_rate": 2e-05,
+      "loss": 0.03068786,
+      "step": 15692
+    },
+    {
+      "epoch": 31.386,
+      "grad_norm": 1.2678953409194946,
+      "learning_rate": 2e-05,
+      "loss": 0.02981563,
+      "step": 15693
+    },
+    {
+      "epoch": 31.388,
+      "grad_norm": 1.2370007038116455,
+      "learning_rate": 2e-05,
+      "loss": 0.04931886,
+      "step": 15694
+    },
+    {
+      "epoch": 31.39,
+      "grad_norm": 1.0795314311981201,
+      "learning_rate": 2e-05,
+      "loss": 0.03978341,
+      "step": 15695
+    },
+    {
+      "epoch": 31.392,
+      "grad_norm": 1.0488916635513306,
+      "learning_rate": 2e-05,
+      "loss": 0.04341751,
+      "step": 15696
+    },
+    {
+      "epoch": 31.394,
+      "grad_norm": 1.9883416891098022,
+      "learning_rate": 2e-05,
+      "loss": 0.04387866,
+      "step": 15697
+    },
+    {
+      "epoch": 31.396,
+      "grad_norm": 1.0542610883712769,
+      "learning_rate": 2e-05,
+      "loss": 0.04586036,
+      "step": 15698
+    },
+    {
+      "epoch": 31.398,
+      "grad_norm": 1.5984288454055786,
+      "learning_rate": 2e-05,
+      "loss": 0.04890535,
+      "step": 15699
+    },
+    {
+      "epoch": 31.4,
+      "grad_norm": 1.2893099784851074,
+      "learning_rate": 2e-05,
+      "loss": 0.0567109,
+      "step": 15700
+    },
+    {
+      "epoch": 31.402,
+      "grad_norm": 1.1580631732940674,
+      "learning_rate": 2e-05,
+      "loss": 0.04688855,
+      "step": 15701
+    },
+    {
+      "epoch": 31.404,
+      "grad_norm": 1.5734524726867676,
+      "learning_rate": 2e-05,
+      "loss": 0.06079507,
+      "step": 15702
+    },
+    {
+      "epoch": 31.406,
+      "grad_norm": 1.2681607007980347,
+      "learning_rate": 2e-05,
+      "loss": 0.0585427,
+      "step": 15703
+    },
+    {
+      "epoch": 31.408,
+      "grad_norm": 1.4479280710220337,
+      "learning_rate": 2e-05,
+      "loss": 0.0500034,
+      "step": 15704
+    },
+    {
+      "epoch": 31.41,
+      "grad_norm": 1.5085829496383667,
+      "learning_rate": 2e-05,
+      "loss": 0.04053114,
+      "step": 15705
+    },
+    {
+      "epoch": 31.412,
+      "grad_norm": 1.2420114278793335,
+      "learning_rate": 2e-05,
+      "loss": 0.0312696,
+      "step": 15706
+    },
+    {
+      "epoch": 31.414,
+      "grad_norm": 1.4845657348632812,
+      "learning_rate": 2e-05,
+      "loss": 0.04520275,
+      "step": 15707
+    },
+    {
+      "epoch": 31.416,
+      "grad_norm": 1.0234241485595703,
+      "learning_rate": 2e-05,
+      "loss": 0.03995465,
+      "step": 15708
+    },
+    {
+      "epoch": 31.418,
+      "grad_norm": 0.9769031405448914,
+      "learning_rate": 2e-05,
+      "loss": 0.03126371,
+      "step": 15709
+    },
+    {
+      "epoch": 31.42,
+      "grad_norm": 1.2428449392318726,
+      "learning_rate": 2e-05,
+      "loss": 0.04003659,
+      "step": 15710
+    },
+    {
+      "epoch": 31.422,
+      "grad_norm": 1.1226387023925781,
+      "learning_rate": 2e-05,
+      "loss": 0.04901841,
+      "step": 15711
+    },
+    {
+      "epoch": 31.424,
+      "grad_norm": 0.9852843284606934,
+      "learning_rate": 2e-05,
+      "loss": 0.03026243,
+      "step": 15712
+    },
+    {
+      "epoch": 31.426,
+      "grad_norm": 1.0646196603775024,
+      "learning_rate": 2e-05,
+      "loss": 0.03818695,
+      "step": 15713
+    },
+    {
+      "epoch": 31.428,
+      "grad_norm": 1.188592553138733,
+      "learning_rate": 2e-05,
+      "loss": 0.03416793,
+      "step": 15714
+    },
+    {
+      "epoch": 31.43,
+      "grad_norm": 1.1710172891616821,
+      "learning_rate": 2e-05,
+      "loss": 0.03767353,
+      "step": 15715
+    },
+    {
+      "epoch": 31.432,
+      "grad_norm": 1.2032878398895264,
+      "learning_rate": 2e-05,
+      "loss": 0.04838914,
+      "step": 15716
+    },
+    {
+      "epoch": 31.434,
+      "grad_norm": 1.038709282875061,
+      "learning_rate": 2e-05,
+      "loss": 0.03334371,
+      "step": 15717
+    },
+    {
+      "epoch": 31.436,
+      "grad_norm": 1.0788524150848389,
+      "learning_rate": 2e-05,
+      "loss": 0.03386791,
+      "step": 15718
+    },
+    {
+      "epoch": 31.438,
+      "grad_norm": 1.163702368736267,
+      "learning_rate": 2e-05,
+      "loss": 0.0470679,
+      "step": 15719
+    },
+    {
+      "epoch": 31.44,
+      "grad_norm": 1.1703226566314697,
+      "learning_rate": 2e-05,
+      "loss": 0.04177292,
+      "step": 15720
+    },
+    {
+      "epoch": 31.442,
+      "grad_norm": 2.1952016353607178,
+      "learning_rate": 2e-05,
+      "loss": 0.03891946,
+      "step": 15721
+    },
+    {
+      "epoch": 31.444,
+      "grad_norm": 1.4332395792007446,
+      "learning_rate": 2e-05,
+      "loss": 0.05187111,
+      "step": 15722
+    },
+    {
+      "epoch": 31.446,
+      "grad_norm": 1.305293083190918,
+      "learning_rate": 2e-05,
+      "loss": 0.04548291,
+      "step": 15723
+    },
+    {
+      "epoch": 31.448,
+      "grad_norm": 1.0815423727035522,
+      "learning_rate": 2e-05,
+      "loss": 0.04857568,
+      "step": 15724
+    },
+    {
+      "epoch": 31.45,
+      "grad_norm": 1.2314715385437012,
+      "learning_rate": 2e-05,
+      "loss": 0.03415262,
+      "step": 15725
+    },
+    {
+      "epoch": 31.452,
+      "grad_norm": 1.8589730262756348,
+      "learning_rate": 2e-05,
+      "loss": 0.04466228,
+      "step": 15726
+    },
+    {
+      "epoch": 31.454,
+      "grad_norm": 1.4462237358093262,
+      "learning_rate": 2e-05,
+      "loss": 0.05154398,
+      "step": 15727
+    },
+    {
+      "epoch": 31.456,
+      "grad_norm": 1.0429617166519165,
+      "learning_rate": 2e-05,
+      "loss": 0.03432951,
+      "step": 15728
+    },
+    {
+      "epoch": 31.458,
+      "grad_norm": 1.2360962629318237,
+      "learning_rate": 2e-05,
+      "loss": 0.04059312,
+      "step": 15729
+    },
+    {
+      "epoch": 31.46,
+      "grad_norm": 1.1760252714157104,
+      "learning_rate": 2e-05,
+      "loss": 0.05076623,
+      "step": 15730
+    },
+    {
+      "epoch": 31.462,
+      "grad_norm": 1.2189456224441528,
+      "learning_rate": 2e-05,
+      "loss": 0.05206157,
+      "step": 15731
+    },
+    {
+      "epoch": 31.464,
+      "grad_norm": 1.028828740119934,
+      "learning_rate": 2e-05,
+      "loss": 0.02622063,
+      "step": 15732
+    },
+    {
+      "epoch": 31.466,
+      "grad_norm": 1.3809655904769897,
+      "learning_rate": 2e-05,
+      "loss": 0.04626153,
+      "step": 15733
+    },
+    {
+      "epoch": 31.468,
+      "grad_norm": 1.0601438283920288,
+      "learning_rate": 2e-05,
+      "loss": 0.03184626,
+      "step": 15734
+    },
+    {
+      "epoch": 31.47,
+      "grad_norm": 1.7758816480636597,
+      "learning_rate": 2e-05,
+      "loss": 0.04071598,
+      "step": 15735
+    },
+    {
+      "epoch": 31.472,
+      "grad_norm": 1.2127716541290283,
+      "learning_rate": 2e-05,
+      "loss": 0.0404386,
+      "step": 15736
+    },
+    {
+      "epoch": 31.474,
+      "grad_norm": 1.1199473142623901,
+      "learning_rate": 2e-05,
+      "loss": 0.03128079,
+      "step": 15737
+    },
+    {
+      "epoch": 31.476,
+      "grad_norm": 0.9239538908004761,
+      "learning_rate": 2e-05,
+      "loss": 0.0214905,
+      "step": 15738
+    },
+    {
+      "epoch": 31.478,
+      "grad_norm": 1.0289355516433716,
+      "learning_rate": 2e-05,
+      "loss": 0.03084357,
+      "step": 15739
+    },
+    {
+      "epoch": 31.48,
+      "grad_norm": 1.7557429075241089,
+      "learning_rate": 2e-05,
+      "loss": 0.04279177,
+      "step": 15740
+    },
+    {
+      "epoch": 31.482,
+      "grad_norm": 2.431920289993286,
+      "learning_rate": 2e-05,
+      "loss": 0.04216561,
+      "step": 15741
+    },
+    {
+      "epoch": 31.484,
+      "grad_norm": 1.3500877618789673,
+      "learning_rate": 2e-05,
+      "loss": 0.04621994,
+      "step": 15742
+    },
+    {
+      "epoch": 31.486,
+      "grad_norm": 1.0111196041107178,
+      "learning_rate": 2e-05,
+      "loss": 0.02924443,
+      "step": 15743
+    },
+    {
+      "epoch": 31.488,
+      "grad_norm": 1.5376261472702026,
+      "learning_rate": 2e-05,
+      "loss": 0.05262484,
+      "step": 15744
+    },
+    {
+      "epoch": 31.49,
+      "grad_norm": 1.3182742595672607,
+      "learning_rate": 2e-05,
+      "loss": 0.0479708,
+      "step": 15745
+    },
+    {
+      "epoch": 31.492,
+      "grad_norm": 1.2966049909591675,
+      "learning_rate": 2e-05,
+      "loss": 0.03965274,
+      "step": 15746
+    },
+    {
+      "epoch": 31.494,
+      "grad_norm": 1.1283825635910034,
+      "learning_rate": 2e-05,
+      "loss": 0.0380425,
+      "step": 15747
+    },
+    {
+      "epoch": 31.496,
+      "grad_norm": 1.2638022899627686,
+      "learning_rate": 2e-05,
+      "loss": 0.0417271,
+      "step": 15748
+    },
+    {
+      "epoch": 31.498,
+      "grad_norm": 1.2083308696746826,
+      "learning_rate": 2e-05,
+      "loss": 0.04929405,
+      "step": 15749
+    },
+    {
+      "epoch": 31.5,
+      "grad_norm": 1.5783114433288574,
+      "learning_rate": 2e-05,
+      "loss": 0.04968805,
+      "step": 15750
+    },
+    {
+      "epoch": 31.502,
+      "grad_norm": 2.3429791927337646,
+      "learning_rate": 2e-05,
+      "loss": 0.0579337,
+      "step": 15751
+    },
+    {
+      "epoch": 31.504,
+      "grad_norm": 1.2848727703094482,
+      "learning_rate": 2e-05,
+      "loss": 0.03835206,
+      "step": 15752
+    },
+    {
+      "epoch": 31.506,
+      "grad_norm": 1.3516781330108643,
+      "learning_rate": 2e-05,
+      "loss": 0.04133689,
+      "step": 15753
+    },
+    {
+      "epoch": 31.508,
+      "grad_norm": 1.0922443866729736,
+      "learning_rate": 2e-05,
+      "loss": 0.03277455,
+      "step": 15754
+    },
+    {
+      "epoch": 31.51,
+      "grad_norm": 1.2325778007507324,
+      "learning_rate": 2e-05,
+      "loss": 0.04386805,
+      "step": 15755
+    },
+    {
+      "epoch": 31.512,
+      "grad_norm": 1.190375566482544,
+      "learning_rate": 2e-05,
+      "loss": 0.04679071,
+      "step": 15756
+    },
+    {
+      "epoch": 31.514,
+      "grad_norm": 1.1857788562774658,
+      "learning_rate": 2e-05,
+      "loss": 0.04052466,
+      "step": 15757
+    },
+    {
+      "epoch": 31.516,
+      "grad_norm": 1.4863660335540771,
+      "learning_rate": 2e-05,
+      "loss": 0.04734954,
+      "step": 15758
+    },
+    {
+      "epoch": 31.518,
+      "grad_norm": 0.9761635661125183,
+      "learning_rate": 2e-05,
+      "loss": 0.04507826,
+      "step": 15759
+    },
+    {
+      "epoch": 31.52,
+      "grad_norm": 3.303678274154663,
+      "learning_rate": 2e-05,
+      "loss": 0.04152103,
+      "step": 15760
+    },
+    {
+      "epoch": 31.522,
+      "grad_norm": 1.3020141124725342,
+      "learning_rate": 2e-05,
+      "loss": 0.03344776,
+      "step": 15761
+    },
+    {
+      "epoch": 31.524,
+      "grad_norm": 1.2744166851043701,
+      "learning_rate": 2e-05,
+      "loss": 0.03905339,
+      "step": 15762
+    },
+    {
+      "epoch": 31.526,
+      "grad_norm": 1.3360401391983032,
+      "learning_rate": 2e-05,
+      "loss": 0.05134521,
+      "step": 15763
+    },
+    {
+      "epoch": 31.528,
+      "grad_norm": 1.5633838176727295,
+      "learning_rate": 2e-05,
+      "loss": 0.04629926,
+      "step": 15764
+    },
+    {
+      "epoch": 31.53,
+      "grad_norm": 1.144903540611267,
+      "learning_rate": 2e-05,
+      "loss": 0.04337474,
+      "step": 15765
+    },
+    {
+      "epoch": 31.532,
+      "grad_norm": 1.549691081047058,
+      "learning_rate": 2e-05,
+      "loss": 0.02903229,
+      "step": 15766
+    },
+    {
+      "epoch": 31.534,
+      "grad_norm": 1.6548807621002197,
+      "learning_rate": 2e-05,
+      "loss": 0.03670297,
+      "step": 15767
+    },
+    {
+      "epoch": 31.536,
+      "grad_norm": 1.8595404624938965,
+      "learning_rate": 2e-05,
+      "loss": 0.04848412,
+      "step": 15768
+    },
+    {
+      "epoch": 31.538,
+      "grad_norm": 0.9308550953865051,
+      "learning_rate": 2e-05,
+      "loss": 0.01970943,
+      "step": 15769
+    },
+    {
+      "epoch": 31.54,
+      "grad_norm": 1.6548179388046265,
+      "learning_rate": 2e-05,
+      "loss": 0.05121253,
+      "step": 15770
+    },
+    {
+      "epoch": 31.542,
+      "grad_norm": 1.1826523542404175,
+      "learning_rate": 2e-05,
+      "loss": 0.03165106,
+      "step": 15771
+    },
+    {
+      "epoch": 31.544,
+      "grad_norm": 2.287088394165039,
+      "learning_rate": 2e-05,
+      "loss": 0.05559523,
+      "step": 15772
+    },
+    {
+      "epoch": 31.546,
+      "grad_norm": 1.249127984046936,
+      "learning_rate": 2e-05,
+      "loss": 0.05151629,
+      "step": 15773
+    },
+    {
+      "epoch": 31.548000000000002,
+      "grad_norm": 2.0954861640930176,
+      "learning_rate": 2e-05,
+      "loss": 0.05453818,
+      "step": 15774
+    },
+    {
+      "epoch": 31.55,
+      "grad_norm": 1.1801806688308716,
+      "learning_rate": 2e-05,
+      "loss": 0.04041165,
+      "step": 15775
+    },
+    {
+      "epoch": 31.552,
+      "grad_norm": 1.1119970083236694,
+      "learning_rate": 2e-05,
+      "loss": 0.0460862,
+      "step": 15776
+    },
+    {
+      "epoch": 31.554,
+      "grad_norm": 2.1345860958099365,
+      "learning_rate": 2e-05,
+      "loss": 0.04225209,
+      "step": 15777
+    },
+    {
+      "epoch": 31.556,
+      "grad_norm": 1.1420177221298218,
+      "learning_rate": 2e-05,
+      "loss": 0.03541413,
+      "step": 15778
+    },
+    {
+      "epoch": 31.558,
+      "grad_norm": 1.9255207777023315,
+      "learning_rate": 2e-05,
+      "loss": 0.05924919,
+      "step": 15779
+    },
+    {
+      "epoch": 31.56,
+      "grad_norm": 3.971301317214966,
+      "learning_rate": 2e-05,
+      "loss": 0.05418454,
+      "step": 15780
+    },
+    {
+      "epoch": 31.562,
+      "grad_norm": 1.384949803352356,
+      "learning_rate": 2e-05,
+      "loss": 0.03826708,
+      "step": 15781
+    },
+    {
+      "epoch": 31.564,
+      "grad_norm": 1.1261556148529053,
+      "learning_rate": 2e-05,
+      "loss": 0.04444794,
+      "step": 15782
+    },
+    {
+      "epoch": 31.566,
+      "grad_norm": 1.696993112564087,
+      "learning_rate": 2e-05,
+      "loss": 0.03568056,
+      "step": 15783
+    },
+    {
+      "epoch": 31.568,
+      "grad_norm": 1.7849308252334595,
+      "learning_rate": 2e-05,
+      "loss": 0.03161497,
+      "step": 15784
+    },
+    {
+      "epoch": 31.57,
+      "grad_norm": 2.218967914581299,
+      "learning_rate": 2e-05,
+      "loss": 0.04789735,
+      "step": 15785
+    },
+    {
+      "epoch": 31.572,
+      "grad_norm": 1.280608892440796,
+      "learning_rate": 2e-05,
+      "loss": 0.04609762,
+      "step": 15786
+    },
+    {
+      "epoch": 31.574,
+      "grad_norm": 1.1164113283157349,
+      "learning_rate": 2e-05,
+      "loss": 0.03556387,
+      "step": 15787
+    },
+    {
+      "epoch": 31.576,
+      "grad_norm": 1.5488619804382324,
+      "learning_rate": 2e-05,
+      "loss": 0.06319657,
+      "step": 15788
+    },
+    {
+      "epoch": 31.578,
+      "grad_norm": 1.0593352317810059,
+      "learning_rate": 2e-05,
+      "loss": 0.0368279,
+      "step": 15789
+    },
+    {
+      "epoch": 31.58,
+      "grad_norm": 1.139189600944519,
+      "learning_rate": 2e-05,
+      "loss": 0.03742014,
+      "step": 15790
+    },
+    {
+      "epoch": 31.582,
+      "grad_norm": 1.1916528940200806,
+      "learning_rate": 2e-05,
+      "loss": 0.05514635,
+      "step": 15791
+    },
+    {
+      "epoch": 31.584,
+      "grad_norm": 2.922518491744995,
+      "learning_rate": 2e-05,
+      "loss": 0.06346483,
+      "step": 15792
+    },
+    {
+      "epoch": 31.586,
+      "grad_norm": 1.1857612133026123,
+      "learning_rate": 2e-05,
+      "loss": 0.04470282,
+      "step": 15793
+    },
+    {
+      "epoch": 31.588,
+      "grad_norm": 1.5709048509597778,
+      "learning_rate": 2e-05,
+      "loss": 0.05266313,
+      "step": 15794
+    },
+    {
+      "epoch": 31.59,
+      "grad_norm": 1.2784346342086792,
+      "learning_rate": 2e-05,
+      "loss": 0.02789311,
+      "step": 15795
+    },
+    {
+      "epoch": 31.592,
+      "grad_norm": 1.4792814254760742,
+      "learning_rate": 2e-05,
+      "loss": 0.06371169,
+      "step": 15796
+    },
+    {
+      "epoch": 31.594,
+      "grad_norm": 1.1451568603515625,
+      "learning_rate": 2e-05,
+      "loss": 0.03802731,
+      "step": 15797
+    },
+    {
+      "epoch": 31.596,
+      "grad_norm": 1.5499827861785889,
+      "learning_rate": 2e-05,
+      "loss": 0.05359296,
+      "step": 15798
+    },
+    {
+      "epoch": 31.598,
+      "grad_norm": 1.11916983127594,
+      "learning_rate": 2e-05,
+      "loss": 0.05549221,
+      "step": 15799
+    },
+    {
+      "epoch": 31.6,
+      "grad_norm": 1.0967146158218384,
+      "learning_rate": 2e-05,
+      "loss": 0.03953513,
+      "step": 15800
+    },
+    {
+      "epoch": 31.602,
+      "grad_norm": 1.4866015911102295,
+      "learning_rate": 2e-05,
+      "loss": 0.0373404,
+      "step": 15801
+    },
+    {
+      "epoch": 31.604,
+      "grad_norm": 1.3071479797363281,
+      "learning_rate": 2e-05,
+      "loss": 0.03140751,
+      "step": 15802
+    },
+    {
+      "epoch": 31.606,
+      "grad_norm": 1.8923872709274292,
+      "learning_rate": 2e-05,
+      "loss": 0.03232081,
+      "step": 15803
+    },
+    {
+      "epoch": 31.608,
+      "grad_norm": 1.0222185850143433,
+      "learning_rate": 2e-05,
+      "loss": 0.03265133,
+      "step": 15804
+    },
+    {
+      "epoch": 31.61,
+      "grad_norm": 1.0511828660964966,
+      "learning_rate": 2e-05,
+      "loss": 0.02982992,
+      "step": 15805
+    },
+    {
+      "epoch": 31.612,
+      "grad_norm": 1.0626296997070312,
+      "learning_rate": 2e-05,
+      "loss": 0.03582488,
+      "step": 15806
+    },
+    {
+      "epoch": 31.614,
+      "grad_norm": 1.239322304725647,
+      "learning_rate": 2e-05,
+      "loss": 0.04029377,
+      "step": 15807
+    },
+    {
+      "epoch": 31.616,
+      "grad_norm": 1.2532638311386108,
+      "learning_rate": 2e-05,
+      "loss": 0.04826434,
+      "step": 15808
+    },
+    {
+      "epoch": 31.618,
+      "grad_norm": 0.907904326915741,
+      "learning_rate": 2e-05,
+      "loss": 0.032974,
+      "step": 15809
+    },
+    {
+      "epoch": 31.62,
+      "grad_norm": 1.3583287000656128,
+      "learning_rate": 2e-05,
+      "loss": 0.05751316,
+      "step": 15810
+    },
+    {
+      "epoch": 31.622,
+      "grad_norm": 0.8747350573539734,
+      "learning_rate": 2e-05,
+      "loss": 0.02985252,
+      "step": 15811
+    },
+    {
+      "epoch": 31.624,
+      "grad_norm": 1.2014013528823853,
+      "learning_rate": 2e-05,
+      "loss": 0.0308294,
+      "step": 15812
+    },
+    {
+      "epoch": 31.626,
+      "grad_norm": 1.563045620918274,
+      "learning_rate": 2e-05,
+      "loss": 0.0466816,
+      "step": 15813
+    },
+    {
+      "epoch": 31.628,
+      "grad_norm": 1.5340209007263184,
+      "learning_rate": 2e-05,
+      "loss": 0.04322568,
+      "step": 15814
+    },
+    {
+      "epoch": 31.63,
+      "grad_norm": 1.1069697141647339,
+      "learning_rate": 2e-05,
+      "loss": 0.02941058,
+      "step": 15815
+    },
+    {
+      "epoch": 31.632,
+      "grad_norm": 1.6199052333831787,
+      "learning_rate": 2e-05,
+      "loss": 0.06167778,
+      "step": 15816
+    },
+    {
+      "epoch": 31.634,
+      "grad_norm": 0.9924496412277222,
+      "learning_rate": 2e-05,
+      "loss": 0.02608846,
+      "step": 15817
+    },
+    {
+      "epoch": 31.636,
+      "grad_norm": 1.212367296218872,
+      "learning_rate": 2e-05,
+      "loss": 0.05580743,
+      "step": 15818
+    },
+    {
+      "epoch": 31.638,
+      "grad_norm": 1.232508897781372,
+      "learning_rate": 2e-05,
+      "loss": 0.04425735,
+      "step": 15819
+    },
+    {
+      "epoch": 31.64,
+      "grad_norm": 2.6619043350219727,
+      "learning_rate": 2e-05,
+      "loss": 0.06107927,
+      "step": 15820
+    },
+    {
+      "epoch": 31.642,
+      "grad_norm": 1.5662249326705933,
+      "learning_rate": 2e-05,
+      "loss": 0.04231949,
+      "step": 15821
+    },
+    {
+      "epoch": 31.644,
+      "grad_norm": 1.066655158996582,
+      "learning_rate": 2e-05,
+      "loss": 0.03843967,
+      "step": 15822
+    },
+    {
+      "epoch": 31.646,
+      "grad_norm": 1.9721606969833374,
+      "learning_rate": 2e-05,
+      "loss": 0.05530218,
+      "step": 15823
+    },
+    {
+      "epoch": 31.648,
+      "grad_norm": 1.093684196472168,
+      "learning_rate": 2e-05,
+      "loss": 0.04849446,
+      "step": 15824
+    },
+    {
+      "epoch": 31.65,
+      "grad_norm": 1.434188961982727,
+      "learning_rate": 2e-05,
+      "loss": 0.05032559,
+      "step": 15825
+    },
+    {
+      "epoch": 31.652,
+      "grad_norm": 1.0273107290267944,
+      "learning_rate": 2e-05,
+      "loss": 0.03886499,
+      "step": 15826
+    },
+    {
+      "epoch": 31.654,
+      "grad_norm": 1.0134077072143555,
+      "learning_rate": 2e-05,
+      "loss": 0.0312471,
+      "step": 15827
+    },
+    {
+      "epoch": 31.656,
+      "grad_norm": 1.69369637966156,
+      "learning_rate": 2e-05,
+      "loss": 0.05736078,
+      "step": 15828
+    },
+    {
+      "epoch": 31.658,
+      "grad_norm": 3.8166019916534424,
+      "learning_rate": 2e-05,
+      "loss": 0.05662319,
+      "step": 15829
+    },
+    {
+      "epoch": 31.66,
+      "grad_norm": 1.4313768148422241,
+      "learning_rate": 2e-05,
+      "loss": 0.04016527,
+      "step": 15830
+    },
+    {
+      "epoch": 31.662,
+      "grad_norm": 1.6341408491134644,
+      "learning_rate": 2e-05,
+      "loss": 0.05576465,
+      "step": 15831
+    },
+    {
+      "epoch": 31.664,
+      "grad_norm": 1.6464592218399048,
+      "learning_rate": 2e-05,
+      "loss": 0.05058682,
+      "step": 15832
+    },
+    {
+      "epoch": 31.666,
+      "grad_norm": 1.6304293870925903,
+      "learning_rate": 2e-05,
+      "loss": 0.04562239,
+      "step": 15833
+    },
+    {
+      "epoch": 31.668,
+      "grad_norm": 1.1134459972381592,
+      "learning_rate": 2e-05,
+      "loss": 0.0356195,
+      "step": 15834
+    },
+    {
+      "epoch": 31.67,
+      "grad_norm": 0.9937390685081482,
+      "learning_rate": 2e-05,
+      "loss": 0.02685843,
+      "step": 15835
+    },
+    {
+      "epoch": 31.672,
+      "grad_norm": 1.066178321838379,
+      "learning_rate": 2e-05,
+      "loss": 0.03931656,
+      "step": 15836
+    },
+    {
+      "epoch": 31.674,
+      "grad_norm": 1.0088249444961548,
+      "learning_rate": 2e-05,
+      "loss": 0.04021916,
+      "step": 15837
+    },
+    {
+      "epoch": 31.676,
+      "grad_norm": 1.028534173965454,
+      "learning_rate": 2e-05,
+      "loss": 0.05189967,
+      "step": 15838
+    },
+    {
+      "epoch": 31.678,
+      "grad_norm": 4.298219203948975,
+      "learning_rate": 2e-05,
+      "loss": 0.03603112,
+      "step": 15839
+    },
+    {
+      "epoch": 31.68,
+      "grad_norm": 1.1019927263259888,
+      "learning_rate": 2e-05,
+      "loss": 0.03563758,
+      "step": 15840
+    },
+    {
+      "epoch": 31.682,
+      "grad_norm": 1.016705870628357,
+      "learning_rate": 2e-05,
+      "loss": 0.03160737,
+      "step": 15841
+    },
+    {
+      "epoch": 31.684,
+      "grad_norm": 1.2760132551193237,
+      "learning_rate": 2e-05,
+      "loss": 0.0512272,
+      "step": 15842
+    },
+    {
+      "epoch": 31.686,
+      "grad_norm": 1.0675709247589111,
+      "learning_rate": 2e-05,
+      "loss": 0.03803336,
+      "step": 15843
+    },
+    {
+      "epoch": 31.688,
+      "grad_norm": 1.0932615995407104,
+      "learning_rate": 2e-05,
+      "loss": 0.03163125,
+      "step": 15844
+    },
+    {
+      "epoch": 31.69,
+      "grad_norm": 1.4388149976730347,
+      "learning_rate": 2e-05,
+      "loss": 0.0546472,
+      "step": 15845
+    },
+    {
+      "epoch": 31.692,
+      "grad_norm": 1.247949242591858,
+      "learning_rate": 2e-05,
+      "loss": 0.04308487,
+      "step": 15846
+    },
+    {
+      "epoch": 31.694,
+      "grad_norm": 1.0364240407943726,
+      "learning_rate": 2e-05,
+      "loss": 0.04179894,
+      "step": 15847
+    },
+    {
+      "epoch": 31.696,
+      "grad_norm": 1.3593528270721436,
+      "learning_rate": 2e-05,
+      "loss": 0.03463472,
+      "step": 15848
+    },
+    {
+      "epoch": 31.698,
+      "grad_norm": 3.9153060913085938,
+      "learning_rate": 2e-05,
+      "loss": 0.0593623,
+      "step": 15849
+    },
+    {
+      "epoch": 31.7,
+      "grad_norm": 1.224266529083252,
+      "learning_rate": 2e-05,
+      "loss": 0.06305296,
+      "step": 15850
+    },
+    {
+      "epoch": 31.701999999999998,
+      "grad_norm": 1.8977373838424683,
+      "learning_rate": 2e-05,
+      "loss": 0.04459673,
+      "step": 15851
+    },
+    {
+      "epoch": 31.704,
+      "grad_norm": 1.1238889694213867,
+      "learning_rate": 2e-05,
+      "loss": 0.03320381,
+      "step": 15852
+    },
+    {
+      "epoch": 31.706,
+      "grad_norm": 1.8452239036560059,
+      "learning_rate": 2e-05,
+      "loss": 0.03556684,
+      "step": 15853
+    },
+    {
+      "epoch": 31.708,
+      "grad_norm": 1.469277024269104,
+      "learning_rate": 2e-05,
+      "loss": 0.05553747,
+      "step": 15854
+    },
+    {
+      "epoch": 31.71,
+      "grad_norm": 1.2231842279434204,
+      "learning_rate": 2e-05,
+      "loss": 0.04826688,
+      "step": 15855
+    },
+    {
+      "epoch": 31.712,
+      "grad_norm": 0.8785945177078247,
+      "learning_rate": 2e-05,
+      "loss": 0.03162833,
+      "step": 15856
+    },
+    {
+      "epoch": 31.714,
+      "grad_norm": 1.2260061502456665,
+      "learning_rate": 2e-05,
+      "loss": 0.04555026,
+      "step": 15857
+    },
+    {
+      "epoch": 31.716,
+      "grad_norm": 1.4658849239349365,
+      "learning_rate": 2e-05,
+      "loss": 0.04575991,
+      "step": 15858
+    },
+    {
+      "epoch": 31.718,
+      "grad_norm": 1.1947789192199707,
+      "learning_rate": 2e-05,
+      "loss": 0.03831368,
+      "step": 15859
+    },
+    {
+      "epoch": 31.72,
+      "grad_norm": 2.118058443069458,
+      "learning_rate": 2e-05,
+      "loss": 0.05042726,
+      "step": 15860
+    },
+    {
+      "epoch": 31.722,
+      "grad_norm": 0.9737758040428162,
+      "learning_rate": 2e-05,
+      "loss": 0.03438409,
+      "step": 15861
+    },
+    {
+      "epoch": 31.724,
+      "grad_norm": 1.606362223625183,
+      "learning_rate": 2e-05,
+      "loss": 0.03443389,
+      "step": 15862
+    },
+    {
+      "epoch": 31.726,
+      "grad_norm": 1.9236388206481934,
+      "learning_rate": 2e-05,
+      "loss": 0.04065321,
+      "step": 15863
+    },
+    {
+      "epoch": 31.728,
+      "grad_norm": 1.1777132749557495,
+      "learning_rate": 2e-05,
+      "loss": 0.04068302,
+      "step": 15864
+    },
+    {
+      "epoch": 31.73,
+      "grad_norm": 1.356837272644043,
+      "learning_rate": 2e-05,
+      "loss": 0.04885522,
+      "step": 15865
+    },
+    {
+      "epoch": 31.732,
+      "grad_norm": 1.158024787902832,
+      "learning_rate": 2e-05,
+      "loss": 0.03375251,
+      "step": 15866
+    },
+    {
+      "epoch": 31.734,
+      "grad_norm": 1.4615778923034668,
+      "learning_rate": 2e-05,
+      "loss": 0.03782218,
+      "step": 15867
+    },
+    {
+      "epoch": 31.736,
+      "grad_norm": 1.936156153678894,
+      "learning_rate": 2e-05,
+      "loss": 0.05020786,
+      "step": 15868
+    },
+    {
+      "epoch": 31.738,
+      "grad_norm": 1.1056143045425415,
+      "learning_rate": 2e-05,
+      "loss": 0.03084729,
+      "step": 15869
+    },
+    {
+      "epoch": 31.74,
+      "grad_norm": 1.153363823890686,
+      "learning_rate": 2e-05,
+      "loss": 0.04279421,
+      "step": 15870
+    },
+    {
+      "epoch": 31.742,
+      "grad_norm": 2.0926640033721924,
+      "learning_rate": 2e-05,
+      "loss": 0.05298992,
+      "step": 15871
+    },
+    {
+      "epoch": 31.744,
+      "grad_norm": 0.9748860597610474,
+      "learning_rate": 2e-05,
+      "loss": 0.02891666,
+      "step": 15872
+    },
+    {
+      "epoch": 31.746,
+      "grad_norm": 2.135936737060547,
+      "learning_rate": 2e-05,
+      "loss": 0.04711745,
+      "step": 15873
+    },
+    {
+      "epoch": 31.748,
+      "grad_norm": 1.9730417728424072,
+      "learning_rate": 2e-05,
+      "loss": 0.0426465,
+      "step": 15874
+    },
+    {
+      "epoch": 31.75,
+      "grad_norm": 2.211915969848633,
+      "learning_rate": 2e-05,
+      "loss": 0.0576695,
+      "step": 15875
+    },
+    {
+      "epoch": 31.752,
+      "grad_norm": 1.4895312786102295,
+      "learning_rate": 2e-05,
+      "loss": 0.03817018,
+      "step": 15876
+    },
+    {
+      "epoch": 31.754,
+      "grad_norm": 1.1091728210449219,
+      "learning_rate": 2e-05,
+      "loss": 0.03473945,
+      "step": 15877
+    },
+    {
+      "epoch": 31.756,
+      "grad_norm": 1.5757970809936523,
+      "learning_rate": 2e-05,
+      "loss": 0.05376294,
+      "step": 15878
+    },
+    {
+      "epoch": 31.758,
+      "grad_norm": 1.4097827672958374,
+      "learning_rate": 2e-05,
+      "loss": 0.0575452,
+      "step": 15879
+    },
+    {
+      "epoch": 31.76,
+      "grad_norm": 1.601607084274292,
+      "learning_rate": 2e-05,
+      "loss": 0.03805475,
+      "step": 15880
+    },
+    {
+      "epoch": 31.762,
+      "grad_norm": 1.084592580795288,
+      "learning_rate": 2e-05,
+      "loss": 0.03560298,
+      "step": 15881
+    },
+    {
+      "epoch": 31.764,
+      "grad_norm": 1.3301918506622314,
+      "learning_rate": 2e-05,
+      "loss": 0.05340309,
+      "step": 15882
+    },
+    {
+      "epoch": 31.766,
+      "grad_norm": 1.21677565574646,
+      "learning_rate": 2e-05,
+      "loss": 0.02901489,
+      "step": 15883
+    },
+    {
+      "epoch": 31.768,
+      "grad_norm": 1.5509135723114014,
+      "learning_rate": 2e-05,
+      "loss": 0.0422479,
+      "step": 15884
+    },
+    {
+      "epoch": 31.77,
+      "grad_norm": 2.1377058029174805,
+      "learning_rate": 2e-05,
+      "loss": 0.03911637,
+      "step": 15885
+    },
+    {
+      "epoch": 31.772,
+      "grad_norm": 0.8442297577857971,
+      "learning_rate": 2e-05,
+      "loss": 0.02588265,
+      "step": 15886
+    },
+    {
+      "epoch": 31.774,
+      "grad_norm": 1.5694454908370972,
+      "learning_rate": 2e-05,
+      "loss": 0.04644613,
+      "step": 15887
+    },
+    {
+      "epoch": 31.776,
+      "grad_norm": 1.1110693216323853,
+      "learning_rate": 2e-05,
+      "loss": 0.03558012,
+      "step": 15888
+    },
+    {
+      "epoch": 31.778,
+      "grad_norm": 1.0614808797836304,
+      "learning_rate": 2e-05,
+      "loss": 0.03117304,
+      "step": 15889
+    },
+    {
+      "epoch": 31.78,
+      "grad_norm": 1.6479281187057495,
+      "learning_rate": 2e-05,
+      "loss": 0.03382275,
+      "step": 15890
+    },
+    {
+      "epoch": 31.782,
+      "grad_norm": 3.055795192718506,
+      "learning_rate": 2e-05,
+      "loss": 0.02290535,
+      "step": 15891
+    },
+    {
+      "epoch": 31.784,
+      "grad_norm": 1.0296449661254883,
+      "learning_rate": 2e-05,
+      "loss": 0.04224445,
+      "step": 15892
+    },
+    {
+      "epoch": 31.786,
+      "grad_norm": 0.8810853958129883,
+      "learning_rate": 2e-05,
+      "loss": 0.03229618,
+      "step": 15893
+    },
+    {
+      "epoch": 31.788,
+      "grad_norm": 2.7027084827423096,
+      "learning_rate": 2e-05,
+      "loss": 0.05507965,
+      "step": 15894
+    },
+    {
+      "epoch": 31.79,
+      "grad_norm": 1.3966178894042969,
+      "learning_rate": 2e-05,
+      "loss": 0.03760487,
+      "step": 15895
+    },
+    {
+      "epoch": 31.792,
+      "grad_norm": 1.139143705368042,
+      "learning_rate": 2e-05,
+      "loss": 0.04034952,
+      "step": 15896
+    },
+    {
+      "epoch": 31.794,
+      "grad_norm": 1.1100138425827026,
+      "learning_rate": 2e-05,
+      "loss": 0.03606164,
+      "step": 15897
+    },
+    {
+      "epoch": 31.796,
+      "grad_norm": 1.0225300788879395,
+      "learning_rate": 2e-05,
+      "loss": 0.03408796,
+      "step": 15898
+    },
+    {
+      "epoch": 31.798000000000002,
+      "grad_norm": 1.2668366432189941,
+      "learning_rate": 2e-05,
+      "loss": 0.05540526,
+      "step": 15899
+    },
+    {
+      "epoch": 31.8,
+      "grad_norm": 1.2877216339111328,
+      "learning_rate": 2e-05,
+      "loss": 0.05548783,
+      "step": 15900
+    },
+    {
+      "epoch": 31.802,
+      "grad_norm": 1.4362618923187256,
+      "learning_rate": 2e-05,
+      "loss": 0.0343156,
+      "step": 15901
+    },
+    {
+      "epoch": 31.804,
+      "grad_norm": 1.0215818881988525,
+      "learning_rate": 2e-05,
+      "loss": 0.03574908,
+      "step": 15902
+    },
+    {
+      "epoch": 31.806,
+      "grad_norm": 1.2409071922302246,
+      "learning_rate": 2e-05,
+      "loss": 0.04538013,
+      "step": 15903
+    },
+    {
+      "epoch": 31.808,
+      "grad_norm": 1.2912746667861938,
+      "learning_rate": 2e-05,
+      "loss": 0.0512158,
+      "step": 15904
+    },
+    {
+      "epoch": 31.81,
+      "grad_norm": 2.350969076156616,
+      "learning_rate": 2e-05,
+      "loss": 0.0492909,
+      "step": 15905
+    },
+    {
+      "epoch": 31.812,
+      "grad_norm": 1.299422264099121,
+      "learning_rate": 2e-05,
+      "loss": 0.05618146,
+      "step": 15906
+    },
+    {
+      "epoch": 31.814,
+      "grad_norm": 1.3270517587661743,
+      "learning_rate": 2e-05,
+      "loss": 0.03777631,
+      "step": 15907
+    },
+    {
+      "epoch": 31.816,
+      "grad_norm": 2.0297582149505615,
+      "learning_rate": 2e-05,
+      "loss": 0.05447558,
+      "step": 15908
+    },
+    {
+      "epoch": 31.818,
+      "grad_norm": 1.204034686088562,
+      "learning_rate": 2e-05,
+      "loss": 0.04712773,
+      "step": 15909
+    },
+    {
+      "epoch": 31.82,
+      "grad_norm": 1.8368512392044067,
+      "learning_rate": 2e-05,
+      "loss": 0.03878936,
+      "step": 15910
+    },
+    {
+      "epoch": 31.822,
+      "grad_norm": 1.3600157499313354,
+      "learning_rate": 2e-05,
+      "loss": 0.04160419,
+      "step": 15911
+    },
+    {
+      "epoch": 31.824,
+      "grad_norm": 0.9613247513771057,
+      "learning_rate": 2e-05,
+      "loss": 0.03210657,
+      "step": 15912
+    },
+    {
+      "epoch": 31.826,
+      "grad_norm": 1.2313897609710693,
+      "learning_rate": 2e-05,
+      "loss": 0.0382159,
+      "step": 15913
+    },
+    {
+      "epoch": 31.828,
+      "grad_norm": 1.0835821628570557,
+      "learning_rate": 2e-05,
+      "loss": 0.04776543,
+      "step": 15914
+    },
+    {
+      "epoch": 31.83,
+      "grad_norm": 1.1532397270202637,
+      "learning_rate": 2e-05,
+      "loss": 0.02791233,
+      "step": 15915
+    },
+    {
+      "epoch": 31.832,
+      "grad_norm": 1.6072558164596558,
+      "learning_rate": 2e-05,
+      "loss": 0.04267748,
+      "step": 15916
+    },
+    {
+      "epoch": 31.834,
+      "grad_norm": 1.2094991207122803,
+      "learning_rate": 2e-05,
+      "loss": 0.0439063,
+      "step": 15917
+    },
+    {
+      "epoch": 31.836,
+      "grad_norm": 1.9390913248062134,
+      "learning_rate": 2e-05,
+      "loss": 0.06620531,
+      "step": 15918
+    },
+    {
+      "epoch": 31.838,
+      "grad_norm": 2.274019479751587,
+      "learning_rate": 2e-05,
+      "loss": 0.04259209,
+      "step": 15919
+    },
+    {
+      "epoch": 31.84,
+      "grad_norm": 1.9428718090057373,
+      "learning_rate": 2e-05,
+      "loss": 0.04785924,
+      "step": 15920
+    },
+    {
+      "epoch": 31.842,
+      "grad_norm": 1.422423243522644,
+      "learning_rate": 2e-05,
+      "loss": 0.04497388,
+      "step": 15921
+    },
+    {
+      "epoch": 31.844,
+      "grad_norm": 1.053911805152893,
+      "learning_rate": 2e-05,
+      "loss": 0.03690197,
+      "step": 15922
+    },
+    {
+      "epoch": 31.846,
+      "grad_norm": 3.13558030128479,
+      "learning_rate": 2e-05,
+      "loss": 0.05500776,
+      "step": 15923
+    },
+    {
+      "epoch": 31.848,
+      "grad_norm": 1.275652527809143,
+      "learning_rate": 2e-05,
+      "loss": 0.04231162,
+      "step": 15924
+    },
+    {
+      "epoch": 31.85,
+      "grad_norm": 1.9165621995925903,
+      "learning_rate": 2e-05,
+      "loss": 0.0441243,
+      "step": 15925
+    },
+    {
+      "epoch": 31.852,
+      "grad_norm": 2.099443197250366,
+      "learning_rate": 2e-05,
+      "loss": 0.04373053,
+      "step": 15926
+    },
+    {
+      "epoch": 31.854,
+      "grad_norm": 1.221465826034546,
+      "learning_rate": 2e-05,
+      "loss": 0.04799281,
+      "step": 15927
+    },
+    {
+      "epoch": 31.856,
+      "grad_norm": 1.7947920560836792,
+      "learning_rate": 2e-05,
+      "loss": 0.04117496,
+      "step": 15928
+    },
+    {
+      "epoch": 31.858,
+      "grad_norm": 1.142938256263733,
+      "learning_rate": 2e-05,
+      "loss": 0.03635032,
+      "step": 15929
+    },
+    {
+      "epoch": 31.86,
+      "grad_norm": 1.0332616567611694,
+      "learning_rate": 2e-05,
+      "loss": 0.04042405,
+      "step": 15930
+    },
+    {
+      "epoch": 31.862,
+      "grad_norm": 1.3157538175582886,
+      "learning_rate": 2e-05,
+      "loss": 0.06223164,
+      "step": 15931
+    },
+    {
+      "epoch": 31.864,
+      "grad_norm": 1.207903265953064,
+      "learning_rate": 2e-05,
+      "loss": 0.05030291,
+      "step": 15932
+    },
+    {
+      "epoch": 31.866,
+      "grad_norm": 1.8377586603164673,
+      "learning_rate": 2e-05,
+      "loss": 0.05019252,
+      "step": 15933
+    },
+    {
+      "epoch": 31.868,
+      "grad_norm": 0.9512478709220886,
+      "learning_rate": 2e-05,
+      "loss": 0.03300637,
+      "step": 15934
+    },
+    {
+      "epoch": 31.87,
+      "grad_norm": 1.3328098058700562,
+      "learning_rate": 2e-05,
+      "loss": 0.04929174,
+      "step": 15935
+    },
+    {
+      "epoch": 31.872,
+      "grad_norm": 1.0890222787857056,
+      "learning_rate": 2e-05,
+      "loss": 0.04026979,
+      "step": 15936
+    },
+    {
+      "epoch": 31.874,
+      "grad_norm": 0.9871217012405396,
+      "learning_rate": 2e-05,
+      "loss": 0.03689782,
+      "step": 15937
+    },
+    {
+      "epoch": 31.876,
+      "grad_norm": 1.0311745405197144,
+      "learning_rate": 2e-05,
+      "loss": 0.03890628,
+      "step": 15938
+    },
+    {
+      "epoch": 31.878,
+      "grad_norm": 1.252621054649353,
+      "learning_rate": 2e-05,
+      "loss": 0.05022133,
+      "step": 15939
+    },
+    {
+      "epoch": 31.88,
+      "grad_norm": 0.8780187964439392,
+      "learning_rate": 2e-05,
+      "loss": 0.02962154,
+      "step": 15940
+    },
+    {
+      "epoch": 31.882,
+      "grad_norm": 1.3588930368423462,
+      "learning_rate": 2e-05,
+      "loss": 0.04873728,
+      "step": 15941
+    },
+    {
+      "epoch": 31.884,
+      "grad_norm": 1.4300963878631592,
+      "learning_rate": 2e-05,
+      "loss": 0.05331673,
+      "step": 15942
+    },
+    {
+      "epoch": 31.886,
+      "grad_norm": 1.7596718072891235,
+      "learning_rate": 2e-05,
+      "loss": 0.05031583,
+      "step": 15943
+    },
+    {
+      "epoch": 31.888,
+      "grad_norm": 1.7224795818328857,
+      "learning_rate": 2e-05,
+      "loss": 0.05422819,
+      "step": 15944
+    },
+    {
+      "epoch": 31.89,
+      "grad_norm": 1.7030835151672363,
+      "learning_rate": 2e-05,
+      "loss": 0.05407435,
+      "step": 15945
+    },
+    {
+      "epoch": 31.892,
+      "grad_norm": 0.9744284749031067,
+      "learning_rate": 2e-05,
+      "loss": 0.03288863,
+      "step": 15946
+    },
+    {
+      "epoch": 31.894,
+      "grad_norm": 1.9340544939041138,
+      "learning_rate": 2e-05,
+      "loss": 0.04774361,
+      "step": 15947
+    },
+    {
+      "epoch": 31.896,
+      "grad_norm": 1.3208523988723755,
+      "learning_rate": 2e-05,
+      "loss": 0.02824499,
+      "step": 15948
+    },
+    {
+      "epoch": 31.898,
+      "grad_norm": 1.0506129264831543,
+      "learning_rate": 2e-05,
+      "loss": 0.03759496,
+      "step": 15949
+    },
+    {
+      "epoch": 31.9,
+      "grad_norm": 1.000072717666626,
+      "learning_rate": 2e-05,
+      "loss": 0.03729441,
+      "step": 15950
+    },
+    {
+      "epoch": 31.902,
+      "grad_norm": 1.8239920139312744,
+      "learning_rate": 2e-05,
+      "loss": 0.0503429,
+      "step": 15951
+    },
+    {
+      "epoch": 31.904,
+      "grad_norm": 0.9123238921165466,
+      "learning_rate": 2e-05,
+      "loss": 0.04094257,
+      "step": 15952
+    },
+    {
+      "epoch": 31.906,
+      "grad_norm": 2.360563278198242,
+      "learning_rate": 2e-05,
+      "loss": 0.04889352,
+      "step": 15953
+    },
+    {
+      "epoch": 31.908,
+      "grad_norm": 1.2828713655471802,
+      "learning_rate": 2e-05,
+      "loss": 0.04768791,
+      "step": 15954
+    },
+    {
+      "epoch": 31.91,
+      "grad_norm": 1.2418571710586548,
+      "learning_rate": 2e-05,
+      "loss": 0.04087563,
+      "step": 15955
+    },
+    {
+      "epoch": 31.912,
+      "grad_norm": 2.2912075519561768,
+      "learning_rate": 2e-05,
+      "loss": 0.05685142,
+      "step": 15956
+    },
+    {
+      "epoch": 31.914,
+      "grad_norm": 0.9731912612915039,
+      "learning_rate": 2e-05,
+      "loss": 0.03105514,
+      "step": 15957
+    },
+    {
+      "epoch": 31.916,
+      "grad_norm": 1.03424072265625,
+      "learning_rate": 2e-05,
+      "loss": 0.0377789,
+      "step": 15958
+    },
+    {
+      "epoch": 31.918,
+      "grad_norm": 1.2611950635910034,
+      "learning_rate": 2e-05,
+      "loss": 0.04463454,
+      "step": 15959
+    },
+    {
+      "epoch": 31.92,
+      "grad_norm": 1.9684666395187378,
+      "learning_rate": 2e-05,
+      "loss": 0.05295739,
+      "step": 15960
+    },
+    {
+      "epoch": 31.922,
+      "grad_norm": 1.3407577276229858,
+      "learning_rate": 2e-05,
+      "loss": 0.04236686,
+      "step": 15961
+    },
+    {
+      "epoch": 31.924,
+      "grad_norm": 1.2318034172058105,
+      "learning_rate": 2e-05,
+      "loss": 0.03647027,
+      "step": 15962
+    },
+    {
+      "epoch": 31.926,
+      "grad_norm": 1.745802402496338,
+      "learning_rate": 2e-05,
+      "loss": 0.04532122,
+      "step": 15963
+    },
+    {
+      "epoch": 31.928,
+      "grad_norm": 1.087468147277832,
+      "learning_rate": 2e-05,
+      "loss": 0.04168069,
+      "step": 15964
+    },
+    {
+      "epoch": 31.93,
+      "grad_norm": 2.0053272247314453,
+      "learning_rate": 2e-05,
+      "loss": 0.03822828,
+      "step": 15965
+    },
+    {
+      "epoch": 31.932,
+      "grad_norm": 1.0389565229415894,
+      "learning_rate": 2e-05,
+      "loss": 0.03328662,
+      "step": 15966
+    },
+    {
+      "epoch": 31.934,
+      "grad_norm": 1.394301414489746,
+      "learning_rate": 2e-05,
+      "loss": 0.05171059,
+      "step": 15967
+    },
+    {
+      "epoch": 31.936,
+      "grad_norm": 1.4605191946029663,
+      "learning_rate": 2e-05,
+      "loss": 0.0633463,
+      "step": 15968
+    },
+    {
+      "epoch": 31.938,
+      "grad_norm": 1.1026382446289062,
+      "learning_rate": 2e-05,
+      "loss": 0.04084253,
+      "step": 15969
+    },
+    {
+      "epoch": 31.94,
+      "grad_norm": 1.5898246765136719,
+      "learning_rate": 2e-05,
+      "loss": 0.06907326,
+      "step": 15970
+    },
+    {
+      "epoch": 31.942,
+      "grad_norm": 1.0612562894821167,
+      "learning_rate": 2e-05,
+      "loss": 0.03758101,
+      "step": 15971
+    },
+    {
+      "epoch": 31.944,
+      "grad_norm": 1.5062930583953857,
+      "learning_rate": 2e-05,
+      "loss": 0.03778771,
+      "step": 15972
+    },
+    {
+      "epoch": 31.946,
+      "grad_norm": 1.8301286697387695,
+      "learning_rate": 2e-05,
+      "loss": 0.03591305,
+      "step": 15973
+    },
+    {
+      "epoch": 31.948,
+      "grad_norm": 1.2562285661697388,
+      "learning_rate": 2e-05,
+      "loss": 0.03053888,
+      "step": 15974
+    },
+    {
+      "epoch": 31.95,
+      "grad_norm": 1.2209889888763428,
+      "learning_rate": 2e-05,
+      "loss": 0.05101029,
+      "step": 15975
+    },
+    {
+      "epoch": 31.951999999999998,
+      "grad_norm": 1.7810455560684204,
+      "learning_rate": 2e-05,
+      "loss": 0.04453781,
+      "step": 15976
+    },
+    {
+      "epoch": 31.954,
+      "grad_norm": 1.6434965133666992,
+      "learning_rate": 2e-05,
+      "loss": 0.0371579,
+      "step": 15977
+    },
+    {
+      "epoch": 31.956,
+      "grad_norm": 1.207597255706787,
+      "learning_rate": 2e-05,
+      "loss": 0.03318027,
+      "step": 15978
+    },
+    {
+      "epoch": 31.958,
+      "grad_norm": 1.3328630924224854,
+      "learning_rate": 2e-05,
+      "loss": 0.05412392,
+      "step": 15979
+    },
+    {
+      "epoch": 31.96,
+      "grad_norm": 1.0309226512908936,
+      "learning_rate": 2e-05,
+      "loss": 0.03783122,
+      "step": 15980
+    },
+    {
+      "epoch": 31.962,
+      "grad_norm": 1.042348861694336,
+      "learning_rate": 2e-05,
+      "loss": 0.03998961,
+      "step": 15981
+    },
+    {
+      "epoch": 31.964,
+      "grad_norm": 1.2534490823745728,
+      "learning_rate": 2e-05,
+      "loss": 0.03557429,
+      "step": 15982
+    },
+    {
+      "epoch": 31.966,
+      "grad_norm": 2.982203722000122,
+      "learning_rate": 2e-05,
+      "loss": 0.06171292,
+      "step": 15983
+    },
+    {
+      "epoch": 31.968,
+      "grad_norm": 1.2755846977233887,
+      "learning_rate": 2e-05,
+      "loss": 0.03844975,
+      "step": 15984
+    },
+    {
+      "epoch": 31.97,
+      "grad_norm": 1.0132505893707275,
+      "learning_rate": 2e-05,
+      "loss": 0.03931327,
+      "step": 15985
+    },
+    {
+      "epoch": 31.972,
+      "grad_norm": 1.2777570486068726,
+      "learning_rate": 2e-05,
+      "loss": 0.04398004,
+      "step": 15986
+    },
+    {
+      "epoch": 31.974,
+      "grad_norm": 1.2927149534225464,
+      "learning_rate": 2e-05,
+      "loss": 0.05497951,
+      "step": 15987
+    },
+    {
+      "epoch": 31.976,
+      "grad_norm": 0.8288196921348572,
+      "learning_rate": 2e-05,
+      "loss": 0.02825504,
+      "step": 15988
+    },
+    {
+      "epoch": 31.978,
+      "grad_norm": 1.1850969791412354,
+      "learning_rate": 2e-05,
+      "loss": 0.03954066,
+      "step": 15989
+    },
+    {
+      "epoch": 31.98,
+      "grad_norm": 0.8509149551391602,
+      "learning_rate": 2e-05,
+      "loss": 0.02893381,
+      "step": 15990
+    },
+    {
+      "epoch": 31.982,
+      "grad_norm": 1.2214093208312988,
+      "learning_rate": 2e-05,
+      "loss": 0.04563006,
+      "step": 15991
+    },
+    {
+      "epoch": 31.984,
+      "grad_norm": 1.3428093194961548,
+      "learning_rate": 2e-05,
+      "loss": 0.05709552,
+      "step": 15992
+    },
+    {
+      "epoch": 31.986,
+      "grad_norm": 1.2619773149490356,
+      "learning_rate": 2e-05,
+      "loss": 0.04833028,
+      "step": 15993
+    },
+    {
+      "epoch": 31.988,
+      "grad_norm": 1.4030076265335083,
+      "learning_rate": 2e-05,
+      "loss": 0.04587878,
+      "step": 15994
+    },
+    {
+      "epoch": 31.99,
+      "grad_norm": 1.113511085510254,
+      "learning_rate": 2e-05,
+      "loss": 0.03980121,
+      "step": 15995
+    },
+    {
+      "epoch": 31.992,
+      "grad_norm": 1.087285041809082,
+      "learning_rate": 2e-05,
+      "loss": 0.04423406,
+      "step": 15996
+    },
+    {
+      "epoch": 31.994,
+      "grad_norm": 1.2081905603408813,
+      "learning_rate": 2e-05,
+      "loss": 0.03930626,
+      "step": 15997
+    },
+    {
+      "epoch": 31.996,
+      "grad_norm": 0.9925308227539062,
+      "learning_rate": 2e-05,
+      "loss": 0.03966329,
+      "step": 15998
+    },
+    {
+      "epoch": 31.998,
+      "grad_norm": 1.5078144073486328,
+      "learning_rate": 2e-05,
+      "loss": 0.06151321,
+      "step": 15999
+    },
+    {
+      "epoch": 32.0,
+      "grad_norm": 1.641327142715454,
+      "learning_rate": 2e-05,
+      "loss": 0.05479438,
+      "step": 16000
+    },
+    {
+      "epoch": 32.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.992,
+        "AngleClassification_2": 0.998,
+        "AngleClassification_3": 0.9780439121756487,
+        "Equal_1": 0.998,
+        "Equal_2": 0.9780439121756487,
+        "Equal_3": 0.9880239520958084,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.998003992015968,
+        "Parallel_1": 0.9939879759519038,
+        "Parallel_2": 0.9919839679358717,
+        "Parallel_3": 0.99,
+        "Perpendicular_1": 0.996,
+        "Perpendicular_2": 0.982,
+        "Perpendicular_3": 0.8767535070140281,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9956666666666667,
+        "PointLiesOnCircle_3": 0.9876,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9979959919839679,
+        "PointLiesOnLine_3": 0.9920159680638723
+      },
+      "eval_runtime": 320.2305,
+      "eval_samples_per_second": 32.789,
+      "eval_steps_per_second": 0.656,
+      "step": 16000
+    },
+    {
+      "epoch": 32.002,
+      "grad_norm": 1.0825588703155518,
+      "learning_rate": 2e-05,
+      "loss": 0.04043002,
+      "step": 16001
+    },
+    {
+      "epoch": 32.004,
+      "grad_norm": 1.116982102394104,
+      "learning_rate": 2e-05,
+      "loss": 0.0345992,
+      "step": 16002
+    },
+    {
+      "epoch": 32.006,
+      "grad_norm": 1.4669443368911743,
+      "learning_rate": 2e-05,
+      "loss": 0.05243903,
+      "step": 16003
+    },
+    {
+      "epoch": 32.008,
+      "grad_norm": 1.504705548286438,
+      "learning_rate": 2e-05,
+      "loss": 0.04028647,
+      "step": 16004
+    },
+    {
+      "epoch": 32.01,
+      "grad_norm": 1.2776727676391602,
+      "learning_rate": 2e-05,
+      "loss": 0.03870967,
+      "step": 16005
+    },
+    {
+      "epoch": 32.012,
+      "grad_norm": 2.4159090518951416,
+      "learning_rate": 2e-05,
+      "loss": 0.04641645,
+      "step": 16006
+    },
+    {
+      "epoch": 32.014,
+      "grad_norm": 1.1506953239440918,
+      "learning_rate": 2e-05,
+      "loss": 0.03702936,
+      "step": 16007
+    },
+    {
+      "epoch": 32.016,
+      "grad_norm": 1.2919611930847168,
+      "learning_rate": 2e-05,
+      "loss": 0.05337862,
+      "step": 16008
+    },
+    {
+      "epoch": 32.018,
+      "grad_norm": 0.9426947832107544,
+      "learning_rate": 2e-05,
+      "loss": 0.02725463,
+      "step": 16009
+    },
+    {
+      "epoch": 32.02,
+      "grad_norm": 1.00312077999115,
+      "learning_rate": 2e-05,
+      "loss": 0.04260667,
+      "step": 16010
+    },
+    {
+      "epoch": 32.022,
+      "grad_norm": 1.1375038623809814,
+      "learning_rate": 2e-05,
+      "loss": 0.03781604,
+      "step": 16011
+    },
+    {
+      "epoch": 32.024,
+      "grad_norm": 1.3393661975860596,
+      "learning_rate": 2e-05,
+      "loss": 0.04303484,
+      "step": 16012
+    },
+    {
+      "epoch": 32.026,
+      "grad_norm": 1.1762498617172241,
+      "learning_rate": 2e-05,
+      "loss": 0.04548337,
+      "step": 16013
+    },
+    {
+      "epoch": 32.028,
+      "grad_norm": 1.3804402351379395,
+      "learning_rate": 2e-05,
+      "loss": 0.0544548,
+      "step": 16014
+    },
+    {
+      "epoch": 32.03,
+      "grad_norm": 1.5973312854766846,
+      "learning_rate": 2e-05,
+      "loss": 0.0446333,
+      "step": 16015
+    },
+    {
+      "epoch": 32.032,
+      "grad_norm": 1.1906533241271973,
+      "learning_rate": 2e-05,
+      "loss": 0.0428127,
+      "step": 16016
+    },
+    {
+      "epoch": 32.034,
+      "grad_norm": 1.367288589477539,
+      "learning_rate": 2e-05,
+      "loss": 0.04731005,
+      "step": 16017
+    },
+    {
+      "epoch": 32.036,
+      "grad_norm": 1.3971643447875977,
+      "learning_rate": 2e-05,
+      "loss": 0.062555,
+      "step": 16018
+    },
+    {
+      "epoch": 32.038,
+      "grad_norm": 1.3580145835876465,
+      "learning_rate": 2e-05,
+      "loss": 0.04744005,
+      "step": 16019
+    },
+    {
+      "epoch": 32.04,
+      "grad_norm": 1.9001810550689697,
+      "learning_rate": 2e-05,
+      "loss": 0.03246989,
+      "step": 16020
+    },
+    {
+      "epoch": 32.042,
+      "grad_norm": 1.1311832666397095,
+      "learning_rate": 2e-05,
+      "loss": 0.04354143,
+      "step": 16021
+    },
+    {
+      "epoch": 32.044,
+      "grad_norm": 1.2986774444580078,
+      "learning_rate": 2e-05,
+      "loss": 0.04035509,
+      "step": 16022
+    },
+    {
+      "epoch": 32.046,
+      "grad_norm": 1.015791654586792,
+      "learning_rate": 2e-05,
+      "loss": 0.03449727,
+      "step": 16023
+    },
+    {
+      "epoch": 32.048,
+      "grad_norm": 1.2315489053726196,
+      "learning_rate": 2e-05,
+      "loss": 0.04751674,
+      "step": 16024
+    },
+    {
+      "epoch": 32.05,
+      "grad_norm": 1.6999454498291016,
+      "learning_rate": 2e-05,
+      "loss": 0.04061841,
+      "step": 16025
+    },
+    {
+      "epoch": 32.052,
+      "grad_norm": 1.1011244058609009,
+      "learning_rate": 2e-05,
+      "loss": 0.04402358,
+      "step": 16026
+    },
+    {
+      "epoch": 32.054,
+      "grad_norm": 0.9123277068138123,
+      "learning_rate": 2e-05,
+      "loss": 0.03008776,
+      "step": 16027
+    },
+    {
+      "epoch": 32.056,
+      "grad_norm": 1.5880179405212402,
+      "learning_rate": 2e-05,
+      "loss": 0.05666232,
+      "step": 16028
+    },
+    {
+      "epoch": 32.058,
+      "grad_norm": 1.4386612176895142,
+      "learning_rate": 2e-05,
+      "loss": 0.04437473,
+      "step": 16029
+    },
+    {
+      "epoch": 32.06,
+      "grad_norm": 1.1378834247589111,
+      "learning_rate": 2e-05,
+      "loss": 0.04078399,
+      "step": 16030
+    },
+    {
+      "epoch": 32.062,
+      "grad_norm": 1.1456459760665894,
+      "learning_rate": 2e-05,
+      "loss": 0.05264393,
+      "step": 16031
+    },
+    {
+      "epoch": 32.064,
+      "grad_norm": 1.2216919660568237,
+      "learning_rate": 2e-05,
+      "loss": 0.04822063,
+      "step": 16032
+    },
+    {
+      "epoch": 32.066,
+      "grad_norm": 1.7245386838912964,
+      "learning_rate": 2e-05,
+      "loss": 0.05039418,
+      "step": 16033
+    },
+    {
+      "epoch": 32.068,
+      "grad_norm": 1.1855204105377197,
+      "learning_rate": 2e-05,
+      "loss": 0.02989644,
+      "step": 16034
+    },
+    {
+      "epoch": 32.07,
+      "grad_norm": 1.0187755823135376,
+      "learning_rate": 2e-05,
+      "loss": 0.03046339,
+      "step": 16035
+    },
+    {
+      "epoch": 32.072,
+      "grad_norm": 1.0775607824325562,
+      "learning_rate": 2e-05,
+      "loss": 0.05659611,
+      "step": 16036
+    },
+    {
+      "epoch": 32.074,
+      "grad_norm": 1.234232783317566,
+      "learning_rate": 2e-05,
+      "loss": 0.0332603,
+      "step": 16037
+    },
+    {
+      "epoch": 32.076,
+      "grad_norm": 1.037575364112854,
+      "learning_rate": 2e-05,
+      "loss": 0.03690245,
+      "step": 16038
+    },
+    {
+      "epoch": 32.078,
+      "grad_norm": 1.1662631034851074,
+      "learning_rate": 2e-05,
+      "loss": 0.04369701,
+      "step": 16039
+    },
+    {
+      "epoch": 32.08,
+      "grad_norm": 1.172985553741455,
+      "learning_rate": 2e-05,
+      "loss": 0.03649969,
+      "step": 16040
+    },
+    {
+      "epoch": 32.082,
+      "grad_norm": 1.399276614189148,
+      "learning_rate": 2e-05,
+      "loss": 0.04782386,
+      "step": 16041
+    },
+    {
+      "epoch": 32.084,
+      "grad_norm": 1.089415431022644,
+      "learning_rate": 2e-05,
+      "loss": 0.0528977,
+      "step": 16042
+    },
+    {
+      "epoch": 32.086,
+      "grad_norm": 1.217584252357483,
+      "learning_rate": 2e-05,
+      "loss": 0.04556941,
+      "step": 16043
+    },
+    {
+      "epoch": 32.088,
+      "grad_norm": 1.1931408643722534,
+      "learning_rate": 2e-05,
+      "loss": 0.04738794,
+      "step": 16044
+    },
+    {
+      "epoch": 32.09,
+      "grad_norm": 1.0571660995483398,
+      "learning_rate": 2e-05,
+      "loss": 0.03878347,
+      "step": 16045
+    },
+    {
+      "epoch": 32.092,
+      "grad_norm": 1.4781181812286377,
+      "learning_rate": 2e-05,
+      "loss": 0.04194687,
+      "step": 16046
+    },
+    {
+      "epoch": 32.094,
+      "grad_norm": 4.250499248504639,
+      "learning_rate": 2e-05,
+      "loss": 0.03575663,
+      "step": 16047
+    },
+    {
+      "epoch": 32.096,
+      "grad_norm": 2.5705862045288086,
+      "learning_rate": 2e-05,
+      "loss": 0.04239442,
+      "step": 16048
+    },
+    {
+      "epoch": 32.098,
+      "grad_norm": 0.9903239607810974,
+      "learning_rate": 2e-05,
+      "loss": 0.04511203,
+      "step": 16049
+    },
+    {
+      "epoch": 32.1,
+      "grad_norm": 1.8635953664779663,
+      "learning_rate": 2e-05,
+      "loss": 0.05268551,
+      "step": 16050
+    },
+    {
+      "epoch": 32.102,
+      "grad_norm": 2.417490005493164,
+      "learning_rate": 2e-05,
+      "loss": 0.0534266,
+      "step": 16051
+    },
+    {
+      "epoch": 32.104,
+      "grad_norm": 1.4200366735458374,
+      "learning_rate": 2e-05,
+      "loss": 0.05390621,
+      "step": 16052
+    },
+    {
+      "epoch": 32.106,
+      "grad_norm": 1.796728491783142,
+      "learning_rate": 2e-05,
+      "loss": 0.05886037,
+      "step": 16053
+    },
+    {
+      "epoch": 32.108,
+      "grad_norm": 1.967190146446228,
+      "learning_rate": 2e-05,
+      "loss": 0.04227813,
+      "step": 16054
+    },
+    {
+      "epoch": 32.11,
+      "grad_norm": 1.7734386920928955,
+      "learning_rate": 2e-05,
+      "loss": 0.03885096,
+      "step": 16055
+    },
+    {
+      "epoch": 32.112,
+      "grad_norm": 1.0239137411117554,
+      "learning_rate": 2e-05,
+      "loss": 0.03256997,
+      "step": 16056
+    },
+    {
+      "epoch": 32.114,
+      "grad_norm": 1.1545599699020386,
+      "learning_rate": 2e-05,
+      "loss": 0.04887954,
+      "step": 16057
+    },
+    {
+      "epoch": 32.116,
+      "grad_norm": 2.557694673538208,
+      "learning_rate": 2e-05,
+      "loss": 0.04597983,
+      "step": 16058
+    },
+    {
+      "epoch": 32.118,
+      "grad_norm": 2.367849826812744,
+      "learning_rate": 2e-05,
+      "loss": 0.07052769,
+      "step": 16059
+    },
+    {
+      "epoch": 32.12,
+      "grad_norm": 1.2989052534103394,
+      "learning_rate": 2e-05,
+      "loss": 0.05320881,
+      "step": 16060
+    },
+    {
+      "epoch": 32.122,
+      "grad_norm": 1.6988813877105713,
+      "learning_rate": 2e-05,
+      "loss": 0.05019864,
+      "step": 16061
+    },
+    {
+      "epoch": 32.124,
+      "grad_norm": 1.2246428728103638,
+      "learning_rate": 2e-05,
+      "loss": 0.03273758,
+      "step": 16062
+    },
+    {
+      "epoch": 32.126,
+      "grad_norm": 0.8918790221214294,
+      "learning_rate": 2e-05,
+      "loss": 0.02186277,
+      "step": 16063
+    },
+    {
+      "epoch": 32.128,
+      "grad_norm": 0.9243330359458923,
+      "learning_rate": 2e-05,
+      "loss": 0.02869822,
+      "step": 16064
+    },
+    {
+      "epoch": 32.13,
+      "grad_norm": 0.9982576370239258,
+      "learning_rate": 2e-05,
+      "loss": 0.03740531,
+      "step": 16065
+    },
+    {
+      "epoch": 32.132,
+      "grad_norm": 1.0650466680526733,
+      "learning_rate": 2e-05,
+      "loss": 0.03073342,
+      "step": 16066
+    },
+    {
+      "epoch": 32.134,
+      "grad_norm": 2.1163761615753174,
+      "learning_rate": 2e-05,
+      "loss": 0.03885228,
+      "step": 16067
+    },
+    {
+      "epoch": 32.136,
+      "grad_norm": 0.9434223771095276,
+      "learning_rate": 2e-05,
+      "loss": 0.03199326,
+      "step": 16068
+    },
+    {
+      "epoch": 32.138,
+      "grad_norm": 0.8779048919677734,
+      "learning_rate": 2e-05,
+      "loss": 0.02758794,
+      "step": 16069
+    },
+    {
+      "epoch": 32.14,
+      "grad_norm": 1.155031442642212,
+      "learning_rate": 2e-05,
+      "loss": 0.04074682,
+      "step": 16070
+    },
+    {
+      "epoch": 32.142,
+      "grad_norm": 1.140641450881958,
+      "learning_rate": 2e-05,
+      "loss": 0.04276387,
+      "step": 16071
+    },
+    {
+      "epoch": 32.144,
+      "grad_norm": 0.932158887386322,
+      "learning_rate": 2e-05,
+      "loss": 0.02548703,
+      "step": 16072
+    },
+    {
+      "epoch": 32.146,
+      "grad_norm": 1.3973239660263062,
+      "learning_rate": 2e-05,
+      "loss": 0.05437903,
+      "step": 16073
+    },
+    {
+      "epoch": 32.148,
+      "grad_norm": 1.164240837097168,
+      "learning_rate": 2e-05,
+      "loss": 0.03863488,
+      "step": 16074
+    },
+    {
+      "epoch": 32.15,
+      "grad_norm": 1.162251591682434,
+      "learning_rate": 2e-05,
+      "loss": 0.05012304,
+      "step": 16075
+    },
+    {
+      "epoch": 32.152,
+      "grad_norm": 1.5069862604141235,
+      "learning_rate": 2e-05,
+      "loss": 0.04389899,
+      "step": 16076
+    },
+    {
+      "epoch": 32.154,
+      "grad_norm": 1.4252487421035767,
+      "learning_rate": 2e-05,
+      "loss": 0.04979219,
+      "step": 16077
+    },
+    {
+      "epoch": 32.156,
+      "grad_norm": 1.2503894567489624,
+      "learning_rate": 2e-05,
+      "loss": 0.04105882,
+      "step": 16078
+    },
+    {
+      "epoch": 32.158,
+      "grad_norm": 1.1806787252426147,
+      "learning_rate": 2e-05,
+      "loss": 0.04621601,
+      "step": 16079
+    },
+    {
+      "epoch": 32.16,
+      "grad_norm": 1.0560612678527832,
+      "learning_rate": 2e-05,
+      "loss": 0.03700335,
+      "step": 16080
+    },
+    {
+      "epoch": 32.162,
+      "grad_norm": 1.3735400438308716,
+      "learning_rate": 2e-05,
+      "loss": 0.04111753,
+      "step": 16081
+    },
+    {
+      "epoch": 32.164,
+      "grad_norm": 1.053409457206726,
+      "learning_rate": 2e-05,
+      "loss": 0.0357594,
+      "step": 16082
+    },
+    {
+      "epoch": 32.166,
+      "grad_norm": 2.5766215324401855,
+      "learning_rate": 2e-05,
+      "loss": 0.05252399,
+      "step": 16083
+    },
+    {
+      "epoch": 32.168,
+      "grad_norm": 1.9363195896148682,
+      "learning_rate": 2e-05,
+      "loss": 0.05619735,
+      "step": 16084
+    },
+    {
+      "epoch": 32.17,
+      "grad_norm": 0.9756081104278564,
+      "learning_rate": 2e-05,
+      "loss": 0.03354892,
+      "step": 16085
+    },
+    {
+      "epoch": 32.172,
+      "grad_norm": 0.9970953464508057,
+      "learning_rate": 2e-05,
+      "loss": 0.03117803,
+      "step": 16086
+    },
+    {
+      "epoch": 32.174,
+      "grad_norm": 1.3192036151885986,
+      "learning_rate": 2e-05,
+      "loss": 0.04156514,
+      "step": 16087
+    },
+    {
+      "epoch": 32.176,
+      "grad_norm": 1.3369141817092896,
+      "learning_rate": 2e-05,
+      "loss": 0.04897435,
+      "step": 16088
+    },
+    {
+      "epoch": 32.178,
+      "grad_norm": 1.4819189310073853,
+      "learning_rate": 2e-05,
+      "loss": 0.04706377,
+      "step": 16089
+    },
+    {
+      "epoch": 32.18,
+      "grad_norm": 1.1336413621902466,
+      "learning_rate": 2e-05,
+      "loss": 0.03167891,
+      "step": 16090
+    },
+    {
+      "epoch": 32.182,
+      "grad_norm": 1.1428834199905396,
+      "learning_rate": 2e-05,
+      "loss": 0.02547898,
+      "step": 16091
+    },
+    {
+      "epoch": 32.184,
+      "grad_norm": 1.2145164012908936,
+      "learning_rate": 2e-05,
+      "loss": 0.03935249,
+      "step": 16092
+    },
+    {
+      "epoch": 32.186,
+      "grad_norm": 1.4651552438735962,
+      "learning_rate": 2e-05,
+      "loss": 0.03130937,
+      "step": 16093
+    },
+    {
+      "epoch": 32.188,
+      "grad_norm": 1.480681300163269,
+      "learning_rate": 2e-05,
+      "loss": 0.03741454,
+      "step": 16094
+    },
+    {
+      "epoch": 32.19,
+      "grad_norm": 2.146885633468628,
+      "learning_rate": 2e-05,
+      "loss": 0.05036818,
+      "step": 16095
+    },
+    {
+      "epoch": 32.192,
+      "grad_norm": 1.1545674800872803,
+      "learning_rate": 2e-05,
+      "loss": 0.03278602,
+      "step": 16096
+    },
+    {
+      "epoch": 32.194,
+      "grad_norm": 0.9393602013587952,
+      "learning_rate": 2e-05,
+      "loss": 0.03584001,
+      "step": 16097
+    },
+    {
+      "epoch": 32.196,
+      "grad_norm": 1.2177561521530151,
+      "learning_rate": 2e-05,
+      "loss": 0.03207495,
+      "step": 16098
+    },
+    {
+      "epoch": 32.198,
+      "grad_norm": 1.0664947032928467,
+      "learning_rate": 2e-05,
+      "loss": 0.03970977,
+      "step": 16099
+    },
+    {
+      "epoch": 32.2,
+      "grad_norm": 1.6877775192260742,
+      "learning_rate": 2e-05,
+      "loss": 0.04805655,
+      "step": 16100
+    },
+    {
+      "epoch": 32.202,
+      "grad_norm": 1.9446738958358765,
+      "learning_rate": 2e-05,
+      "loss": 0.06819682,
+      "step": 16101
+    },
+    {
+      "epoch": 32.204,
+      "grad_norm": 1.219976782798767,
+      "learning_rate": 2e-05,
+      "loss": 0.04143453,
+      "step": 16102
+    },
+    {
+      "epoch": 32.206,
+      "grad_norm": 1.203502893447876,
+      "learning_rate": 2e-05,
+      "loss": 0.03769158,
+      "step": 16103
+    },
+    {
+      "epoch": 32.208,
+      "grad_norm": 1.0326348543167114,
+      "learning_rate": 2e-05,
+      "loss": 0.02815155,
+      "step": 16104
+    },
+    {
+      "epoch": 32.21,
+      "grad_norm": 0.9759824872016907,
+      "learning_rate": 2e-05,
+      "loss": 0.03706375,
+      "step": 16105
+    },
+    {
+      "epoch": 32.212,
+      "grad_norm": 1.9715107679367065,
+      "learning_rate": 2e-05,
+      "loss": 0.0670663,
+      "step": 16106
+    },
+    {
+      "epoch": 32.214,
+      "grad_norm": 2.088665723800659,
+      "learning_rate": 2e-05,
+      "loss": 0.04498452,
+      "step": 16107
+    },
+    {
+      "epoch": 32.216,
+      "grad_norm": 1.9537535905838013,
+      "learning_rate": 2e-05,
+      "loss": 0.06290226,
+      "step": 16108
+    },
+    {
+      "epoch": 32.218,
+      "grad_norm": 1.1504062414169312,
+      "learning_rate": 2e-05,
+      "loss": 0.0422533,
+      "step": 16109
+    },
+    {
+      "epoch": 32.22,
+      "grad_norm": 1.2755309343338013,
+      "learning_rate": 2e-05,
+      "loss": 0.05198646,
+      "step": 16110
+    },
+    {
+      "epoch": 32.222,
+      "grad_norm": 1.7408345937728882,
+      "learning_rate": 2e-05,
+      "loss": 0.03385472,
+      "step": 16111
+    },
+    {
+      "epoch": 32.224,
+      "grad_norm": 1.2028616666793823,
+      "learning_rate": 2e-05,
+      "loss": 0.03718933,
+      "step": 16112
+    },
+    {
+      "epoch": 32.226,
+      "grad_norm": 1.2710516452789307,
+      "learning_rate": 2e-05,
+      "loss": 0.04598017,
+      "step": 16113
+    },
+    {
+      "epoch": 32.228,
+      "grad_norm": 1.0447970628738403,
+      "learning_rate": 2e-05,
+      "loss": 0.04060001,
+      "step": 16114
+    },
+    {
+      "epoch": 32.23,
+      "grad_norm": 1.267838478088379,
+      "learning_rate": 2e-05,
+      "loss": 0.06093229,
+      "step": 16115
+    },
+    {
+      "epoch": 32.232,
+      "grad_norm": 1.31565260887146,
+      "learning_rate": 2e-05,
+      "loss": 0.03904974,
+      "step": 16116
+    },
+    {
+      "epoch": 32.234,
+      "grad_norm": 1.2240655422210693,
+      "learning_rate": 2e-05,
+      "loss": 0.03763383,
+      "step": 16117
+    },
+    {
+      "epoch": 32.236,
+      "grad_norm": 0.9750972390174866,
+      "learning_rate": 2e-05,
+      "loss": 0.03341448,
+      "step": 16118
+    },
+    {
+      "epoch": 32.238,
+      "grad_norm": 1.330909013748169,
+      "learning_rate": 2e-05,
+      "loss": 0.04453826,
+      "step": 16119
+    },
+    {
+      "epoch": 32.24,
+      "grad_norm": 2.050285816192627,
+      "learning_rate": 2e-05,
+      "loss": 0.06148023,
+      "step": 16120
+    },
+    {
+      "epoch": 32.242,
+      "grad_norm": 1.2034765481948853,
+      "learning_rate": 2e-05,
+      "loss": 0.03787356,
+      "step": 16121
+    },
+    {
+      "epoch": 32.244,
+      "grad_norm": 2.573892831802368,
+      "learning_rate": 2e-05,
+      "loss": 0.0428374,
+      "step": 16122
+    },
+    {
+      "epoch": 32.246,
+      "grad_norm": 1.063158631324768,
+      "learning_rate": 2e-05,
+      "loss": 0.03356419,
+      "step": 16123
+    },
+    {
+      "epoch": 32.248,
+      "grad_norm": 1.1658393144607544,
+      "learning_rate": 2e-05,
+      "loss": 0.02797179,
+      "step": 16124
+    },
+    {
+      "epoch": 32.25,
+      "grad_norm": 1.3309680223464966,
+      "learning_rate": 2e-05,
+      "loss": 0.06812909,
+      "step": 16125
+    },
+    {
+      "epoch": 32.252,
+      "grad_norm": 0.8895295858383179,
+      "learning_rate": 2e-05,
+      "loss": 0.03186877,
+      "step": 16126
+    },
+    {
+      "epoch": 32.254,
+      "grad_norm": 1.42412269115448,
+      "learning_rate": 2e-05,
+      "loss": 0.05221164,
+      "step": 16127
+    },
+    {
+      "epoch": 32.256,
+      "grad_norm": 1.7970479726791382,
+      "learning_rate": 2e-05,
+      "loss": 0.04089088,
+      "step": 16128
+    },
+    {
+      "epoch": 32.258,
+      "grad_norm": 1.4141465425491333,
+      "learning_rate": 2e-05,
+      "loss": 0.03509438,
+      "step": 16129
+    },
+    {
+      "epoch": 32.26,
+      "grad_norm": 1.255109190940857,
+      "learning_rate": 2e-05,
+      "loss": 0.03642782,
+      "step": 16130
+    },
+    {
+      "epoch": 32.262,
+      "grad_norm": 2.4638772010803223,
+      "learning_rate": 2e-05,
+      "loss": 0.06095183,
+      "step": 16131
+    },
+    {
+      "epoch": 32.264,
+      "grad_norm": 1.527658224105835,
+      "learning_rate": 2e-05,
+      "loss": 0.03274827,
+      "step": 16132
+    },
+    {
+      "epoch": 32.266,
+      "grad_norm": 1.2536580562591553,
+      "learning_rate": 2e-05,
+      "loss": 0.0438434,
+      "step": 16133
+    },
+    {
+      "epoch": 32.268,
+      "grad_norm": 1.317441701889038,
+      "learning_rate": 2e-05,
+      "loss": 0.05264898,
+      "step": 16134
+    },
+    {
+      "epoch": 32.27,
+      "grad_norm": 1.0908536911010742,
+      "learning_rate": 2e-05,
+      "loss": 0.04722644,
+      "step": 16135
+    },
+    {
+      "epoch": 32.272,
+      "grad_norm": 1.3812543153762817,
+      "learning_rate": 2e-05,
+      "loss": 0.04213402,
+      "step": 16136
+    },
+    {
+      "epoch": 32.274,
+      "grad_norm": 1.2621177434921265,
+      "learning_rate": 2e-05,
+      "loss": 0.04359785,
+      "step": 16137
+    },
+    {
+      "epoch": 32.276,
+      "grad_norm": 1.1774916648864746,
+      "learning_rate": 2e-05,
+      "loss": 0.03530777,
+      "step": 16138
+    },
+    {
+      "epoch": 32.278,
+      "grad_norm": 1.1465319395065308,
+      "learning_rate": 2e-05,
+      "loss": 0.04779294,
+      "step": 16139
+    },
+    {
+      "epoch": 32.28,
+      "grad_norm": 1.1407089233398438,
+      "learning_rate": 2e-05,
+      "loss": 0.04244178,
+      "step": 16140
+    },
+    {
+      "epoch": 32.282,
+      "grad_norm": 1.1952414512634277,
+      "learning_rate": 2e-05,
+      "loss": 0.03308697,
+      "step": 16141
+    },
+    {
+      "epoch": 32.284,
+      "grad_norm": 1.1828652620315552,
+      "learning_rate": 2e-05,
+      "loss": 0.03380813,
+      "step": 16142
+    },
+    {
+      "epoch": 32.286,
+      "grad_norm": 2.2525599002838135,
+      "learning_rate": 2e-05,
+      "loss": 0.05679112,
+      "step": 16143
+    },
+    {
+      "epoch": 32.288,
+      "grad_norm": 2.428598642349243,
+      "learning_rate": 2e-05,
+      "loss": 0.05677919,
+      "step": 16144
+    },
+    {
+      "epoch": 32.29,
+      "grad_norm": 1.4035303592681885,
+      "learning_rate": 2e-05,
+      "loss": 0.0381352,
+      "step": 16145
+    },
+    {
+      "epoch": 32.292,
+      "grad_norm": 1.2325247526168823,
+      "learning_rate": 2e-05,
+      "loss": 0.04354863,
+      "step": 16146
+    },
+    {
+      "epoch": 32.294,
+      "grad_norm": 1.1236069202423096,
+      "learning_rate": 2e-05,
+      "loss": 0.04897562,
+      "step": 16147
+    },
+    {
+      "epoch": 32.296,
+      "grad_norm": 1.730275273323059,
+      "learning_rate": 2e-05,
+      "loss": 0.03983387,
+      "step": 16148
+    },
+    {
+      "epoch": 32.298,
+      "grad_norm": 1.732548713684082,
+      "learning_rate": 2e-05,
+      "loss": 0.04608201,
+      "step": 16149
+    },
+    {
+      "epoch": 32.3,
+      "grad_norm": 1.9877454042434692,
+      "learning_rate": 2e-05,
+      "loss": 0.03996104,
+      "step": 16150
+    },
+    {
+      "epoch": 32.302,
+      "grad_norm": 1.0003015995025635,
+      "learning_rate": 2e-05,
+      "loss": 0.03154821,
+      "step": 16151
+    },
+    {
+      "epoch": 32.304,
+      "grad_norm": 1.7395548820495605,
+      "learning_rate": 2e-05,
+      "loss": 0.05859068,
+      "step": 16152
+    },
+    {
+      "epoch": 32.306,
+      "grad_norm": 1.7625404596328735,
+      "learning_rate": 2e-05,
+      "loss": 0.04158505,
+      "step": 16153
+    },
+    {
+      "epoch": 32.308,
+      "grad_norm": 1.3035045862197876,
+      "learning_rate": 2e-05,
+      "loss": 0.03832395,
+      "step": 16154
+    },
+    {
+      "epoch": 32.31,
+      "grad_norm": 1.2695159912109375,
+      "learning_rate": 2e-05,
+      "loss": 0.05645255,
+      "step": 16155
+    },
+    {
+      "epoch": 32.312,
+      "grad_norm": 1.68500816822052,
+      "learning_rate": 2e-05,
+      "loss": 0.07209504,
+      "step": 16156
+    },
+    {
+      "epoch": 32.314,
+      "grad_norm": 1.1984082460403442,
+      "learning_rate": 2e-05,
+      "loss": 0.03992949,
+      "step": 16157
+    },
+    {
+      "epoch": 32.316,
+      "grad_norm": 1.199971079826355,
+      "learning_rate": 2e-05,
+      "loss": 0.04131328,
+      "step": 16158
+    },
+    {
+      "epoch": 32.318,
+      "grad_norm": 1.0048577785491943,
+      "learning_rate": 2e-05,
+      "loss": 0.03791015,
+      "step": 16159
+    },
+    {
+      "epoch": 32.32,
+      "grad_norm": 1.0711760520935059,
+      "learning_rate": 2e-05,
+      "loss": 0.03038563,
+      "step": 16160
+    },
+    {
+      "epoch": 32.322,
+      "grad_norm": 1.5091052055358887,
+      "learning_rate": 2e-05,
+      "loss": 0.05084156,
+      "step": 16161
+    },
+    {
+      "epoch": 32.324,
+      "grad_norm": 1.170743703842163,
+      "learning_rate": 2e-05,
+      "loss": 0.04108653,
+      "step": 16162
+    },
+    {
+      "epoch": 32.326,
+      "grad_norm": 0.9663565158843994,
+      "learning_rate": 2e-05,
+      "loss": 0.03507331,
+      "step": 16163
+    },
+    {
+      "epoch": 32.328,
+      "grad_norm": 1.7884368896484375,
+      "learning_rate": 2e-05,
+      "loss": 0.03116245,
+      "step": 16164
+    },
+    {
+      "epoch": 32.33,
+      "grad_norm": 1.2119756937026978,
+      "learning_rate": 2e-05,
+      "loss": 0.03635027,
+      "step": 16165
+    },
+    {
+      "epoch": 32.332,
+      "grad_norm": 1.2660174369812012,
+      "learning_rate": 2e-05,
+      "loss": 0.05102555,
+      "step": 16166
+    },
+    {
+      "epoch": 32.334,
+      "grad_norm": 1.6132479906082153,
+      "learning_rate": 2e-05,
+      "loss": 0.05014277,
+      "step": 16167
+    },
+    {
+      "epoch": 32.336,
+      "grad_norm": 1.722839117050171,
+      "learning_rate": 2e-05,
+      "loss": 0.04384566,
+      "step": 16168
+    },
+    {
+      "epoch": 32.338,
+      "grad_norm": 1.3902097940444946,
+      "learning_rate": 2e-05,
+      "loss": 0.03447413,
+      "step": 16169
+    },
+    {
+      "epoch": 32.34,
+      "grad_norm": 1.2110004425048828,
+      "learning_rate": 2e-05,
+      "loss": 0.03868734,
+      "step": 16170
+    },
+    {
+      "epoch": 32.342,
+      "grad_norm": 1.4563666582107544,
+      "learning_rate": 2e-05,
+      "loss": 0.04213499,
+      "step": 16171
+    },
+    {
+      "epoch": 32.344,
+      "grad_norm": 1.213379144668579,
+      "learning_rate": 2e-05,
+      "loss": 0.05638143,
+      "step": 16172
+    },
+    {
+      "epoch": 32.346,
+      "grad_norm": 1.09957754611969,
+      "learning_rate": 2e-05,
+      "loss": 0.04616127,
+      "step": 16173
+    },
+    {
+      "epoch": 32.348,
+      "grad_norm": 1.0122456550598145,
+      "learning_rate": 2e-05,
+      "loss": 0.0424768,
+      "step": 16174
+    },
+    {
+      "epoch": 32.35,
+      "grad_norm": 1.259447455406189,
+      "learning_rate": 2e-05,
+      "loss": 0.04152884,
+      "step": 16175
+    },
+    {
+      "epoch": 32.352,
+      "grad_norm": 1.1292533874511719,
+      "learning_rate": 2e-05,
+      "loss": 0.04704421,
+      "step": 16176
+    },
+    {
+      "epoch": 32.354,
+      "grad_norm": 1.455267071723938,
+      "learning_rate": 2e-05,
+      "loss": 0.05591056,
+      "step": 16177
+    },
+    {
+      "epoch": 32.356,
+      "grad_norm": 1.2393004894256592,
+      "learning_rate": 2e-05,
+      "loss": 0.03246272,
+      "step": 16178
+    },
+    {
+      "epoch": 32.358,
+      "grad_norm": 1.4194928407669067,
+      "learning_rate": 2e-05,
+      "loss": 0.04797504,
+      "step": 16179
+    },
+    {
+      "epoch": 32.36,
+      "grad_norm": 1.3719854354858398,
+      "learning_rate": 2e-05,
+      "loss": 0.05387818,
+      "step": 16180
+    },
+    {
+      "epoch": 32.362,
+      "grad_norm": 1.0265268087387085,
+      "learning_rate": 2e-05,
+      "loss": 0.03407123,
+      "step": 16181
+    },
+    {
+      "epoch": 32.364,
+      "grad_norm": 1.1385420560836792,
+      "learning_rate": 2e-05,
+      "loss": 0.04104249,
+      "step": 16182
+    },
+    {
+      "epoch": 32.366,
+      "grad_norm": 1.4606928825378418,
+      "learning_rate": 2e-05,
+      "loss": 0.05414707,
+      "step": 16183
+    },
+    {
+      "epoch": 32.368,
+      "grad_norm": 1.4488290548324585,
+      "learning_rate": 2e-05,
+      "loss": 0.05977223,
+      "step": 16184
+    },
+    {
+      "epoch": 32.37,
+      "grad_norm": 1.556361436843872,
+      "learning_rate": 2e-05,
+      "loss": 0.03753366,
+      "step": 16185
+    },
+    {
+      "epoch": 32.372,
+      "grad_norm": 1.0535379648208618,
+      "learning_rate": 2e-05,
+      "loss": 0.038872,
+      "step": 16186
+    },
+    {
+      "epoch": 32.374,
+      "grad_norm": 1.3477610349655151,
+      "learning_rate": 2e-05,
+      "loss": 0.04552489,
+      "step": 16187
+    },
+    {
+      "epoch": 32.376,
+      "grad_norm": 0.9643598198890686,
+      "learning_rate": 2e-05,
+      "loss": 0.03325019,
+      "step": 16188
+    },
+    {
+      "epoch": 32.378,
+      "grad_norm": 2.3846733570098877,
+      "learning_rate": 2e-05,
+      "loss": 0.03706906,
+      "step": 16189
+    },
+    {
+      "epoch": 32.38,
+      "grad_norm": 1.3666075468063354,
+      "learning_rate": 2e-05,
+      "loss": 0.05090994,
+      "step": 16190
+    },
+    {
+      "epoch": 32.382,
+      "grad_norm": 1.3230539560317993,
+      "learning_rate": 2e-05,
+      "loss": 0.04045279,
+      "step": 16191
+    },
+    {
+      "epoch": 32.384,
+      "grad_norm": 1.1574872732162476,
+      "learning_rate": 2e-05,
+      "loss": 0.043694,
+      "step": 16192
+    },
+    {
+      "epoch": 32.386,
+      "grad_norm": 1.11777663230896,
+      "learning_rate": 2e-05,
+      "loss": 0.05323738,
+      "step": 16193
+    },
+    {
+      "epoch": 32.388,
+      "grad_norm": 1.1913745403289795,
+      "learning_rate": 2e-05,
+      "loss": 0.04276652,
+      "step": 16194
+    },
+    {
+      "epoch": 32.39,
+      "grad_norm": 2.0520830154418945,
+      "learning_rate": 2e-05,
+      "loss": 0.05131931,
+      "step": 16195
+    },
+    {
+      "epoch": 32.392,
+      "grad_norm": 1.3622318506240845,
+      "learning_rate": 2e-05,
+      "loss": 0.03571715,
+      "step": 16196
+    },
+    {
+      "epoch": 32.394,
+      "grad_norm": 1.278665542602539,
+      "learning_rate": 2e-05,
+      "loss": 0.04110078,
+      "step": 16197
+    },
+    {
+      "epoch": 32.396,
+      "grad_norm": 1.3019253015518188,
+      "learning_rate": 2e-05,
+      "loss": 0.03151403,
+      "step": 16198
+    },
+    {
+      "epoch": 32.398,
+      "grad_norm": 1.0655170679092407,
+      "learning_rate": 2e-05,
+      "loss": 0.03590965,
+      "step": 16199
+    },
+    {
+      "epoch": 32.4,
+      "grad_norm": 1.0709350109100342,
+      "learning_rate": 2e-05,
+      "loss": 0.03407409,
+      "step": 16200
+    },
+    {
+      "epoch": 32.402,
+      "grad_norm": 1.1004314422607422,
+      "learning_rate": 2e-05,
+      "loss": 0.04275486,
+      "step": 16201
+    },
+    {
+      "epoch": 32.404,
+      "grad_norm": 1.122451901435852,
+      "learning_rate": 2e-05,
+      "loss": 0.0433308,
+      "step": 16202
+    },
+    {
+      "epoch": 32.406,
+      "grad_norm": 1.286196231842041,
+      "learning_rate": 2e-05,
+      "loss": 0.05405517,
+      "step": 16203
+    },
+    {
+      "epoch": 32.408,
+      "grad_norm": 1.3067890405654907,
+      "learning_rate": 2e-05,
+      "loss": 0.04709282,
+      "step": 16204
+    },
+    {
+      "epoch": 32.41,
+      "grad_norm": 1.2169573307037354,
+      "learning_rate": 2e-05,
+      "loss": 0.04799424,
+      "step": 16205
+    },
+    {
+      "epoch": 32.412,
+      "grad_norm": 1.0351089239120483,
+      "learning_rate": 2e-05,
+      "loss": 0.02892509,
+      "step": 16206
+    },
+    {
+      "epoch": 32.414,
+      "grad_norm": 1.1305210590362549,
+      "learning_rate": 2e-05,
+      "loss": 0.03046542,
+      "step": 16207
+    },
+    {
+      "epoch": 32.416,
+      "grad_norm": 1.049100637435913,
+      "learning_rate": 2e-05,
+      "loss": 0.03685327,
+      "step": 16208
+    },
+    {
+      "epoch": 32.418,
+      "grad_norm": 2.5545873641967773,
+      "learning_rate": 2e-05,
+      "loss": 0.05403619,
+      "step": 16209
+    },
+    {
+      "epoch": 32.42,
+      "grad_norm": 1.5375778675079346,
+      "learning_rate": 2e-05,
+      "loss": 0.0551252,
+      "step": 16210
+    },
+    {
+      "epoch": 32.422,
+      "grad_norm": 1.355461597442627,
+      "learning_rate": 2e-05,
+      "loss": 0.03641314,
+      "step": 16211
+    },
+    {
+      "epoch": 32.424,
+      "grad_norm": 2.1974892616271973,
+      "learning_rate": 2e-05,
+      "loss": 0.0479245,
+      "step": 16212
+    },
+    {
+      "epoch": 32.426,
+      "grad_norm": 2.6640677452087402,
+      "learning_rate": 2e-05,
+      "loss": 0.05790597,
+      "step": 16213
+    },
+    {
+      "epoch": 32.428,
+      "grad_norm": 1.0314759016036987,
+      "learning_rate": 2e-05,
+      "loss": 0.02948291,
+      "step": 16214
+    },
+    {
+      "epoch": 32.43,
+      "grad_norm": 1.1263335943222046,
+      "learning_rate": 2e-05,
+      "loss": 0.03837535,
+      "step": 16215
+    },
+    {
+      "epoch": 32.432,
+      "grad_norm": 0.9716264605522156,
+      "learning_rate": 2e-05,
+      "loss": 0.03991567,
+      "step": 16216
+    },
+    {
+      "epoch": 32.434,
+      "grad_norm": 1.3745099306106567,
+      "learning_rate": 2e-05,
+      "loss": 0.05104405,
+      "step": 16217
+    },
+    {
+      "epoch": 32.436,
+      "grad_norm": 1.1306488513946533,
+      "learning_rate": 2e-05,
+      "loss": 0.04568043,
+      "step": 16218
+    },
+    {
+      "epoch": 32.438,
+      "grad_norm": 1.738258719444275,
+      "learning_rate": 2e-05,
+      "loss": 0.06466265,
+      "step": 16219
+    },
+    {
+      "epoch": 32.44,
+      "grad_norm": 1.6814087629318237,
+      "learning_rate": 2e-05,
+      "loss": 0.04136957,
+      "step": 16220
+    },
+    {
+      "epoch": 32.442,
+      "grad_norm": 1.1837775707244873,
+      "learning_rate": 2e-05,
+      "loss": 0.0486927,
+      "step": 16221
+    },
+    {
+      "epoch": 32.444,
+      "grad_norm": 1.257043719291687,
+      "learning_rate": 2e-05,
+      "loss": 0.04791939,
+      "step": 16222
+    },
+    {
+      "epoch": 32.446,
+      "grad_norm": 1.8833086490631104,
+      "learning_rate": 2e-05,
+      "loss": 0.03800988,
+      "step": 16223
+    },
+    {
+      "epoch": 32.448,
+      "grad_norm": 1.1962319612503052,
+      "learning_rate": 2e-05,
+      "loss": 0.03850331,
+      "step": 16224
+    },
+    {
+      "epoch": 32.45,
+      "grad_norm": 1.2660398483276367,
+      "learning_rate": 2e-05,
+      "loss": 0.03511299,
+      "step": 16225
+    },
+    {
+      "epoch": 32.452,
+      "grad_norm": 1.2324693202972412,
+      "learning_rate": 2e-05,
+      "loss": 0.04578947,
+      "step": 16226
+    },
+    {
+      "epoch": 32.454,
+      "grad_norm": 0.8969568014144897,
+      "learning_rate": 2e-05,
+      "loss": 0.03069787,
+      "step": 16227
+    },
+    {
+      "epoch": 32.456,
+      "grad_norm": 1.4285138845443726,
+      "learning_rate": 2e-05,
+      "loss": 0.06139054,
+      "step": 16228
+    },
+    {
+      "epoch": 32.458,
+      "grad_norm": 0.9827262163162231,
+      "learning_rate": 2e-05,
+      "loss": 0.03358797,
+      "step": 16229
+    },
+    {
+      "epoch": 32.46,
+      "grad_norm": 1.6161268949508667,
+      "learning_rate": 2e-05,
+      "loss": 0.05113133,
+      "step": 16230
+    },
+    {
+      "epoch": 32.462,
+      "grad_norm": 1.2342438697814941,
+      "learning_rate": 2e-05,
+      "loss": 0.04767845,
+      "step": 16231
+    },
+    {
+      "epoch": 32.464,
+      "grad_norm": 1.4637534618377686,
+      "learning_rate": 2e-05,
+      "loss": 0.05040576,
+      "step": 16232
+    },
+    {
+      "epoch": 32.466,
+      "grad_norm": 1.251790165901184,
+      "learning_rate": 2e-05,
+      "loss": 0.04769853,
+      "step": 16233
+    },
+    {
+      "epoch": 32.468,
+      "grad_norm": 2.195897102355957,
+      "learning_rate": 2e-05,
+      "loss": 0.04872566,
+      "step": 16234
+    },
+    {
+      "epoch": 32.47,
+      "grad_norm": 0.8366017937660217,
+      "learning_rate": 2e-05,
+      "loss": 0.03476872,
+      "step": 16235
+    },
+    {
+      "epoch": 32.472,
+      "grad_norm": 3.469074249267578,
+      "learning_rate": 2e-05,
+      "loss": 0.04331683,
+      "step": 16236
+    },
+    {
+      "epoch": 32.474,
+      "grad_norm": 2.101489782333374,
+      "learning_rate": 2e-05,
+      "loss": 0.06111233,
+      "step": 16237
+    },
+    {
+      "epoch": 32.476,
+      "grad_norm": 1.5878509283065796,
+      "learning_rate": 2e-05,
+      "loss": 0.04456938,
+      "step": 16238
+    },
+    {
+      "epoch": 32.478,
+      "grad_norm": 1.6882450580596924,
+      "learning_rate": 2e-05,
+      "loss": 0.06533901,
+      "step": 16239
+    },
+    {
+      "epoch": 32.48,
+      "grad_norm": 1.980782389640808,
+      "learning_rate": 2e-05,
+      "loss": 0.02859255,
+      "step": 16240
+    },
+    {
+      "epoch": 32.482,
+      "grad_norm": 2.289910316467285,
+      "learning_rate": 2e-05,
+      "loss": 0.04708671,
+      "step": 16241
+    },
+    {
+      "epoch": 32.484,
+      "grad_norm": 1.2156428098678589,
+      "learning_rate": 2e-05,
+      "loss": 0.06721392,
+      "step": 16242
+    },
+    {
+      "epoch": 32.486,
+      "grad_norm": 2.177699327468872,
+      "learning_rate": 2e-05,
+      "loss": 0.03834368,
+      "step": 16243
+    },
+    {
+      "epoch": 32.488,
+      "grad_norm": 1.3921353816986084,
+      "learning_rate": 2e-05,
+      "loss": 0.05225492,
+      "step": 16244
+    },
+    {
+      "epoch": 32.49,
+      "grad_norm": 1.0509675741195679,
+      "learning_rate": 2e-05,
+      "loss": 0.03782216,
+      "step": 16245
+    },
+    {
+      "epoch": 32.492,
+      "grad_norm": 0.8965580463409424,
+      "learning_rate": 2e-05,
+      "loss": 0.03118903,
+      "step": 16246
+    },
+    {
+      "epoch": 32.494,
+      "grad_norm": 1.2393367290496826,
+      "learning_rate": 2e-05,
+      "loss": 0.04286648,
+      "step": 16247
+    },
+    {
+      "epoch": 32.496,
+      "grad_norm": 4.540200233459473,
+      "learning_rate": 2e-05,
+      "loss": 0.05019096,
+      "step": 16248
+    },
+    {
+      "epoch": 32.498,
+      "grad_norm": 1.4414736032485962,
+      "learning_rate": 2e-05,
+      "loss": 0.03688266,
+      "step": 16249
+    },
+    {
+      "epoch": 32.5,
+      "grad_norm": 1.0771220922470093,
+      "learning_rate": 2e-05,
+      "loss": 0.04028372,
+      "step": 16250
+    },
+    {
+      "epoch": 32.502,
+      "grad_norm": 1.2341318130493164,
+      "learning_rate": 2e-05,
+      "loss": 0.04482314,
+      "step": 16251
+    },
+    {
+      "epoch": 32.504,
+      "grad_norm": 1.4283705949783325,
+      "learning_rate": 2e-05,
+      "loss": 0.04541865,
+      "step": 16252
+    },
+    {
+      "epoch": 32.506,
+      "grad_norm": 1.5509240627288818,
+      "learning_rate": 2e-05,
+      "loss": 0.06562129,
+      "step": 16253
+    },
+    {
+      "epoch": 32.508,
+      "grad_norm": 1.07283353805542,
+      "learning_rate": 2e-05,
+      "loss": 0.04550766,
+      "step": 16254
+    },
+    {
+      "epoch": 32.51,
+      "grad_norm": 1.1795142889022827,
+      "learning_rate": 2e-05,
+      "loss": 0.05442602,
+      "step": 16255
+    },
+    {
+      "epoch": 32.512,
+      "grad_norm": 1.1918917894363403,
+      "learning_rate": 2e-05,
+      "loss": 0.04531888,
+      "step": 16256
+    },
+    {
+      "epoch": 32.514,
+      "grad_norm": 1.6722873449325562,
+      "learning_rate": 2e-05,
+      "loss": 0.04310684,
+      "step": 16257
+    },
+    {
+      "epoch": 32.516,
+      "grad_norm": 1.2727532386779785,
+      "learning_rate": 2e-05,
+      "loss": 0.04338937,
+      "step": 16258
+    },
+    {
+      "epoch": 32.518,
+      "grad_norm": 1.411271095275879,
+      "learning_rate": 2e-05,
+      "loss": 0.0337945,
+      "step": 16259
+    },
+    {
+      "epoch": 32.52,
+      "grad_norm": 1.1750949621200562,
+      "learning_rate": 2e-05,
+      "loss": 0.03457836,
+      "step": 16260
+    },
+    {
+      "epoch": 32.522,
+      "grad_norm": 3.289924144744873,
+      "learning_rate": 2e-05,
+      "loss": 0.04035544,
+      "step": 16261
+    },
+    {
+      "epoch": 32.524,
+      "grad_norm": 0.9971786737442017,
+      "learning_rate": 2e-05,
+      "loss": 0.03331866,
+      "step": 16262
+    },
+    {
+      "epoch": 32.526,
+      "grad_norm": 1.3523638248443604,
+      "learning_rate": 2e-05,
+      "loss": 0.04086154,
+      "step": 16263
+    },
+    {
+      "epoch": 32.528,
+      "grad_norm": 1.1040973663330078,
+      "learning_rate": 2e-05,
+      "loss": 0.0335994,
+      "step": 16264
+    },
+    {
+      "epoch": 32.53,
+      "grad_norm": 1.7474960088729858,
+      "learning_rate": 2e-05,
+      "loss": 0.04292202,
+      "step": 16265
+    },
+    {
+      "epoch": 32.532,
+      "grad_norm": 1.0948997735977173,
+      "learning_rate": 2e-05,
+      "loss": 0.03585258,
+      "step": 16266
+    },
+    {
+      "epoch": 32.534,
+      "grad_norm": 1.512582778930664,
+      "learning_rate": 2e-05,
+      "loss": 0.04289878,
+      "step": 16267
+    },
+    {
+      "epoch": 32.536,
+      "grad_norm": 1.9142391681671143,
+      "learning_rate": 2e-05,
+      "loss": 0.04722939,
+      "step": 16268
+    },
+    {
+      "epoch": 32.538,
+      "grad_norm": 1.0834648609161377,
+      "learning_rate": 2e-05,
+      "loss": 0.03546905,
+      "step": 16269
+    },
+    {
+      "epoch": 32.54,
+      "grad_norm": 1.8602557182312012,
+      "learning_rate": 2e-05,
+      "loss": 0.03079529,
+      "step": 16270
+    },
+    {
+      "epoch": 32.542,
+      "grad_norm": 1.2603952884674072,
+      "learning_rate": 2e-05,
+      "loss": 0.04889859,
+      "step": 16271
+    },
+    {
+      "epoch": 32.544,
+      "grad_norm": 0.9242783188819885,
+      "learning_rate": 2e-05,
+      "loss": 0.02692599,
+      "step": 16272
+    },
+    {
+      "epoch": 32.546,
+      "grad_norm": 1.140777349472046,
+      "learning_rate": 2e-05,
+      "loss": 0.04153073,
+      "step": 16273
+    },
+    {
+      "epoch": 32.548,
+      "grad_norm": 1.153734564781189,
+      "learning_rate": 2e-05,
+      "loss": 0.04307986,
+      "step": 16274
+    },
+    {
+      "epoch": 32.55,
+      "grad_norm": 1.1327687501907349,
+      "learning_rate": 2e-05,
+      "loss": 0.04815378,
+      "step": 16275
+    },
+    {
+      "epoch": 32.552,
+      "grad_norm": 1.2450306415557861,
+      "learning_rate": 2e-05,
+      "loss": 0.0305282,
+      "step": 16276
+    },
+    {
+      "epoch": 32.554,
+      "grad_norm": 1.2972357273101807,
+      "learning_rate": 2e-05,
+      "loss": 0.05288756,
+      "step": 16277
+    },
+    {
+      "epoch": 32.556,
+      "grad_norm": 1.880336880683899,
+      "learning_rate": 2e-05,
+      "loss": 0.02928368,
+      "step": 16278
+    },
+    {
+      "epoch": 32.558,
+      "grad_norm": 1.2293763160705566,
+      "learning_rate": 2e-05,
+      "loss": 0.03244419,
+      "step": 16279
+    },
+    {
+      "epoch": 32.56,
+      "grad_norm": 1.0479623079299927,
+      "learning_rate": 2e-05,
+      "loss": 0.03361559,
+      "step": 16280
+    },
+    {
+      "epoch": 32.562,
+      "grad_norm": 1.0732969045639038,
+      "learning_rate": 2e-05,
+      "loss": 0.0405861,
+      "step": 16281
+    },
+    {
+      "epoch": 32.564,
+      "grad_norm": 1.251217007637024,
+      "learning_rate": 2e-05,
+      "loss": 0.03988477,
+      "step": 16282
+    },
+    {
+      "epoch": 32.566,
+      "grad_norm": 1.0673967599868774,
+      "learning_rate": 2e-05,
+      "loss": 0.03976289,
+      "step": 16283
+    },
+    {
+      "epoch": 32.568,
+      "grad_norm": 1.1585348844528198,
+      "learning_rate": 2e-05,
+      "loss": 0.0360926,
+      "step": 16284
+    },
+    {
+      "epoch": 32.57,
+      "grad_norm": 1.051295518875122,
+      "learning_rate": 2e-05,
+      "loss": 0.03265924,
+      "step": 16285
+    },
+    {
+      "epoch": 32.572,
+      "grad_norm": 1.1959424018859863,
+      "learning_rate": 2e-05,
+      "loss": 0.03449826,
+      "step": 16286
+    },
+    {
+      "epoch": 32.574,
+      "grad_norm": 2.357743978500366,
+      "learning_rate": 2e-05,
+      "loss": 0.06374232,
+      "step": 16287
+    },
+    {
+      "epoch": 32.576,
+      "grad_norm": 1.252048373222351,
+      "learning_rate": 2e-05,
+      "loss": 0.04862808,
+      "step": 16288
+    },
+    {
+      "epoch": 32.578,
+      "grad_norm": 1.318522572517395,
+      "learning_rate": 2e-05,
+      "loss": 0.05090618,
+      "step": 16289
+    },
+    {
+      "epoch": 32.58,
+      "grad_norm": 1.414203405380249,
+      "learning_rate": 2e-05,
+      "loss": 0.03399177,
+      "step": 16290
+    },
+    {
+      "epoch": 32.582,
+      "grad_norm": 1.1485605239868164,
+      "learning_rate": 2e-05,
+      "loss": 0.03333712,
+      "step": 16291
+    },
+    {
+      "epoch": 32.584,
+      "grad_norm": 1.231711506843567,
+      "learning_rate": 2e-05,
+      "loss": 0.04081344,
+      "step": 16292
+    },
+    {
+      "epoch": 32.586,
+      "grad_norm": 1.4785326719284058,
+      "learning_rate": 2e-05,
+      "loss": 0.04804887,
+      "step": 16293
+    },
+    {
+      "epoch": 32.588,
+      "grad_norm": 1.113040566444397,
+      "learning_rate": 2e-05,
+      "loss": 0.04360687,
+      "step": 16294
+    },
+    {
+      "epoch": 32.59,
+      "grad_norm": 1.2656840085983276,
+      "learning_rate": 2e-05,
+      "loss": 0.03424615,
+      "step": 16295
+    },
+    {
+      "epoch": 32.592,
+      "grad_norm": 2.090235471725464,
+      "learning_rate": 2e-05,
+      "loss": 0.03176118,
+      "step": 16296
+    },
+    {
+      "epoch": 32.594,
+      "grad_norm": 1.2419369220733643,
+      "learning_rate": 2e-05,
+      "loss": 0.0560364,
+      "step": 16297
+    },
+    {
+      "epoch": 32.596,
+      "grad_norm": 1.8632920980453491,
+      "learning_rate": 2e-05,
+      "loss": 0.04524683,
+      "step": 16298
+    },
+    {
+      "epoch": 32.598,
+      "grad_norm": 1.5768665075302124,
+      "learning_rate": 2e-05,
+      "loss": 0.04588313,
+      "step": 16299
+    },
+    {
+      "epoch": 32.6,
+      "grad_norm": 2.376133918762207,
+      "learning_rate": 2e-05,
+      "loss": 0.06110063,
+      "step": 16300
+    },
+    {
+      "epoch": 32.602,
+      "grad_norm": 1.2238625288009644,
+      "learning_rate": 2e-05,
+      "loss": 0.04093513,
+      "step": 16301
+    },
+    {
+      "epoch": 32.604,
+      "grad_norm": 1.8445336818695068,
+      "learning_rate": 2e-05,
+      "loss": 0.03895678,
+      "step": 16302
+    },
+    {
+      "epoch": 32.606,
+      "grad_norm": 1.5113000869750977,
+      "learning_rate": 2e-05,
+      "loss": 0.04946197,
+      "step": 16303
+    },
+    {
+      "epoch": 32.608,
+      "grad_norm": 1.107370376586914,
+      "learning_rate": 2e-05,
+      "loss": 0.02631177,
+      "step": 16304
+    },
+    {
+      "epoch": 32.61,
+      "grad_norm": 1.319182276725769,
+      "learning_rate": 2e-05,
+      "loss": 0.03960687,
+      "step": 16305
+    },
+    {
+      "epoch": 32.612,
+      "grad_norm": 1.1253376007080078,
+      "learning_rate": 2e-05,
+      "loss": 0.03078615,
+      "step": 16306
+    },
+    {
+      "epoch": 32.614,
+      "grad_norm": 1.0863707065582275,
+      "learning_rate": 2e-05,
+      "loss": 0.03195829,
+      "step": 16307
+    },
+    {
+      "epoch": 32.616,
+      "grad_norm": 1.9176464080810547,
+      "learning_rate": 2e-05,
+      "loss": 0.04252298,
+      "step": 16308
+    },
+    {
+      "epoch": 32.618,
+      "grad_norm": 1.123355746269226,
+      "learning_rate": 2e-05,
+      "loss": 0.03697478,
+      "step": 16309
+    },
+    {
+      "epoch": 32.62,
+      "grad_norm": 3.388070583343506,
+      "learning_rate": 2e-05,
+      "loss": 0.06398418,
+      "step": 16310
+    },
+    {
+      "epoch": 32.622,
+      "grad_norm": 1.3932557106018066,
+      "learning_rate": 2e-05,
+      "loss": 0.05024934,
+      "step": 16311
+    },
+    {
+      "epoch": 32.624,
+      "grad_norm": 1.2493484020233154,
+      "learning_rate": 2e-05,
+      "loss": 0.04943375,
+      "step": 16312
+    },
+    {
+      "epoch": 32.626,
+      "grad_norm": 1.1439610719680786,
+      "learning_rate": 2e-05,
+      "loss": 0.0385169,
+      "step": 16313
+    },
+    {
+      "epoch": 32.628,
+      "grad_norm": 2.4775614738464355,
+      "learning_rate": 2e-05,
+      "loss": 0.06101976,
+      "step": 16314
+    },
+    {
+      "epoch": 32.63,
+      "grad_norm": 2.373351573944092,
+      "learning_rate": 2e-05,
+      "loss": 0.04415338,
+      "step": 16315
+    },
+    {
+      "epoch": 32.632,
+      "grad_norm": 1.7569504976272583,
+      "learning_rate": 2e-05,
+      "loss": 0.04195081,
+      "step": 16316
+    },
+    {
+      "epoch": 32.634,
+      "grad_norm": 1.1292805671691895,
+      "learning_rate": 2e-05,
+      "loss": 0.035651,
+      "step": 16317
+    },
+    {
+      "epoch": 32.636,
+      "grad_norm": 1.251456618309021,
+      "learning_rate": 2e-05,
+      "loss": 0.04292186,
+      "step": 16318
+    },
+    {
+      "epoch": 32.638,
+      "grad_norm": 0.9629459977149963,
+      "learning_rate": 2e-05,
+      "loss": 0.02403329,
+      "step": 16319
+    },
+    {
+      "epoch": 32.64,
+      "grad_norm": 1.001502275466919,
+      "learning_rate": 2e-05,
+      "loss": 0.03040902,
+      "step": 16320
+    },
+    {
+      "epoch": 32.642,
+      "grad_norm": 1.3388837575912476,
+      "learning_rate": 2e-05,
+      "loss": 0.0465103,
+      "step": 16321
+    },
+    {
+      "epoch": 32.644,
+      "grad_norm": 1.0850396156311035,
+      "learning_rate": 2e-05,
+      "loss": 0.03716237,
+      "step": 16322
+    },
+    {
+      "epoch": 32.646,
+      "grad_norm": 3.381190776824951,
+      "learning_rate": 2e-05,
+      "loss": 0.06202536,
+      "step": 16323
+    },
+    {
+      "epoch": 32.648,
+      "grad_norm": 1.3530367612838745,
+      "learning_rate": 2e-05,
+      "loss": 0.03582821,
+      "step": 16324
+    },
+    {
+      "epoch": 32.65,
+      "grad_norm": 2.433734893798828,
+      "learning_rate": 2e-05,
+      "loss": 0.05068055,
+      "step": 16325
+    },
+    {
+      "epoch": 32.652,
+      "grad_norm": 1.7328464984893799,
+      "learning_rate": 2e-05,
+      "loss": 0.0590675,
+      "step": 16326
+    },
+    {
+      "epoch": 32.654,
+      "grad_norm": 1.1107189655303955,
+      "learning_rate": 2e-05,
+      "loss": 0.05116561,
+      "step": 16327
+    },
+    {
+      "epoch": 32.656,
+      "grad_norm": 1.2130955457687378,
+      "learning_rate": 2e-05,
+      "loss": 0.03328083,
+      "step": 16328
+    },
+    {
+      "epoch": 32.658,
+      "grad_norm": 1.7198522090911865,
+      "learning_rate": 2e-05,
+      "loss": 0.04785996,
+      "step": 16329
+    },
+    {
+      "epoch": 32.66,
+      "grad_norm": 1.5912532806396484,
+      "learning_rate": 2e-05,
+      "loss": 0.04132064,
+      "step": 16330
+    },
+    {
+      "epoch": 32.662,
+      "grad_norm": 1.930972933769226,
+      "learning_rate": 2e-05,
+      "loss": 0.03286422,
+      "step": 16331
+    },
+    {
+      "epoch": 32.664,
+      "grad_norm": 1.51142156124115,
+      "learning_rate": 2e-05,
+      "loss": 0.05630952,
+      "step": 16332
+    },
+    {
+      "epoch": 32.666,
+      "grad_norm": 1.2394816875457764,
+      "learning_rate": 2e-05,
+      "loss": 0.05155517,
+      "step": 16333
+    },
+    {
+      "epoch": 32.668,
+      "grad_norm": 2.041574716567993,
+      "learning_rate": 2e-05,
+      "loss": 0.04154414,
+      "step": 16334
+    },
+    {
+      "epoch": 32.67,
+      "grad_norm": 3.471482992172241,
+      "learning_rate": 2e-05,
+      "loss": 0.06497285,
+      "step": 16335
+    },
+    {
+      "epoch": 32.672,
+      "grad_norm": 1.1447969675064087,
+      "learning_rate": 2e-05,
+      "loss": 0.04535265,
+      "step": 16336
+    },
+    {
+      "epoch": 32.674,
+      "grad_norm": 1.678331732749939,
+      "learning_rate": 2e-05,
+      "loss": 0.0535218,
+      "step": 16337
+    },
+    {
+      "epoch": 32.676,
+      "grad_norm": 1.378172755241394,
+      "learning_rate": 2e-05,
+      "loss": 0.04844815,
+      "step": 16338
+    },
+    {
+      "epoch": 32.678,
+      "grad_norm": 0.9606616497039795,
+      "learning_rate": 2e-05,
+      "loss": 0.03191324,
+      "step": 16339
+    },
+    {
+      "epoch": 32.68,
+      "grad_norm": 1.0323848724365234,
+      "learning_rate": 2e-05,
+      "loss": 0.0283333,
+      "step": 16340
+    },
+    {
+      "epoch": 32.682,
+      "grad_norm": 1.1071007251739502,
+      "learning_rate": 2e-05,
+      "loss": 0.03747611,
+      "step": 16341
+    },
+    {
+      "epoch": 32.684,
+      "grad_norm": 1.0263428688049316,
+      "learning_rate": 2e-05,
+      "loss": 0.03178697,
+      "step": 16342
+    },
+    {
+      "epoch": 32.686,
+      "grad_norm": 1.2423036098480225,
+      "learning_rate": 2e-05,
+      "loss": 0.03112799,
+      "step": 16343
+    },
+    {
+      "epoch": 32.688,
+      "grad_norm": 1.1887843608856201,
+      "learning_rate": 2e-05,
+      "loss": 0.04574025,
+      "step": 16344
+    },
+    {
+      "epoch": 32.69,
+      "grad_norm": 1.2458415031433105,
+      "learning_rate": 2e-05,
+      "loss": 0.04187585,
+      "step": 16345
+    },
+    {
+      "epoch": 32.692,
+      "grad_norm": 1.3959715366363525,
+      "learning_rate": 2e-05,
+      "loss": 0.0457012,
+      "step": 16346
+    },
+    {
+      "epoch": 32.694,
+      "grad_norm": 1.9177616834640503,
+      "learning_rate": 2e-05,
+      "loss": 0.03737821,
+      "step": 16347
+    },
+    {
+      "epoch": 32.696,
+      "grad_norm": 1.9234297275543213,
+      "learning_rate": 2e-05,
+      "loss": 0.05901012,
+      "step": 16348
+    },
+    {
+      "epoch": 32.698,
+      "grad_norm": 2.6709837913513184,
+      "learning_rate": 2e-05,
+      "loss": 0.05703682,
+      "step": 16349
+    },
+    {
+      "epoch": 32.7,
+      "grad_norm": 1.9380576610565186,
+      "learning_rate": 2e-05,
+      "loss": 0.05842331,
+      "step": 16350
+    },
+    {
+      "epoch": 32.702,
+      "grad_norm": 1.0618531703948975,
+      "learning_rate": 2e-05,
+      "loss": 0.03581022,
+      "step": 16351
+    },
+    {
+      "epoch": 32.704,
+      "grad_norm": 1.2853940725326538,
+      "learning_rate": 2e-05,
+      "loss": 0.04471534,
+      "step": 16352
+    },
+    {
+      "epoch": 32.706,
+      "grad_norm": 1.9387348890304565,
+      "learning_rate": 2e-05,
+      "loss": 0.04033763,
+      "step": 16353
+    },
+    {
+      "epoch": 32.708,
+      "grad_norm": 1.651381015777588,
+      "learning_rate": 2e-05,
+      "loss": 0.04997224,
+      "step": 16354
+    },
+    {
+      "epoch": 32.71,
+      "grad_norm": 1.1719062328338623,
+      "learning_rate": 2e-05,
+      "loss": 0.04294977,
+      "step": 16355
+    },
+    {
+      "epoch": 32.712,
+      "grad_norm": 1.1827785968780518,
+      "learning_rate": 2e-05,
+      "loss": 0.06157464,
+      "step": 16356
+    },
+    {
+      "epoch": 32.714,
+      "grad_norm": 2.0270729064941406,
+      "learning_rate": 2e-05,
+      "loss": 0.04811145,
+      "step": 16357
+    },
+    {
+      "epoch": 32.716,
+      "grad_norm": 2.081660032272339,
+      "learning_rate": 2e-05,
+      "loss": 0.03992213,
+      "step": 16358
+    },
+    {
+      "epoch": 32.718,
+      "grad_norm": 1.2889872789382935,
+      "learning_rate": 2e-05,
+      "loss": 0.05166517,
+      "step": 16359
+    },
+    {
+      "epoch": 32.72,
+      "grad_norm": 1.5523995161056519,
+      "learning_rate": 2e-05,
+      "loss": 0.04683542,
+      "step": 16360
+    },
+    {
+      "epoch": 32.722,
+      "grad_norm": 1.2523963451385498,
+      "learning_rate": 2e-05,
+      "loss": 0.03490978,
+      "step": 16361
+    },
+    {
+      "epoch": 32.724,
+      "grad_norm": 1.1765249967575073,
+      "learning_rate": 2e-05,
+      "loss": 0.04525149,
+      "step": 16362
+    },
+    {
+      "epoch": 32.726,
+      "grad_norm": 2.704094409942627,
+      "learning_rate": 2e-05,
+      "loss": 0.05494591,
+      "step": 16363
+    },
+    {
+      "epoch": 32.728,
+      "grad_norm": 1.1506249904632568,
+      "learning_rate": 2e-05,
+      "loss": 0.04181469,
+      "step": 16364
+    },
+    {
+      "epoch": 32.73,
+      "grad_norm": 1.0797075033187866,
+      "learning_rate": 2e-05,
+      "loss": 0.03218354,
+      "step": 16365
+    },
+    {
+      "epoch": 32.732,
+      "grad_norm": 1.3011102676391602,
+      "learning_rate": 2e-05,
+      "loss": 0.03808787,
+      "step": 16366
+    },
+    {
+      "epoch": 32.734,
+      "grad_norm": 1.300958514213562,
+      "learning_rate": 2e-05,
+      "loss": 0.0296345,
+      "step": 16367
+    },
+    {
+      "epoch": 32.736,
+      "grad_norm": 1.4178552627563477,
+      "learning_rate": 2e-05,
+      "loss": 0.05248306,
+      "step": 16368
+    },
+    {
+      "epoch": 32.738,
+      "grad_norm": 1.0817713737487793,
+      "learning_rate": 2e-05,
+      "loss": 0.03825011,
+      "step": 16369
+    },
+    {
+      "epoch": 32.74,
+      "grad_norm": 1.0695898532867432,
+      "learning_rate": 2e-05,
+      "loss": 0.04175761,
+      "step": 16370
+    },
+    {
+      "epoch": 32.742,
+      "grad_norm": 2.968357801437378,
+      "learning_rate": 2e-05,
+      "loss": 0.05305419,
+      "step": 16371
+    },
+    {
+      "epoch": 32.744,
+      "grad_norm": 1.0489188432693481,
+      "learning_rate": 2e-05,
+      "loss": 0.03766435,
+      "step": 16372
+    },
+    {
+      "epoch": 32.746,
+      "grad_norm": 0.9474694132804871,
+      "learning_rate": 2e-05,
+      "loss": 0.03348034,
+      "step": 16373
+    },
+    {
+      "epoch": 32.748,
+      "grad_norm": 0.9930300116539001,
+      "learning_rate": 2e-05,
+      "loss": 0.03745917,
+      "step": 16374
+    },
+    {
+      "epoch": 32.75,
+      "grad_norm": 1.0736403465270996,
+      "learning_rate": 2e-05,
+      "loss": 0.0384927,
+      "step": 16375
+    },
+    {
+      "epoch": 32.752,
+      "grad_norm": 1.2080556154251099,
+      "learning_rate": 2e-05,
+      "loss": 0.05006144,
+      "step": 16376
+    },
+    {
+      "epoch": 32.754,
+      "grad_norm": 0.9601371884346008,
+      "learning_rate": 2e-05,
+      "loss": 0.03716236,
+      "step": 16377
+    },
+    {
+      "epoch": 32.756,
+      "grad_norm": 1.128725528717041,
+      "learning_rate": 2e-05,
+      "loss": 0.04063039,
+      "step": 16378
+    },
+    {
+      "epoch": 32.758,
+      "grad_norm": 1.2183740139007568,
+      "learning_rate": 2e-05,
+      "loss": 0.05256324,
+      "step": 16379
+    },
+    {
+      "epoch": 32.76,
+      "grad_norm": 3.258612632751465,
+      "learning_rate": 2e-05,
+      "loss": 0.04376873,
+      "step": 16380
+    },
+    {
+      "epoch": 32.762,
+      "grad_norm": 1.1058506965637207,
+      "learning_rate": 2e-05,
+      "loss": 0.04281379,
+      "step": 16381
+    },
+    {
+      "epoch": 32.764,
+      "grad_norm": 1.1320183277130127,
+      "learning_rate": 2e-05,
+      "loss": 0.04657583,
+      "step": 16382
+    },
+    {
+      "epoch": 32.766,
+      "grad_norm": 1.4201374053955078,
+      "learning_rate": 2e-05,
+      "loss": 0.04276565,
+      "step": 16383
+    },
+    {
+      "epoch": 32.768,
+      "grad_norm": 1.0543874502182007,
+      "learning_rate": 2e-05,
+      "loss": 0.04018901,
+      "step": 16384
+    },
+    {
+      "epoch": 32.77,
+      "grad_norm": 1.4601715803146362,
+      "learning_rate": 2e-05,
+      "loss": 0.0541468,
+      "step": 16385
+    },
+    {
+      "epoch": 32.772,
+      "grad_norm": 2.787783145904541,
+      "learning_rate": 2e-05,
+      "loss": 0.0562762,
+      "step": 16386
+    },
+    {
+      "epoch": 32.774,
+      "grad_norm": 1.4297536611557007,
+      "learning_rate": 2e-05,
+      "loss": 0.04877428,
+      "step": 16387
+    },
+    {
+      "epoch": 32.776,
+      "grad_norm": 1.5178031921386719,
+      "learning_rate": 2e-05,
+      "loss": 0.03471943,
+      "step": 16388
+    },
+    {
+      "epoch": 32.778,
+      "grad_norm": 1.089928388595581,
+      "learning_rate": 2e-05,
+      "loss": 0.04103519,
+      "step": 16389
+    },
+    {
+      "epoch": 32.78,
+      "grad_norm": 0.9991024136543274,
+      "learning_rate": 2e-05,
+      "loss": 0.03855386,
+      "step": 16390
+    },
+    {
+      "epoch": 32.782,
+      "grad_norm": 1.0029306411743164,
+      "learning_rate": 2e-05,
+      "loss": 0.04183157,
+      "step": 16391
+    },
+    {
+      "epoch": 32.784,
+      "grad_norm": 1.2355519533157349,
+      "learning_rate": 2e-05,
+      "loss": 0.05071557,
+      "step": 16392
+    },
+    {
+      "epoch": 32.786,
+      "grad_norm": 0.9400306940078735,
+      "learning_rate": 2e-05,
+      "loss": 0.03195136,
+      "step": 16393
+    },
+    {
+      "epoch": 32.788,
+      "grad_norm": 1.4382003545761108,
+      "learning_rate": 2e-05,
+      "loss": 0.04054421,
+      "step": 16394
+    },
+    {
+      "epoch": 32.79,
+      "grad_norm": 1.1275018453598022,
+      "learning_rate": 2e-05,
+      "loss": 0.03660734,
+      "step": 16395
+    },
+    {
+      "epoch": 32.792,
+      "grad_norm": 1.1579735279083252,
+      "learning_rate": 2e-05,
+      "loss": 0.03553727,
+      "step": 16396
+    },
+    {
+      "epoch": 32.794,
+      "grad_norm": 1.317656397819519,
+      "learning_rate": 2e-05,
+      "loss": 0.03424529,
+      "step": 16397
+    },
+    {
+      "epoch": 32.796,
+      "grad_norm": 2.1389896869659424,
+      "learning_rate": 2e-05,
+      "loss": 0.04343967,
+      "step": 16398
+    },
+    {
+      "epoch": 32.798,
+      "grad_norm": 1.9410218000411987,
+      "learning_rate": 2e-05,
+      "loss": 0.05163146,
+      "step": 16399
+    },
+    {
+      "epoch": 32.8,
+      "grad_norm": 1.146166205406189,
+      "learning_rate": 2e-05,
+      "loss": 0.03072646,
+      "step": 16400
+    },
+    {
+      "epoch": 32.802,
+      "grad_norm": 1.333733081817627,
+      "learning_rate": 2e-05,
+      "loss": 0.0516637,
+      "step": 16401
+    },
+    {
+      "epoch": 32.804,
+      "grad_norm": 1.0944757461547852,
+      "learning_rate": 2e-05,
+      "loss": 0.03826744,
+      "step": 16402
+    },
+    {
+      "epoch": 32.806,
+      "grad_norm": 1.241289734840393,
+      "learning_rate": 2e-05,
+      "loss": 0.04015851,
+      "step": 16403
+    },
+    {
+      "epoch": 32.808,
+      "grad_norm": 1.1392414569854736,
+      "learning_rate": 2e-05,
+      "loss": 0.04946699,
+      "step": 16404
+    },
+    {
+      "epoch": 32.81,
+      "grad_norm": 0.9604151844978333,
+      "learning_rate": 2e-05,
+      "loss": 0.02486083,
+      "step": 16405
+    },
+    {
+      "epoch": 32.812,
+      "grad_norm": 1.4133819341659546,
+      "learning_rate": 2e-05,
+      "loss": 0.03912335,
+      "step": 16406
+    },
+    {
+      "epoch": 32.814,
+      "grad_norm": 3.016547441482544,
+      "learning_rate": 2e-05,
+      "loss": 0.05131318,
+      "step": 16407
+    },
+    {
+      "epoch": 32.816,
+      "grad_norm": 1.6161396503448486,
+      "learning_rate": 2e-05,
+      "loss": 0.05032074,
+      "step": 16408
+    },
+    {
+      "epoch": 32.818,
+      "grad_norm": 1.1903536319732666,
+      "learning_rate": 2e-05,
+      "loss": 0.0486296,
+      "step": 16409
+    },
+    {
+      "epoch": 32.82,
+      "grad_norm": 1.2930749654769897,
+      "learning_rate": 2e-05,
+      "loss": 0.04178453,
+      "step": 16410
+    },
+    {
+      "epoch": 32.822,
+      "grad_norm": 0.9721649885177612,
+      "learning_rate": 2e-05,
+      "loss": 0.03203877,
+      "step": 16411
+    },
+    {
+      "epoch": 32.824,
+      "grad_norm": 1.4896728992462158,
+      "learning_rate": 2e-05,
+      "loss": 0.0417724,
+      "step": 16412
+    },
+    {
+      "epoch": 32.826,
+      "grad_norm": 0.9889609217643738,
+      "learning_rate": 2e-05,
+      "loss": 0.03722448,
+      "step": 16413
+    },
+    {
+      "epoch": 32.828,
+      "grad_norm": 1.1123636960983276,
+      "learning_rate": 2e-05,
+      "loss": 0.04222369,
+      "step": 16414
+    },
+    {
+      "epoch": 32.83,
+      "grad_norm": 1.0436445474624634,
+      "learning_rate": 2e-05,
+      "loss": 0.03794225,
+      "step": 16415
+    },
+    {
+      "epoch": 32.832,
+      "grad_norm": 1.930905818939209,
+      "learning_rate": 2e-05,
+      "loss": 0.04170159,
+      "step": 16416
+    },
+    {
+      "epoch": 32.834,
+      "grad_norm": 1.0021107196807861,
+      "learning_rate": 2e-05,
+      "loss": 0.03140467,
+      "step": 16417
+    },
+    {
+      "epoch": 32.836,
+      "grad_norm": 1.0246645212173462,
+      "learning_rate": 2e-05,
+      "loss": 0.03313623,
+      "step": 16418
+    },
+    {
+      "epoch": 32.838,
+      "grad_norm": 1.666858196258545,
+      "learning_rate": 2e-05,
+      "loss": 0.04175393,
+      "step": 16419
+    },
+    {
+      "epoch": 32.84,
+      "grad_norm": 1.1152204275131226,
+      "learning_rate": 2e-05,
+      "loss": 0.02505697,
+      "step": 16420
+    },
+    {
+      "epoch": 32.842,
+      "grad_norm": 0.992870032787323,
+      "learning_rate": 2e-05,
+      "loss": 0.03373595,
+      "step": 16421
+    },
+    {
+      "epoch": 32.844,
+      "grad_norm": 1.0322585105895996,
+      "learning_rate": 2e-05,
+      "loss": 0.03882974,
+      "step": 16422
+    },
+    {
+      "epoch": 32.846,
+      "grad_norm": 2.256075620651245,
+      "learning_rate": 2e-05,
+      "loss": 0.05817116,
+      "step": 16423
+    },
+    {
+      "epoch": 32.848,
+      "grad_norm": 5.627259254455566,
+      "learning_rate": 2e-05,
+      "loss": 0.06864762,
+      "step": 16424
+    },
+    {
+      "epoch": 32.85,
+      "grad_norm": 1.0085660219192505,
+      "learning_rate": 2e-05,
+      "loss": 0.04046705,
+      "step": 16425
+    },
+    {
+      "epoch": 32.852,
+      "grad_norm": 1.0974931716918945,
+      "learning_rate": 2e-05,
+      "loss": 0.03791347,
+      "step": 16426
+    },
+    {
+      "epoch": 32.854,
+      "grad_norm": 1.2425390481948853,
+      "learning_rate": 2e-05,
+      "loss": 0.03167582,
+      "step": 16427
+    },
+    {
+      "epoch": 32.856,
+      "grad_norm": 1.2744793891906738,
+      "learning_rate": 2e-05,
+      "loss": 0.04943582,
+      "step": 16428
+    },
+    {
+      "epoch": 32.858,
+      "grad_norm": 0.9825481176376343,
+      "learning_rate": 2e-05,
+      "loss": 0.02305823,
+      "step": 16429
+    },
+    {
+      "epoch": 32.86,
+      "grad_norm": 1.1719574928283691,
+      "learning_rate": 2e-05,
+      "loss": 0.04331547,
+      "step": 16430
+    },
+    {
+      "epoch": 32.862,
+      "grad_norm": 1.6855008602142334,
+      "learning_rate": 2e-05,
+      "loss": 0.04401325,
+      "step": 16431
+    },
+    {
+      "epoch": 32.864,
+      "grad_norm": 1.4206801652908325,
+      "learning_rate": 2e-05,
+      "loss": 0.04856441,
+      "step": 16432
+    },
+    {
+      "epoch": 32.866,
+      "grad_norm": 1.7253879308700562,
+      "learning_rate": 2e-05,
+      "loss": 0.03633871,
+      "step": 16433
+    },
+    {
+      "epoch": 32.868,
+      "grad_norm": 1.3289132118225098,
+      "learning_rate": 2e-05,
+      "loss": 0.03001435,
+      "step": 16434
+    },
+    {
+      "epoch": 32.87,
+      "grad_norm": 1.1799901723861694,
+      "learning_rate": 2e-05,
+      "loss": 0.03081987,
+      "step": 16435
+    },
+    {
+      "epoch": 32.872,
+      "grad_norm": 1.9092236757278442,
+      "learning_rate": 2e-05,
+      "loss": 0.03654516,
+      "step": 16436
+    },
+    {
+      "epoch": 32.874,
+      "grad_norm": 1.1453834772109985,
+      "learning_rate": 2e-05,
+      "loss": 0.04329348,
+      "step": 16437
+    },
+    {
+      "epoch": 32.876,
+      "grad_norm": 1.3431400060653687,
+      "learning_rate": 2e-05,
+      "loss": 0.04679033,
+      "step": 16438
+    },
+    {
+      "epoch": 32.878,
+      "grad_norm": 1.5955402851104736,
+      "learning_rate": 2e-05,
+      "loss": 0.07649103,
+      "step": 16439
+    },
+    {
+      "epoch": 32.88,
+      "grad_norm": 1.485243797302246,
+      "learning_rate": 2e-05,
+      "loss": 0.03790029,
+      "step": 16440
+    },
+    {
+      "epoch": 32.882,
+      "grad_norm": 1.3634883165359497,
+      "learning_rate": 2e-05,
+      "loss": 0.06115822,
+      "step": 16441
+    },
+    {
+      "epoch": 32.884,
+      "grad_norm": 1.1459612846374512,
+      "learning_rate": 2e-05,
+      "loss": 0.04496354,
+      "step": 16442
+    },
+    {
+      "epoch": 32.886,
+      "grad_norm": 1.2098444700241089,
+      "learning_rate": 2e-05,
+      "loss": 0.03849549,
+      "step": 16443
+    },
+    {
+      "epoch": 32.888,
+      "grad_norm": 1.6543594598770142,
+      "learning_rate": 2e-05,
+      "loss": 0.05749209,
+      "step": 16444
+    },
+    {
+      "epoch": 32.89,
+      "grad_norm": 1.1916124820709229,
+      "learning_rate": 2e-05,
+      "loss": 0.02707171,
+      "step": 16445
+    },
+    {
+      "epoch": 32.892,
+      "grad_norm": 1.328437328338623,
+      "learning_rate": 2e-05,
+      "loss": 0.04334527,
+      "step": 16446
+    },
+    {
+      "epoch": 32.894,
+      "grad_norm": 1.472269058227539,
+      "learning_rate": 2e-05,
+      "loss": 0.02811809,
+      "step": 16447
+    },
+    {
+      "epoch": 32.896,
+      "grad_norm": 1.1660832166671753,
+      "learning_rate": 2e-05,
+      "loss": 0.04044122,
+      "step": 16448
+    },
+    {
+      "epoch": 32.898,
+      "grad_norm": 4.141280174255371,
+      "learning_rate": 2e-05,
+      "loss": 0.04939477,
+      "step": 16449
+    },
+    {
+      "epoch": 32.9,
+      "grad_norm": 1.0704823732376099,
+      "learning_rate": 2e-05,
+      "loss": 0.03728681,
+      "step": 16450
+    },
+    {
+      "epoch": 32.902,
+      "grad_norm": 1.1379095315933228,
+      "learning_rate": 2e-05,
+      "loss": 0.04447804,
+      "step": 16451
+    },
+    {
+      "epoch": 32.904,
+      "grad_norm": 1.5318787097930908,
+      "learning_rate": 2e-05,
+      "loss": 0.03792928,
+      "step": 16452
+    },
+    {
+      "epoch": 32.906,
+      "grad_norm": 1.7072391510009766,
+      "learning_rate": 2e-05,
+      "loss": 0.03659452,
+      "step": 16453
+    },
+    {
+      "epoch": 32.908,
+      "grad_norm": 1.2457821369171143,
+      "learning_rate": 2e-05,
+      "loss": 0.03751954,
+      "step": 16454
+    },
+    {
+      "epoch": 32.91,
+      "grad_norm": 1.167049527168274,
+      "learning_rate": 2e-05,
+      "loss": 0.03698363,
+      "step": 16455
+    },
+    {
+      "epoch": 32.912,
+      "grad_norm": 1.0011647939682007,
+      "learning_rate": 2e-05,
+      "loss": 0.03130367,
+      "step": 16456
+    },
+    {
+      "epoch": 32.914,
+      "grad_norm": 0.9641328454017639,
+      "learning_rate": 2e-05,
+      "loss": 0.03359494,
+      "step": 16457
+    },
+    {
+      "epoch": 32.916,
+      "grad_norm": 1.1731847524642944,
+      "learning_rate": 2e-05,
+      "loss": 0.02458394,
+      "step": 16458
+    },
+    {
+      "epoch": 32.918,
+      "grad_norm": 1.1044111251831055,
+      "learning_rate": 2e-05,
+      "loss": 0.04152217,
+      "step": 16459
+    },
+    {
+      "epoch": 32.92,
+      "grad_norm": 1.6179026365280151,
+      "learning_rate": 2e-05,
+      "loss": 0.03974091,
+      "step": 16460
+    },
+    {
+      "epoch": 32.922,
+      "grad_norm": 1.6373225450515747,
+      "learning_rate": 2e-05,
+      "loss": 0.05506727,
+      "step": 16461
+    },
+    {
+      "epoch": 32.924,
+      "grad_norm": 1.1272046566009521,
+      "learning_rate": 2e-05,
+      "loss": 0.03284402,
+      "step": 16462
+    },
+    {
+      "epoch": 32.926,
+      "grad_norm": 1.5268757343292236,
+      "learning_rate": 2e-05,
+      "loss": 0.05714603,
+      "step": 16463
+    },
+    {
+      "epoch": 32.928,
+      "grad_norm": 1.1075844764709473,
+      "learning_rate": 2e-05,
+      "loss": 0.03451537,
+      "step": 16464
+    },
+    {
+      "epoch": 32.93,
+      "grad_norm": 1.25146484375,
+      "learning_rate": 2e-05,
+      "loss": 0.05552392,
+      "step": 16465
+    },
+    {
+      "epoch": 32.932,
+      "grad_norm": 1.9479658603668213,
+      "learning_rate": 2e-05,
+      "loss": 0.06264338,
+      "step": 16466
+    },
+    {
+      "epoch": 32.934,
+      "grad_norm": 1.2056105136871338,
+      "learning_rate": 2e-05,
+      "loss": 0.05005362,
+      "step": 16467
+    },
+    {
+      "epoch": 32.936,
+      "grad_norm": 1.3711949586868286,
+      "learning_rate": 2e-05,
+      "loss": 0.04337479,
+      "step": 16468
+    },
+    {
+      "epoch": 32.938,
+      "grad_norm": 1.2532182931900024,
+      "learning_rate": 2e-05,
+      "loss": 0.04488462,
+      "step": 16469
+    },
+    {
+      "epoch": 32.94,
+      "grad_norm": 1.350342035293579,
+      "learning_rate": 2e-05,
+      "loss": 0.04863258,
+      "step": 16470
+    },
+    {
+      "epoch": 32.942,
+      "grad_norm": 1.5152119398117065,
+      "learning_rate": 2e-05,
+      "loss": 0.03797042,
+      "step": 16471
+    },
+    {
+      "epoch": 32.944,
+      "grad_norm": 1.1705118417739868,
+      "learning_rate": 2e-05,
+      "loss": 0.04442852,
+      "step": 16472
+    },
+    {
+      "epoch": 32.946,
+      "grad_norm": 1.4079266786575317,
+      "learning_rate": 2e-05,
+      "loss": 0.0416984,
+      "step": 16473
+    },
+    {
+      "epoch": 32.948,
+      "grad_norm": 1.1473650932312012,
+      "learning_rate": 2e-05,
+      "loss": 0.0417373,
+      "step": 16474
+    },
+    {
+      "epoch": 32.95,
+      "grad_norm": 1.14018976688385,
+      "learning_rate": 2e-05,
+      "loss": 0.03788053,
+      "step": 16475
+    },
+    {
+      "epoch": 32.952,
+      "grad_norm": 1.0621168613433838,
+      "learning_rate": 2e-05,
+      "loss": 0.02839708,
+      "step": 16476
+    },
+    {
+      "epoch": 32.954,
+      "grad_norm": 1.2420809268951416,
+      "learning_rate": 2e-05,
+      "loss": 0.04673456,
+      "step": 16477
+    },
+    {
+      "epoch": 32.956,
+      "grad_norm": 1.4332877397537231,
+      "learning_rate": 2e-05,
+      "loss": 0.05584753,
+      "step": 16478
+    },
+    {
+      "epoch": 32.958,
+      "grad_norm": 0.9597718119621277,
+      "learning_rate": 2e-05,
+      "loss": 0.03618815,
+      "step": 16479
+    },
+    {
+      "epoch": 32.96,
+      "grad_norm": 1.8332748413085938,
+      "learning_rate": 2e-05,
+      "loss": 0.04130953,
+      "step": 16480
+    },
+    {
+      "epoch": 32.962,
+      "grad_norm": 3.7896127700805664,
+      "learning_rate": 2e-05,
+      "loss": 0.048007,
+      "step": 16481
+    },
+    {
+      "epoch": 32.964,
+      "grad_norm": 1.2358813285827637,
+      "learning_rate": 2e-05,
+      "loss": 0.04505859,
+      "step": 16482
+    },
+    {
+      "epoch": 32.966,
+      "grad_norm": 1.2916022539138794,
+      "learning_rate": 2e-05,
+      "loss": 0.02319424,
+      "step": 16483
+    },
+    {
+      "epoch": 32.968,
+      "grad_norm": 1.1144905090332031,
+      "learning_rate": 2e-05,
+      "loss": 0.04152855,
+      "step": 16484
+    },
+    {
+      "epoch": 32.97,
+      "grad_norm": 1.2684904336929321,
+      "learning_rate": 2e-05,
+      "loss": 0.04438142,
+      "step": 16485
+    },
+    {
+      "epoch": 32.972,
+      "grad_norm": 2.002365827560425,
+      "learning_rate": 2e-05,
+      "loss": 0.06185231,
+      "step": 16486
+    },
+    {
+      "epoch": 32.974,
+      "grad_norm": 1.3114463090896606,
+      "learning_rate": 2e-05,
+      "loss": 0.0507057,
+      "step": 16487
+    },
+    {
+      "epoch": 32.976,
+      "grad_norm": 0.9986602663993835,
+      "learning_rate": 2e-05,
+      "loss": 0.04370682,
+      "step": 16488
+    },
+    {
+      "epoch": 32.978,
+      "grad_norm": 0.9797910451889038,
+      "learning_rate": 2e-05,
+      "loss": 0.03410993,
+      "step": 16489
+    },
+    {
+      "epoch": 32.98,
+      "grad_norm": 1.8880815505981445,
+      "learning_rate": 2e-05,
+      "loss": 0.050763,
+      "step": 16490
+    },
+    {
+      "epoch": 32.982,
+      "grad_norm": 1.241797685623169,
+      "learning_rate": 2e-05,
+      "loss": 0.05129294,
+      "step": 16491
+    },
+    {
+      "epoch": 32.984,
+      "grad_norm": 1.9726879596710205,
+      "learning_rate": 2e-05,
+      "loss": 0.06170304,
+      "step": 16492
+    },
+    {
+      "epoch": 32.986,
+      "grad_norm": 1.1492379903793335,
+      "learning_rate": 2e-05,
+      "loss": 0.05158865,
+      "step": 16493
+    },
+    {
+      "epoch": 32.988,
+      "grad_norm": 1.4059292078018188,
+      "learning_rate": 2e-05,
+      "loss": 0.04178857,
+      "step": 16494
+    },
+    {
+      "epoch": 32.99,
+      "grad_norm": 1.3059656620025635,
+      "learning_rate": 2e-05,
+      "loss": 0.03553785,
+      "step": 16495
+    },
+    {
+      "epoch": 32.992,
+      "grad_norm": 1.3720580339431763,
+      "learning_rate": 2e-05,
+      "loss": 0.04478185,
+      "step": 16496
+    },
+    {
+      "epoch": 32.994,
+      "grad_norm": 1.007232427597046,
+      "learning_rate": 2e-05,
+      "loss": 0.03115787,
+      "step": 16497
+    },
+    {
+      "epoch": 32.996,
+      "grad_norm": 1.1226294040679932,
+      "learning_rate": 2e-05,
+      "loss": 0.05025791,
+      "step": 16498
+    },
+    {
+      "epoch": 32.998,
+      "grad_norm": 1.0878342390060425,
+      "learning_rate": 2e-05,
+      "loss": 0.04733447,
+      "step": 16499
+    },
+    {
+      "epoch": 33.0,
+      "grad_norm": 1.013754963874817,
+      "learning_rate": 2e-05,
+      "loss": 0.03848787,
+      "step": 16500
+    },
+    {
+      "epoch": 33.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.998,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9720558882235529,
+        "Equal_1": 1.0,
+        "Equal_2": 0.9800399201596807,
+        "Equal_3": 0.9840319361277445,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9960079840319361,
+        "Parallel_1": 0.9859719438877755,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.992,
+        "Perpendicular_1": 0.996,
+        "Perpendicular_2": 0.988,
+        "Perpendicular_3": 0.9138276553106213,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 0.9996666666666667,
+        "PointLiesOnCircle_3": 0.992,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9920159680638723
+      },
+      "eval_runtime": 320.8236,
+      "eval_samples_per_second": 32.728,
+      "eval_steps_per_second": 0.655,
+      "step": 16500
+    },
+    {
+      "epoch": 33.002,
+      "grad_norm": 1.0145152807235718,
+      "learning_rate": 2e-05,
+      "loss": 0.03352314,
+      "step": 16501
+    },
+    {
+      "epoch": 33.004,
+      "grad_norm": 2.6528162956237793,
+      "learning_rate": 2e-05,
+      "loss": 0.04622073,
+      "step": 16502
+    },
+    {
+      "epoch": 33.006,
+      "grad_norm": 1.5733745098114014,
+      "learning_rate": 2e-05,
+      "loss": 0.04280098,
+      "step": 16503
+    },
+    {
+      "epoch": 33.008,
+      "grad_norm": 2.088067054748535,
+      "learning_rate": 2e-05,
+      "loss": 0.04422682,
+      "step": 16504
+    },
+    {
+      "epoch": 33.01,
+      "grad_norm": 0.7775627970695496,
+      "learning_rate": 2e-05,
+      "loss": 0.01831872,
+      "step": 16505
+    },
+    {
+      "epoch": 33.012,
+      "grad_norm": 0.9788329601287842,
+      "learning_rate": 2e-05,
+      "loss": 0.03603578,
+      "step": 16506
+    },
+    {
+      "epoch": 33.014,
+      "grad_norm": 1.1010586023330688,
+      "learning_rate": 2e-05,
+      "loss": 0.03720935,
+      "step": 16507
+    },
+    {
+      "epoch": 33.016,
+      "grad_norm": 1.1118940114974976,
+      "learning_rate": 2e-05,
+      "loss": 0.04310315,
+      "step": 16508
+    },
+    {
+      "epoch": 33.018,
+      "grad_norm": 1.3882803916931152,
+      "learning_rate": 2e-05,
+      "loss": 0.04679479,
+      "step": 16509
+    },
+    {
+      "epoch": 33.02,
+      "grad_norm": 1.1422741413116455,
+      "learning_rate": 2e-05,
+      "loss": 0.03853836,
+      "step": 16510
+    },
+    {
+      "epoch": 33.022,
+      "grad_norm": 0.9372391104698181,
+      "learning_rate": 2e-05,
+      "loss": 0.0216747,
+      "step": 16511
+    },
+    {
+      "epoch": 33.024,
+      "grad_norm": 1.2110828161239624,
+      "learning_rate": 2e-05,
+      "loss": 0.04518433,
+      "step": 16512
+    },
+    {
+      "epoch": 33.026,
+      "grad_norm": 1.14997136592865,
+      "learning_rate": 2e-05,
+      "loss": 0.04907893,
+      "step": 16513
+    },
+    {
+      "epoch": 33.028,
+      "grad_norm": 1.4301350116729736,
+      "learning_rate": 2e-05,
+      "loss": 0.05764885,
+      "step": 16514
+    },
+    {
+      "epoch": 33.03,
+      "grad_norm": 1.0221556425094604,
+      "learning_rate": 2e-05,
+      "loss": 0.04257384,
+      "step": 16515
+    },
+    {
+      "epoch": 33.032,
+      "grad_norm": 1.0143057107925415,
+      "learning_rate": 2e-05,
+      "loss": 0.03372315,
+      "step": 16516
+    },
+    {
+      "epoch": 33.034,
+      "grad_norm": 1.1732292175292969,
+      "learning_rate": 2e-05,
+      "loss": 0.04997346,
+      "step": 16517
+    },
+    {
+      "epoch": 33.036,
+      "grad_norm": 1.0122156143188477,
+      "learning_rate": 2e-05,
+      "loss": 0.03192022,
+      "step": 16518
+    },
+    {
+      "epoch": 33.038,
+      "grad_norm": 1.9777421951293945,
+      "learning_rate": 2e-05,
+      "loss": 0.03694803,
+      "step": 16519
+    },
+    {
+      "epoch": 33.04,
+      "grad_norm": 1.1994304656982422,
+      "learning_rate": 2e-05,
+      "loss": 0.05166928,
+      "step": 16520
+    },
+    {
+      "epoch": 33.042,
+      "grad_norm": 0.8912333250045776,
+      "learning_rate": 2e-05,
+      "loss": 0.03158681,
+      "step": 16521
+    },
+    {
+      "epoch": 33.044,
+      "grad_norm": 1.4108567237854004,
+      "learning_rate": 2e-05,
+      "loss": 0.03951647,
+      "step": 16522
+    },
+    {
+      "epoch": 33.046,
+      "grad_norm": 1.5426366329193115,
+      "learning_rate": 2e-05,
+      "loss": 0.05633399,
+      "step": 16523
+    },
+    {
+      "epoch": 33.048,
+      "grad_norm": 1.044606328010559,
+      "learning_rate": 2e-05,
+      "loss": 0.03324772,
+      "step": 16524
+    },
+    {
+      "epoch": 33.05,
+      "grad_norm": 2.6143667697906494,
+      "learning_rate": 2e-05,
+      "loss": 0.06454179,
+      "step": 16525
+    },
+    {
+      "epoch": 33.052,
+      "grad_norm": 1.6238820552825928,
+      "learning_rate": 2e-05,
+      "loss": 0.02863217,
+      "step": 16526
+    },
+    {
+      "epoch": 33.054,
+      "grad_norm": 1.1710536479949951,
+      "learning_rate": 2e-05,
+      "loss": 0.05588196,
+      "step": 16527
+    },
+    {
+      "epoch": 33.056,
+      "grad_norm": 1.5752817392349243,
+      "learning_rate": 2e-05,
+      "loss": 0.03416557,
+      "step": 16528
+    },
+    {
+      "epoch": 33.058,
+      "grad_norm": 1.6088008880615234,
+      "learning_rate": 2e-05,
+      "loss": 0.05284679,
+      "step": 16529
+    },
+    {
+      "epoch": 33.06,
+      "grad_norm": 1.1662729978561401,
+      "learning_rate": 2e-05,
+      "loss": 0.03968053,
+      "step": 16530
+    },
+    {
+      "epoch": 33.062,
+      "grad_norm": 1.0963259935379028,
+      "learning_rate": 2e-05,
+      "loss": 0.04498053,
+      "step": 16531
+    },
+    {
+      "epoch": 33.064,
+      "grad_norm": 0.9817715287208557,
+      "learning_rate": 2e-05,
+      "loss": 0.03069879,
+      "step": 16532
+    },
+    {
+      "epoch": 33.066,
+      "grad_norm": 1.3074853420257568,
+      "learning_rate": 2e-05,
+      "loss": 0.06249923,
+      "step": 16533
+    },
+    {
+      "epoch": 33.068,
+      "grad_norm": 0.9941977262496948,
+      "learning_rate": 2e-05,
+      "loss": 0.03796024,
+      "step": 16534
+    },
+    {
+      "epoch": 33.07,
+      "grad_norm": 1.546793818473816,
+      "learning_rate": 2e-05,
+      "loss": 0.04255424,
+      "step": 16535
+    },
+    {
+      "epoch": 33.072,
+      "grad_norm": 1.2092461585998535,
+      "learning_rate": 2e-05,
+      "loss": 0.04925638,
+      "step": 16536
+    },
+    {
+      "epoch": 33.074,
+      "grad_norm": 1.3330050706863403,
+      "learning_rate": 2e-05,
+      "loss": 0.04654669,
+      "step": 16537
+    },
+    {
+      "epoch": 33.076,
+      "grad_norm": 1.539285659790039,
+      "learning_rate": 2e-05,
+      "loss": 0.04258667,
+      "step": 16538
+    },
+    {
+      "epoch": 33.078,
+      "grad_norm": 1.1344162225723267,
+      "learning_rate": 2e-05,
+      "loss": 0.03024994,
+      "step": 16539
+    },
+    {
+      "epoch": 33.08,
+      "grad_norm": 1.3139582872390747,
+      "learning_rate": 2e-05,
+      "loss": 0.06406873,
+      "step": 16540
+    },
+    {
+      "epoch": 33.082,
+      "grad_norm": 1.342620611190796,
+      "learning_rate": 2e-05,
+      "loss": 0.04298013,
+      "step": 16541
+    },
+    {
+      "epoch": 33.084,
+      "grad_norm": 1.1698533296585083,
+      "learning_rate": 2e-05,
+      "loss": 0.0356679,
+      "step": 16542
+    },
+    {
+      "epoch": 33.086,
+      "grad_norm": 1.3078049421310425,
+      "learning_rate": 2e-05,
+      "loss": 0.03047599,
+      "step": 16543
+    },
+    {
+      "epoch": 33.088,
+      "grad_norm": 1.1613730192184448,
+      "learning_rate": 2e-05,
+      "loss": 0.04012228,
+      "step": 16544
+    },
+    {
+      "epoch": 33.09,
+      "grad_norm": 1.5841779708862305,
+      "learning_rate": 2e-05,
+      "loss": 0.05739962,
+      "step": 16545
+    },
+    {
+      "epoch": 33.092,
+      "grad_norm": 1.175606608390808,
+      "learning_rate": 2e-05,
+      "loss": 0.0417013,
+      "step": 16546
+    },
+    {
+      "epoch": 33.094,
+      "grad_norm": 1.0316969156265259,
+      "learning_rate": 2e-05,
+      "loss": 0.02844536,
+      "step": 16547
+    },
+    {
+      "epoch": 33.096,
+      "grad_norm": 1.6557471752166748,
+      "learning_rate": 2e-05,
+      "loss": 0.04946776,
+      "step": 16548
+    },
+    {
+      "epoch": 33.098,
+      "grad_norm": 1.2517582178115845,
+      "learning_rate": 2e-05,
+      "loss": 0.03600208,
+      "step": 16549
+    },
+    {
+      "epoch": 33.1,
+      "grad_norm": 1.3024152517318726,
+      "learning_rate": 2e-05,
+      "loss": 0.0431541,
+      "step": 16550
+    },
+    {
+      "epoch": 33.102,
+      "grad_norm": 1.2932987213134766,
+      "learning_rate": 2e-05,
+      "loss": 0.04457521,
+      "step": 16551
+    },
+    {
+      "epoch": 33.104,
+      "grad_norm": 1.2561955451965332,
+      "learning_rate": 2e-05,
+      "loss": 0.05188843,
+      "step": 16552
+    },
+    {
+      "epoch": 33.106,
+      "grad_norm": 1.1888597011566162,
+      "learning_rate": 2e-05,
+      "loss": 0.03158218,
+      "step": 16553
+    },
+    {
+      "epoch": 33.108,
+      "grad_norm": 1.1802350282669067,
+      "learning_rate": 2e-05,
+      "loss": 0.04162861,
+      "step": 16554
+    },
+    {
+      "epoch": 33.11,
+      "grad_norm": 1.2707432508468628,
+      "learning_rate": 2e-05,
+      "loss": 0.0347266,
+      "step": 16555
+    },
+    {
+      "epoch": 33.112,
+      "grad_norm": 1.2336151599884033,
+      "learning_rate": 2e-05,
+      "loss": 0.03552887,
+      "step": 16556
+    },
+    {
+      "epoch": 33.114,
+      "grad_norm": 1.348934292793274,
+      "learning_rate": 2e-05,
+      "loss": 0.04676005,
+      "step": 16557
+    },
+    {
+      "epoch": 33.116,
+      "grad_norm": 0.9855770468711853,
+      "learning_rate": 2e-05,
+      "loss": 0.02963222,
+      "step": 16558
+    },
+    {
+      "epoch": 33.118,
+      "grad_norm": 0.9198537468910217,
+      "learning_rate": 2e-05,
+      "loss": 0.03035306,
+      "step": 16559
+    },
+    {
+      "epoch": 33.12,
+      "grad_norm": 1.309364914894104,
+      "learning_rate": 2e-05,
+      "loss": 0.05264792,
+      "step": 16560
+    },
+    {
+      "epoch": 33.122,
+      "grad_norm": 1.5512858629226685,
+      "learning_rate": 2e-05,
+      "loss": 0.05176829,
+      "step": 16561
+    },
+    {
+      "epoch": 33.124,
+      "grad_norm": 1.2333922386169434,
+      "learning_rate": 2e-05,
+      "loss": 0.04524909,
+      "step": 16562
+    },
+    {
+      "epoch": 33.126,
+      "grad_norm": 1.252514362335205,
+      "learning_rate": 2e-05,
+      "loss": 0.0393767,
+      "step": 16563
+    },
+    {
+      "epoch": 33.128,
+      "grad_norm": 0.95073002576828,
+      "learning_rate": 2e-05,
+      "loss": 0.0251915,
+      "step": 16564
+    },
+    {
+      "epoch": 33.13,
+      "grad_norm": 2.1087992191314697,
+      "learning_rate": 2e-05,
+      "loss": 0.04168628,
+      "step": 16565
+    },
+    {
+      "epoch": 33.132,
+      "grad_norm": 1.073662519454956,
+      "learning_rate": 2e-05,
+      "loss": 0.03738192,
+      "step": 16566
+    },
+    {
+      "epoch": 33.134,
+      "grad_norm": 2.494189977645874,
+      "learning_rate": 2e-05,
+      "loss": 0.03791768,
+      "step": 16567
+    },
+    {
+      "epoch": 33.136,
+      "grad_norm": 2.3711354732513428,
+      "learning_rate": 2e-05,
+      "loss": 0.05274618,
+      "step": 16568
+    },
+    {
+      "epoch": 33.138,
+      "grad_norm": 1.1939685344696045,
+      "learning_rate": 2e-05,
+      "loss": 0.05177773,
+      "step": 16569
+    },
+    {
+      "epoch": 33.14,
+      "grad_norm": 1.0729763507843018,
+      "learning_rate": 2e-05,
+      "loss": 0.04366021,
+      "step": 16570
+    },
+    {
+      "epoch": 33.142,
+      "grad_norm": 1.9815735816955566,
+      "learning_rate": 2e-05,
+      "loss": 0.05311869,
+      "step": 16571
+    },
+    {
+      "epoch": 33.144,
+      "grad_norm": 2.0230648517608643,
+      "learning_rate": 2e-05,
+      "loss": 0.05889687,
+      "step": 16572
+    },
+    {
+      "epoch": 33.146,
+      "grad_norm": 0.9048420190811157,
+      "learning_rate": 2e-05,
+      "loss": 0.02796107,
+      "step": 16573
+    },
+    {
+      "epoch": 33.148,
+      "grad_norm": 1.2179429531097412,
+      "learning_rate": 2e-05,
+      "loss": 0.04486234,
+      "step": 16574
+    },
+    {
+      "epoch": 33.15,
+      "grad_norm": 1.1371427774429321,
+      "learning_rate": 2e-05,
+      "loss": 0.03180134,
+      "step": 16575
+    },
+    {
+      "epoch": 33.152,
+      "grad_norm": 1.2799369096755981,
+      "learning_rate": 2e-05,
+      "loss": 0.04838013,
+      "step": 16576
+    },
+    {
+      "epoch": 33.154,
+      "grad_norm": 1.4155060052871704,
+      "learning_rate": 2e-05,
+      "loss": 0.03349139,
+      "step": 16577
+    },
+    {
+      "epoch": 33.156,
+      "grad_norm": 1.009975790977478,
+      "learning_rate": 2e-05,
+      "loss": 0.02858535,
+      "step": 16578
+    },
+    {
+      "epoch": 33.158,
+      "grad_norm": 1.0067310333251953,
+      "learning_rate": 2e-05,
+      "loss": 0.02488973,
+      "step": 16579
+    },
+    {
+      "epoch": 33.16,
+      "grad_norm": 1.6880412101745605,
+      "learning_rate": 2e-05,
+      "loss": 0.04501157,
+      "step": 16580
+    },
+    {
+      "epoch": 33.162,
+      "grad_norm": 1.3325822353363037,
+      "learning_rate": 2e-05,
+      "loss": 0.0444175,
+      "step": 16581
+    },
+    {
+      "epoch": 33.164,
+      "grad_norm": 5.777720928192139,
+      "learning_rate": 2e-05,
+      "loss": 0.0530759,
+      "step": 16582
+    },
+    {
+      "epoch": 33.166,
+      "grad_norm": 2.4529478549957275,
+      "learning_rate": 2e-05,
+      "loss": 0.04354583,
+      "step": 16583
+    },
+    {
+      "epoch": 33.168,
+      "grad_norm": 1.3145490884780884,
+      "learning_rate": 2e-05,
+      "loss": 0.03570919,
+      "step": 16584
+    },
+    {
+      "epoch": 33.17,
+      "grad_norm": 1.2409156560897827,
+      "learning_rate": 2e-05,
+      "loss": 0.04724094,
+      "step": 16585
+    },
+    {
+      "epoch": 33.172,
+      "grad_norm": 0.8115535974502563,
+      "learning_rate": 2e-05,
+      "loss": 0.03618132,
+      "step": 16586
+    },
+    {
+      "epoch": 33.174,
+      "grad_norm": 0.9847850799560547,
+      "learning_rate": 2e-05,
+      "loss": 0.03718961,
+      "step": 16587
+    },
+    {
+      "epoch": 33.176,
+      "grad_norm": 1.0901992321014404,
+      "learning_rate": 2e-05,
+      "loss": 0.03891124,
+      "step": 16588
+    },
+    {
+      "epoch": 33.178,
+      "grad_norm": 1.2606801986694336,
+      "learning_rate": 2e-05,
+      "loss": 0.04655527,
+      "step": 16589
+    },
+    {
+      "epoch": 33.18,
+      "grad_norm": 1.6437278985977173,
+      "learning_rate": 2e-05,
+      "loss": 0.04169775,
+      "step": 16590
+    },
+    {
+      "epoch": 33.182,
+      "grad_norm": 1.0829734802246094,
+      "learning_rate": 2e-05,
+      "loss": 0.03909619,
+      "step": 16591
+    },
+    {
+      "epoch": 33.184,
+      "grad_norm": 1.9588395357131958,
+      "learning_rate": 2e-05,
+      "loss": 0.05959759,
+      "step": 16592
+    },
+    {
+      "epoch": 33.186,
+      "grad_norm": 2.825683355331421,
+      "learning_rate": 2e-05,
+      "loss": 0.05063524,
+      "step": 16593
+    },
+    {
+      "epoch": 33.188,
+      "grad_norm": 1.0350728034973145,
+      "learning_rate": 2e-05,
+      "loss": 0.03315126,
+      "step": 16594
+    },
+    {
+      "epoch": 33.19,
+      "grad_norm": 2.2779905796051025,
+      "learning_rate": 2e-05,
+      "loss": 0.04792159,
+      "step": 16595
+    },
+    {
+      "epoch": 33.192,
+      "grad_norm": 1.0446611642837524,
+      "learning_rate": 2e-05,
+      "loss": 0.03796531,
+      "step": 16596
+    },
+    {
+      "epoch": 33.194,
+      "grad_norm": 1.1090408563613892,
+      "learning_rate": 2e-05,
+      "loss": 0.03750544,
+      "step": 16597
+    },
+    {
+      "epoch": 33.196,
+      "grad_norm": 1.0616990327835083,
+      "learning_rate": 2e-05,
+      "loss": 0.04366959,
+      "step": 16598
+    },
+    {
+      "epoch": 33.198,
+      "grad_norm": 1.0580826997756958,
+      "learning_rate": 2e-05,
+      "loss": 0.03785151,
+      "step": 16599
+    },
+    {
+      "epoch": 33.2,
+      "grad_norm": 1.757721185684204,
+      "learning_rate": 2e-05,
+      "loss": 0.04076345,
+      "step": 16600
+    },
+    {
+      "epoch": 33.202,
+      "grad_norm": 1.2572243213653564,
+      "learning_rate": 2e-05,
+      "loss": 0.04631349,
+      "step": 16601
+    },
+    {
+      "epoch": 33.204,
+      "grad_norm": 1.4636191129684448,
+      "learning_rate": 2e-05,
+      "loss": 0.04375532,
+      "step": 16602
+    },
+    {
+      "epoch": 33.206,
+      "grad_norm": 1.7812341451644897,
+      "learning_rate": 2e-05,
+      "loss": 0.03871331,
+      "step": 16603
+    },
+    {
+      "epoch": 33.208,
+      "grad_norm": 1.4072896242141724,
+      "learning_rate": 2e-05,
+      "loss": 0.05194937,
+      "step": 16604
+    },
+    {
+      "epoch": 33.21,
+      "grad_norm": 0.9747751951217651,
+      "learning_rate": 2e-05,
+      "loss": 0.02706435,
+      "step": 16605
+    },
+    {
+      "epoch": 33.212,
+      "grad_norm": 1.1808767318725586,
+      "learning_rate": 2e-05,
+      "loss": 0.03670924,
+      "step": 16606
+    },
+    {
+      "epoch": 33.214,
+      "grad_norm": 1.2573798894882202,
+      "learning_rate": 2e-05,
+      "loss": 0.03263208,
+      "step": 16607
+    },
+    {
+      "epoch": 33.216,
+      "grad_norm": 1.379146695137024,
+      "learning_rate": 2e-05,
+      "loss": 0.04529354,
+      "step": 16608
+    },
+    {
+      "epoch": 33.218,
+      "grad_norm": 1.177720308303833,
+      "learning_rate": 2e-05,
+      "loss": 0.03827986,
+      "step": 16609
+    },
+    {
+      "epoch": 33.22,
+      "grad_norm": 1.1560696363449097,
+      "learning_rate": 2e-05,
+      "loss": 0.04032604,
+      "step": 16610
+    },
+    {
+      "epoch": 33.222,
+      "grad_norm": 1.2086445093154907,
+      "learning_rate": 2e-05,
+      "loss": 0.04973095,
+      "step": 16611
+    },
+    {
+      "epoch": 33.224,
+      "grad_norm": 1.0575134754180908,
+      "learning_rate": 2e-05,
+      "loss": 0.03815421,
+      "step": 16612
+    },
+    {
+      "epoch": 33.226,
+      "grad_norm": 1.3156135082244873,
+      "learning_rate": 2e-05,
+      "loss": 0.0418507,
+      "step": 16613
+    },
+    {
+      "epoch": 33.228,
+      "grad_norm": 1.3325525522232056,
+      "learning_rate": 2e-05,
+      "loss": 0.0365724,
+      "step": 16614
+    },
+    {
+      "epoch": 33.23,
+      "grad_norm": 1.4213200807571411,
+      "learning_rate": 2e-05,
+      "loss": 0.04568385,
+      "step": 16615
+    },
+    {
+      "epoch": 33.232,
+      "grad_norm": 1.8256237506866455,
+      "learning_rate": 2e-05,
+      "loss": 0.06385624,
+      "step": 16616
+    },
+    {
+      "epoch": 33.234,
+      "grad_norm": 2.2751426696777344,
+      "learning_rate": 2e-05,
+      "loss": 0.04987566,
+      "step": 16617
+    },
+    {
+      "epoch": 33.236,
+      "grad_norm": 0.9569184184074402,
+      "learning_rate": 2e-05,
+      "loss": 0.02542327,
+      "step": 16618
+    },
+    {
+      "epoch": 33.238,
+      "grad_norm": 1.0519928932189941,
+      "learning_rate": 2e-05,
+      "loss": 0.03581095,
+      "step": 16619
+    },
+    {
+      "epoch": 33.24,
+      "grad_norm": 1.1077507734298706,
+      "learning_rate": 2e-05,
+      "loss": 0.03992999,
+      "step": 16620
+    },
+    {
+      "epoch": 33.242,
+      "grad_norm": 1.6223715543746948,
+      "learning_rate": 2e-05,
+      "loss": 0.02992079,
+      "step": 16621
+    },
+    {
+      "epoch": 33.244,
+      "grad_norm": 1.5266954898834229,
+      "learning_rate": 2e-05,
+      "loss": 0.03266663,
+      "step": 16622
+    },
+    {
+      "epoch": 33.246,
+      "grad_norm": 1.642143964767456,
+      "learning_rate": 2e-05,
+      "loss": 0.05565649,
+      "step": 16623
+    },
+    {
+      "epoch": 33.248,
+      "grad_norm": 2.065688371658325,
+      "learning_rate": 2e-05,
+      "loss": 0.0459539,
+      "step": 16624
+    },
+    {
+      "epoch": 33.25,
+      "grad_norm": 1.1380778551101685,
+      "learning_rate": 2e-05,
+      "loss": 0.04274027,
+      "step": 16625
+    },
+    {
+      "epoch": 33.252,
+      "grad_norm": 1.23936927318573,
+      "learning_rate": 2e-05,
+      "loss": 0.05491698,
+      "step": 16626
+    },
+    {
+      "epoch": 33.254,
+      "grad_norm": 1.0805511474609375,
+      "learning_rate": 2e-05,
+      "loss": 0.04202371,
+      "step": 16627
+    },
+    {
+      "epoch": 33.256,
+      "grad_norm": 1.1573930978775024,
+      "learning_rate": 2e-05,
+      "loss": 0.04801682,
+      "step": 16628
+    },
+    {
+      "epoch": 33.258,
+      "grad_norm": 1.1957443952560425,
+      "learning_rate": 2e-05,
+      "loss": 0.02971683,
+      "step": 16629
+    },
+    {
+      "epoch": 33.26,
+      "grad_norm": 1.1596754789352417,
+      "learning_rate": 2e-05,
+      "loss": 0.03906877,
+      "step": 16630
+    },
+    {
+      "epoch": 33.262,
+      "grad_norm": 1.0122718811035156,
+      "learning_rate": 2e-05,
+      "loss": 0.0367621,
+      "step": 16631
+    },
+    {
+      "epoch": 33.264,
+      "grad_norm": 1.2783515453338623,
+      "learning_rate": 2e-05,
+      "loss": 0.06680563,
+      "step": 16632
+    },
+    {
+      "epoch": 33.266,
+      "grad_norm": 1.4091695547103882,
+      "learning_rate": 2e-05,
+      "loss": 0.05388515,
+      "step": 16633
+    },
+    {
+      "epoch": 33.268,
+      "grad_norm": 1.1046702861785889,
+      "learning_rate": 2e-05,
+      "loss": 0.04830401,
+      "step": 16634
+    },
+    {
+      "epoch": 33.27,
+      "grad_norm": 1.6029551029205322,
+      "learning_rate": 2e-05,
+      "loss": 0.04491821,
+      "step": 16635
+    },
+    {
+      "epoch": 33.272,
+      "grad_norm": 1.1932250261306763,
+      "learning_rate": 2e-05,
+      "loss": 0.04055645,
+      "step": 16636
+    },
+    {
+      "epoch": 33.274,
+      "grad_norm": 1.7215577363967896,
+      "learning_rate": 2e-05,
+      "loss": 0.04914788,
+      "step": 16637
+    },
+    {
+      "epoch": 33.276,
+      "grad_norm": 0.9664545059204102,
+      "learning_rate": 2e-05,
+      "loss": 0.04409215,
+      "step": 16638
+    },
+    {
+      "epoch": 33.278,
+      "grad_norm": 1.0649373531341553,
+      "learning_rate": 2e-05,
+      "loss": 0.04012241,
+      "step": 16639
+    },
+    {
+      "epoch": 33.28,
+      "grad_norm": 1.5425022840499878,
+      "learning_rate": 2e-05,
+      "loss": 0.0414533,
+      "step": 16640
+    },
+    {
+      "epoch": 33.282,
+      "grad_norm": 1.4415966272354126,
+      "learning_rate": 2e-05,
+      "loss": 0.05605222,
+      "step": 16641
+    },
+    {
+      "epoch": 33.284,
+      "grad_norm": 1.9677499532699585,
+      "learning_rate": 2e-05,
+      "loss": 0.05646674,
+      "step": 16642
+    },
+    {
+      "epoch": 33.286,
+      "grad_norm": 1.0647591352462769,
+      "learning_rate": 2e-05,
+      "loss": 0.04393759,
+      "step": 16643
+    },
+    {
+      "epoch": 33.288,
+      "grad_norm": 0.9724460244178772,
+      "learning_rate": 2e-05,
+      "loss": 0.0201349,
+      "step": 16644
+    },
+    {
+      "epoch": 33.29,
+      "grad_norm": 1.3234846591949463,
+      "learning_rate": 2e-05,
+      "loss": 0.04536884,
+      "step": 16645
+    },
+    {
+      "epoch": 33.292,
+      "grad_norm": 1.315469741821289,
+      "learning_rate": 2e-05,
+      "loss": 0.03901994,
+      "step": 16646
+    },
+    {
+      "epoch": 33.294,
+      "grad_norm": 1.1740449666976929,
+      "learning_rate": 2e-05,
+      "loss": 0.04922675,
+      "step": 16647
+    },
+    {
+      "epoch": 33.296,
+      "grad_norm": 2.73299241065979,
+      "learning_rate": 2e-05,
+      "loss": 0.04908387,
+      "step": 16648
+    },
+    {
+      "epoch": 33.298,
+      "grad_norm": 1.0189504623413086,
+      "learning_rate": 2e-05,
+      "loss": 0.04484309,
+      "step": 16649
+    },
+    {
+      "epoch": 33.3,
+      "grad_norm": 2.3086318969726562,
+      "learning_rate": 2e-05,
+      "loss": 0.06764978,
+      "step": 16650
+    },
+    {
+      "epoch": 33.302,
+      "grad_norm": 0.8751531839370728,
+      "learning_rate": 2e-05,
+      "loss": 0.03591763,
+      "step": 16651
+    },
+    {
+      "epoch": 33.304,
+      "grad_norm": 1.2427144050598145,
+      "learning_rate": 2e-05,
+      "loss": 0.05099815,
+      "step": 16652
+    },
+    {
+      "epoch": 33.306,
+      "grad_norm": 1.8863247632980347,
+      "learning_rate": 2e-05,
+      "loss": 0.0360302,
+      "step": 16653
+    },
+    {
+      "epoch": 33.308,
+      "grad_norm": 1.1457570791244507,
+      "learning_rate": 2e-05,
+      "loss": 0.03447718,
+      "step": 16654
+    },
+    {
+      "epoch": 33.31,
+      "grad_norm": 1.273047685623169,
+      "learning_rate": 2e-05,
+      "loss": 0.03491887,
+      "step": 16655
+    },
+    {
+      "epoch": 33.312,
+      "grad_norm": 1.65647554397583,
+      "learning_rate": 2e-05,
+      "loss": 0.05843513,
+      "step": 16656
+    },
+    {
+      "epoch": 33.314,
+      "grad_norm": 1.1409825086593628,
+      "learning_rate": 2e-05,
+      "loss": 0.04555066,
+      "step": 16657
+    },
+    {
+      "epoch": 33.316,
+      "grad_norm": 3.0305325984954834,
+      "learning_rate": 2e-05,
+      "loss": 0.05683912,
+      "step": 16658
+    },
+    {
+      "epoch": 33.318,
+      "grad_norm": 2.7458958625793457,
+      "learning_rate": 2e-05,
+      "loss": 0.0510499,
+      "step": 16659
+    },
+    {
+      "epoch": 33.32,
+      "grad_norm": 2.914576768875122,
+      "learning_rate": 2e-05,
+      "loss": 0.04127868,
+      "step": 16660
+    },
+    {
+      "epoch": 33.322,
+      "grad_norm": 1.22411048412323,
+      "learning_rate": 2e-05,
+      "loss": 0.0514855,
+      "step": 16661
+    },
+    {
+      "epoch": 33.324,
+      "grad_norm": 1.5475612878799438,
+      "learning_rate": 2e-05,
+      "loss": 0.05710922,
+      "step": 16662
+    },
+    {
+      "epoch": 33.326,
+      "grad_norm": 1.0445101261138916,
+      "learning_rate": 2e-05,
+      "loss": 0.0416135,
+      "step": 16663
+    },
+    {
+      "epoch": 33.328,
+      "grad_norm": 1.168892502784729,
+      "learning_rate": 2e-05,
+      "loss": 0.02608304,
+      "step": 16664
+    },
+    {
+      "epoch": 33.33,
+      "grad_norm": 1.1032596826553345,
+      "learning_rate": 2e-05,
+      "loss": 0.03604908,
+      "step": 16665
+    },
+    {
+      "epoch": 33.332,
+      "grad_norm": 1.2727869749069214,
+      "learning_rate": 2e-05,
+      "loss": 0.06578045,
+      "step": 16666
+    },
+    {
+      "epoch": 33.334,
+      "grad_norm": 0.9599734544754028,
+      "learning_rate": 2e-05,
+      "loss": 0.03464708,
+      "step": 16667
+    },
+    {
+      "epoch": 33.336,
+      "grad_norm": 0.9870487451553345,
+      "learning_rate": 2e-05,
+      "loss": 0.04101057,
+      "step": 16668
+    },
+    {
+      "epoch": 33.338,
+      "grad_norm": 1.5237897634506226,
+      "learning_rate": 2e-05,
+      "loss": 0.06414343,
+      "step": 16669
+    },
+    {
+      "epoch": 33.34,
+      "grad_norm": 1.0481436252593994,
+      "learning_rate": 2e-05,
+      "loss": 0.03079265,
+      "step": 16670
+    },
+    {
+      "epoch": 33.342,
+      "grad_norm": 1.1099966764450073,
+      "learning_rate": 2e-05,
+      "loss": 0.04527031,
+      "step": 16671
+    },
+    {
+      "epoch": 33.344,
+      "grad_norm": 1.3195089101791382,
+      "learning_rate": 2e-05,
+      "loss": 0.02745794,
+      "step": 16672
+    },
+    {
+      "epoch": 33.346,
+      "grad_norm": 0.8879894018173218,
+      "learning_rate": 2e-05,
+      "loss": 0.03119539,
+      "step": 16673
+    },
+    {
+      "epoch": 33.348,
+      "grad_norm": 1.2862367630004883,
+      "learning_rate": 2e-05,
+      "loss": 0.04255934,
+      "step": 16674
+    },
+    {
+      "epoch": 33.35,
+      "grad_norm": 1.123695731163025,
+      "learning_rate": 2e-05,
+      "loss": 0.04763044,
+      "step": 16675
+    },
+    {
+      "epoch": 33.352,
+      "grad_norm": 1.058069109916687,
+      "learning_rate": 2e-05,
+      "loss": 0.03856095,
+      "step": 16676
+    },
+    {
+      "epoch": 33.354,
+      "grad_norm": 1.1873466968536377,
+      "learning_rate": 2e-05,
+      "loss": 0.04556847,
+      "step": 16677
+    },
+    {
+      "epoch": 33.356,
+      "grad_norm": 1.5896145105361938,
+      "learning_rate": 2e-05,
+      "loss": 0.06114259,
+      "step": 16678
+    },
+    {
+      "epoch": 33.358,
+      "grad_norm": 1.1013940572738647,
+      "learning_rate": 2e-05,
+      "loss": 0.04544736,
+      "step": 16679
+    },
+    {
+      "epoch": 33.36,
+      "grad_norm": 1.1660057306289673,
+      "learning_rate": 2e-05,
+      "loss": 0.04236136,
+      "step": 16680
+    },
+    {
+      "epoch": 33.362,
+      "grad_norm": 1.0070010423660278,
+      "learning_rate": 2e-05,
+      "loss": 0.03095785,
+      "step": 16681
+    },
+    {
+      "epoch": 33.364,
+      "grad_norm": 1.2435543537139893,
+      "learning_rate": 2e-05,
+      "loss": 0.05413376,
+      "step": 16682
+    },
+    {
+      "epoch": 33.366,
+      "grad_norm": 1.076140284538269,
+      "learning_rate": 2e-05,
+      "loss": 0.0455126,
+      "step": 16683
+    },
+    {
+      "epoch": 33.368,
+      "grad_norm": 0.9917166829109192,
+      "learning_rate": 2e-05,
+      "loss": 0.03302848,
+      "step": 16684
+    },
+    {
+      "epoch": 33.37,
+      "grad_norm": 1.7563289403915405,
+      "learning_rate": 2e-05,
+      "loss": 0.03998546,
+      "step": 16685
+    },
+    {
+      "epoch": 33.372,
+      "grad_norm": 1.0916879177093506,
+      "learning_rate": 2e-05,
+      "loss": 0.03832256,
+      "step": 16686
+    },
+    {
+      "epoch": 33.374,
+      "grad_norm": 1.0437517166137695,
+      "learning_rate": 2e-05,
+      "loss": 0.04042347,
+      "step": 16687
+    },
+    {
+      "epoch": 33.376,
+      "grad_norm": 2.85295033454895,
+      "learning_rate": 2e-05,
+      "loss": 0.04512815,
+      "step": 16688
+    },
+    {
+      "epoch": 33.378,
+      "grad_norm": 1.3456438779830933,
+      "learning_rate": 2e-05,
+      "loss": 0.05866044,
+      "step": 16689
+    },
+    {
+      "epoch": 33.38,
+      "grad_norm": 1.2579126358032227,
+      "learning_rate": 2e-05,
+      "loss": 0.04194574,
+      "step": 16690
+    },
+    {
+      "epoch": 33.382,
+      "grad_norm": 1.2284680604934692,
+      "learning_rate": 2e-05,
+      "loss": 0.04293683,
+      "step": 16691
+    },
+    {
+      "epoch": 33.384,
+      "grad_norm": 1.0738353729248047,
+      "learning_rate": 2e-05,
+      "loss": 0.03847217,
+      "step": 16692
+    },
+    {
+      "epoch": 33.386,
+      "grad_norm": 0.9319986701011658,
+      "learning_rate": 2e-05,
+      "loss": 0.02663419,
+      "step": 16693
+    },
+    {
+      "epoch": 33.388,
+      "grad_norm": 2.2574198246002197,
+      "learning_rate": 2e-05,
+      "loss": 0.03809108,
+      "step": 16694
+    },
+    {
+      "epoch": 33.39,
+      "grad_norm": 1.3886204957962036,
+      "learning_rate": 2e-05,
+      "loss": 0.04164458,
+      "step": 16695
+    },
+    {
+      "epoch": 33.392,
+      "grad_norm": 1.673781156539917,
+      "learning_rate": 2e-05,
+      "loss": 0.04233087,
+      "step": 16696
+    },
+    {
+      "epoch": 33.394,
+      "grad_norm": 1.2593107223510742,
+      "learning_rate": 2e-05,
+      "loss": 0.04103605,
+      "step": 16697
+    },
+    {
+      "epoch": 33.396,
+      "grad_norm": 2.0994770526885986,
+      "learning_rate": 2e-05,
+      "loss": 0.0581164,
+      "step": 16698
+    },
+    {
+      "epoch": 33.398,
+      "grad_norm": 1.1525026559829712,
+      "learning_rate": 2e-05,
+      "loss": 0.0525818,
+      "step": 16699
+    },
+    {
+      "epoch": 33.4,
+      "grad_norm": 1.141378402709961,
+      "learning_rate": 2e-05,
+      "loss": 0.05826852,
+      "step": 16700
+    },
+    {
+      "epoch": 33.402,
+      "grad_norm": 1.3574016094207764,
+      "learning_rate": 2e-05,
+      "loss": 0.05125176,
+      "step": 16701
+    },
+    {
+      "epoch": 33.404,
+      "grad_norm": 1.2207424640655518,
+      "learning_rate": 2e-05,
+      "loss": 0.03801119,
+      "step": 16702
+    },
+    {
+      "epoch": 33.406,
+      "grad_norm": 1.1650470495224,
+      "learning_rate": 2e-05,
+      "loss": 0.04888697,
+      "step": 16703
+    },
+    {
+      "epoch": 33.408,
+      "grad_norm": 1.325939416885376,
+      "learning_rate": 2e-05,
+      "loss": 0.05456465,
+      "step": 16704
+    },
+    {
+      "epoch": 33.41,
+      "grad_norm": 1.076897382736206,
+      "learning_rate": 2e-05,
+      "loss": 0.04185375,
+      "step": 16705
+    },
+    {
+      "epoch": 33.412,
+      "grad_norm": 1.0933011770248413,
+      "learning_rate": 2e-05,
+      "loss": 0.03941853,
+      "step": 16706
+    },
+    {
+      "epoch": 33.414,
+      "grad_norm": 1.4302270412445068,
+      "learning_rate": 2e-05,
+      "loss": 0.04674981,
+      "step": 16707
+    },
+    {
+      "epoch": 33.416,
+      "grad_norm": 5.180481910705566,
+      "learning_rate": 2e-05,
+      "loss": 0.05447872,
+      "step": 16708
+    },
+    {
+      "epoch": 33.418,
+      "grad_norm": 0.997200071811676,
+      "learning_rate": 2e-05,
+      "loss": 0.04152042,
+      "step": 16709
+    },
+    {
+      "epoch": 33.42,
+      "grad_norm": 1.2832751274108887,
+      "learning_rate": 2e-05,
+      "loss": 0.05069866,
+      "step": 16710
+    },
+    {
+      "epoch": 33.422,
+      "grad_norm": 1.0193212032318115,
+      "learning_rate": 2e-05,
+      "loss": 0.0379979,
+      "step": 16711
+    },
+    {
+      "epoch": 33.424,
+      "grad_norm": 1.0952942371368408,
+      "learning_rate": 2e-05,
+      "loss": 0.03752263,
+      "step": 16712
+    },
+    {
+      "epoch": 33.426,
+      "grad_norm": 1.048114538192749,
+      "learning_rate": 2e-05,
+      "loss": 0.05350174,
+      "step": 16713
+    },
+    {
+      "epoch": 33.428,
+      "grad_norm": 1.142867088317871,
+      "learning_rate": 2e-05,
+      "loss": 0.04559736,
+      "step": 16714
+    },
+    {
+      "epoch": 33.43,
+      "grad_norm": 1.1532340049743652,
+      "learning_rate": 2e-05,
+      "loss": 0.03369208,
+      "step": 16715
+    },
+    {
+      "epoch": 33.432,
+      "grad_norm": 2.2827866077423096,
+      "learning_rate": 2e-05,
+      "loss": 0.04446258,
+      "step": 16716
+    },
+    {
+      "epoch": 33.434,
+      "grad_norm": 1.1983774900436401,
+      "learning_rate": 2e-05,
+      "loss": 0.04126913,
+      "step": 16717
+    },
+    {
+      "epoch": 33.436,
+      "grad_norm": 1.3406853675842285,
+      "learning_rate": 2e-05,
+      "loss": 0.03477804,
+      "step": 16718
+    },
+    {
+      "epoch": 33.438,
+      "grad_norm": 1.4163920879364014,
+      "learning_rate": 2e-05,
+      "loss": 0.05400594,
+      "step": 16719
+    },
+    {
+      "epoch": 33.44,
+      "grad_norm": 1.202052116394043,
+      "learning_rate": 2e-05,
+      "loss": 0.04666064,
+      "step": 16720
+    },
+    {
+      "epoch": 33.442,
+      "grad_norm": 1.020628809928894,
+      "learning_rate": 2e-05,
+      "loss": 0.03358646,
+      "step": 16721
+    },
+    {
+      "epoch": 33.444,
+      "grad_norm": 1.0128796100616455,
+      "learning_rate": 2e-05,
+      "loss": 0.03198342,
+      "step": 16722
+    },
+    {
+      "epoch": 33.446,
+      "grad_norm": 1.3667223453521729,
+      "learning_rate": 2e-05,
+      "loss": 0.04862178,
+      "step": 16723
+    },
+    {
+      "epoch": 33.448,
+      "grad_norm": 1.0447344779968262,
+      "learning_rate": 2e-05,
+      "loss": 0.0359532,
+      "step": 16724
+    },
+    {
+      "epoch": 33.45,
+      "grad_norm": 1.2238010168075562,
+      "learning_rate": 2e-05,
+      "loss": 0.03374971,
+      "step": 16725
+    },
+    {
+      "epoch": 33.452,
+      "grad_norm": 1.7347222566604614,
+      "learning_rate": 2e-05,
+      "loss": 0.05317023,
+      "step": 16726
+    },
+    {
+      "epoch": 33.454,
+      "grad_norm": 1.2322536706924438,
+      "learning_rate": 2e-05,
+      "loss": 0.05196625,
+      "step": 16727
+    },
+    {
+      "epoch": 33.456,
+      "grad_norm": 1.0215041637420654,
+      "learning_rate": 2e-05,
+      "loss": 0.03156345,
+      "step": 16728
+    },
+    {
+      "epoch": 33.458,
+      "grad_norm": 1.5477778911590576,
+      "learning_rate": 2e-05,
+      "loss": 0.04214332,
+      "step": 16729
+    },
+    {
+      "epoch": 33.46,
+      "grad_norm": 1.1270982027053833,
+      "learning_rate": 2e-05,
+      "loss": 0.04706374,
+      "step": 16730
+    },
+    {
+      "epoch": 33.462,
+      "grad_norm": 1.2014801502227783,
+      "learning_rate": 2e-05,
+      "loss": 0.04483427,
+      "step": 16731
+    },
+    {
+      "epoch": 33.464,
+      "grad_norm": 1.4638102054595947,
+      "learning_rate": 2e-05,
+      "loss": 0.05931569,
+      "step": 16732
+    },
+    {
+      "epoch": 33.466,
+      "grad_norm": 1.084399700164795,
+      "learning_rate": 2e-05,
+      "loss": 0.04061939,
+      "step": 16733
+    },
+    {
+      "epoch": 33.468,
+      "grad_norm": 1.254996657371521,
+      "learning_rate": 2e-05,
+      "loss": 0.03688417,
+      "step": 16734
+    },
+    {
+      "epoch": 33.47,
+      "grad_norm": 1.2726435661315918,
+      "learning_rate": 2e-05,
+      "loss": 0.04665891,
+      "step": 16735
+    },
+    {
+      "epoch": 33.472,
+      "grad_norm": 1.0594907999038696,
+      "learning_rate": 2e-05,
+      "loss": 0.03953811,
+      "step": 16736
+    },
+    {
+      "epoch": 33.474,
+      "grad_norm": 1.8973904848098755,
+      "learning_rate": 2e-05,
+      "loss": 0.04582148,
+      "step": 16737
+    },
+    {
+      "epoch": 33.476,
+      "grad_norm": 1.1382102966308594,
+      "learning_rate": 2e-05,
+      "loss": 0.03187426,
+      "step": 16738
+    },
+    {
+      "epoch": 33.478,
+      "grad_norm": 1.8633939027786255,
+      "learning_rate": 2e-05,
+      "loss": 0.04606871,
+      "step": 16739
+    },
+    {
+      "epoch": 33.48,
+      "grad_norm": 1.2930206060409546,
+      "learning_rate": 2e-05,
+      "loss": 0.03444067,
+      "step": 16740
+    },
+    {
+      "epoch": 33.482,
+      "grad_norm": 1.696944236755371,
+      "learning_rate": 2e-05,
+      "loss": 0.04227548,
+      "step": 16741
+    },
+    {
+      "epoch": 33.484,
+      "grad_norm": 1.115965723991394,
+      "learning_rate": 2e-05,
+      "loss": 0.03508355,
+      "step": 16742
+    },
+    {
+      "epoch": 33.486,
+      "grad_norm": 1.2587809562683105,
+      "learning_rate": 2e-05,
+      "loss": 0.03677014,
+      "step": 16743
+    },
+    {
+      "epoch": 33.488,
+      "grad_norm": 1.0906723737716675,
+      "learning_rate": 2e-05,
+      "loss": 0.03932015,
+      "step": 16744
+    },
+    {
+      "epoch": 33.49,
+      "grad_norm": 1.193812370300293,
+      "learning_rate": 2e-05,
+      "loss": 0.04173867,
+      "step": 16745
+    },
+    {
+      "epoch": 33.492,
+      "grad_norm": 1.3376166820526123,
+      "learning_rate": 2e-05,
+      "loss": 0.03034848,
+      "step": 16746
+    },
+    {
+      "epoch": 33.494,
+      "grad_norm": 1.1713814735412598,
+      "learning_rate": 2e-05,
+      "loss": 0.04791263,
+      "step": 16747
+    },
+    {
+      "epoch": 33.496,
+      "grad_norm": 1.723981261253357,
+      "learning_rate": 2e-05,
+      "loss": 0.04253115,
+      "step": 16748
+    },
+    {
+      "epoch": 33.498,
+      "grad_norm": 1.1139684915542603,
+      "learning_rate": 2e-05,
+      "loss": 0.04166551,
+      "step": 16749
+    },
+    {
+      "epoch": 33.5,
+      "grad_norm": 1.292219877243042,
+      "learning_rate": 2e-05,
+      "loss": 0.03971953,
+      "step": 16750
+    },
+    {
+      "epoch": 33.502,
+      "grad_norm": 1.2598776817321777,
+      "learning_rate": 2e-05,
+      "loss": 0.04661015,
+      "step": 16751
+    },
+    {
+      "epoch": 33.504,
+      "grad_norm": 1.248761534690857,
+      "learning_rate": 2e-05,
+      "loss": 0.04709421,
+      "step": 16752
+    },
+    {
+      "epoch": 33.506,
+      "grad_norm": 1.290848970413208,
+      "learning_rate": 2e-05,
+      "loss": 0.04546386,
+      "step": 16753
+    },
+    {
+      "epoch": 33.508,
+      "grad_norm": 3.039777994155884,
+      "learning_rate": 2e-05,
+      "loss": 0.04197717,
+      "step": 16754
+    },
+    {
+      "epoch": 33.51,
+      "grad_norm": 1.0628236532211304,
+      "learning_rate": 2e-05,
+      "loss": 0.03273289,
+      "step": 16755
+    },
+    {
+      "epoch": 33.512,
+      "grad_norm": 1.1584326028823853,
+      "learning_rate": 2e-05,
+      "loss": 0.03990288,
+      "step": 16756
+    },
+    {
+      "epoch": 33.514,
+      "grad_norm": 1.013412356376648,
+      "learning_rate": 2e-05,
+      "loss": 0.02777958,
+      "step": 16757
+    },
+    {
+      "epoch": 33.516,
+      "grad_norm": 2.4592092037200928,
+      "learning_rate": 2e-05,
+      "loss": 0.03798764,
+      "step": 16758
+    },
+    {
+      "epoch": 33.518,
+      "grad_norm": 1.4188690185546875,
+      "learning_rate": 2e-05,
+      "loss": 0.04493807,
+      "step": 16759
+    },
+    {
+      "epoch": 33.52,
+      "grad_norm": 4.044872760772705,
+      "learning_rate": 2e-05,
+      "loss": 0.05675915,
+      "step": 16760
+    },
+    {
+      "epoch": 33.522,
+      "grad_norm": 1.4888228178024292,
+      "learning_rate": 2e-05,
+      "loss": 0.039264,
+      "step": 16761
+    },
+    {
+      "epoch": 33.524,
+      "grad_norm": 1.2201298475265503,
+      "learning_rate": 2e-05,
+      "loss": 0.04121239,
+      "step": 16762
+    },
+    {
+      "epoch": 33.526,
+      "grad_norm": 1.0518842935562134,
+      "learning_rate": 2e-05,
+      "loss": 0.03466121,
+      "step": 16763
+    },
+    {
+      "epoch": 33.528,
+      "grad_norm": 0.8494116067886353,
+      "learning_rate": 2e-05,
+      "loss": 0.02845479,
+      "step": 16764
+    },
+    {
+      "epoch": 33.53,
+      "grad_norm": 1.2486729621887207,
+      "learning_rate": 2e-05,
+      "loss": 0.0528804,
+      "step": 16765
+    },
+    {
+      "epoch": 33.532,
+      "grad_norm": 1.1569898128509521,
+      "learning_rate": 2e-05,
+      "loss": 0.03768142,
+      "step": 16766
+    },
+    {
+      "epoch": 33.534,
+      "grad_norm": 0.9712448120117188,
+      "learning_rate": 2e-05,
+      "loss": 0.03684617,
+      "step": 16767
+    },
+    {
+      "epoch": 33.536,
+      "grad_norm": 3.453179121017456,
+      "learning_rate": 2e-05,
+      "loss": 0.05391577,
+      "step": 16768
+    },
+    {
+      "epoch": 33.538,
+      "grad_norm": 1.2368735074996948,
+      "learning_rate": 2e-05,
+      "loss": 0.04700865,
+      "step": 16769
+    },
+    {
+      "epoch": 33.54,
+      "grad_norm": 1.745194435119629,
+      "learning_rate": 2e-05,
+      "loss": 0.04905911,
+      "step": 16770
+    },
+    {
+      "epoch": 33.542,
+      "grad_norm": 1.4005979299545288,
+      "learning_rate": 2e-05,
+      "loss": 0.05808295,
+      "step": 16771
+    },
+    {
+      "epoch": 33.544,
+      "grad_norm": 1.6781764030456543,
+      "learning_rate": 2e-05,
+      "loss": 0.03419307,
+      "step": 16772
+    },
+    {
+      "epoch": 33.546,
+      "grad_norm": 1.5262959003448486,
+      "learning_rate": 2e-05,
+      "loss": 0.04930171,
+      "step": 16773
+    },
+    {
+      "epoch": 33.548,
+      "grad_norm": 1.185598373413086,
+      "learning_rate": 2e-05,
+      "loss": 0.04440915,
+      "step": 16774
+    },
+    {
+      "epoch": 33.55,
+      "grad_norm": 1.0013930797576904,
+      "learning_rate": 2e-05,
+      "loss": 0.03913631,
+      "step": 16775
+    },
+    {
+      "epoch": 33.552,
+      "grad_norm": 1.5372309684753418,
+      "learning_rate": 2e-05,
+      "loss": 0.04706179,
+      "step": 16776
+    },
+    {
+      "epoch": 33.554,
+      "grad_norm": 1.3070640563964844,
+      "learning_rate": 2e-05,
+      "loss": 0.03147201,
+      "step": 16777
+    },
+    {
+      "epoch": 33.556,
+      "grad_norm": 1.387190818786621,
+      "learning_rate": 2e-05,
+      "loss": 0.04918057,
+      "step": 16778
+    },
+    {
+      "epoch": 33.558,
+      "grad_norm": 2.1223700046539307,
+      "learning_rate": 2e-05,
+      "loss": 0.05167126,
+      "step": 16779
+    },
+    {
+      "epoch": 33.56,
+      "grad_norm": 1.3010870218276978,
+      "learning_rate": 2e-05,
+      "loss": 0.04664216,
+      "step": 16780
+    },
+    {
+      "epoch": 33.562,
+      "grad_norm": 1.6655634641647339,
+      "learning_rate": 2e-05,
+      "loss": 0.04345709,
+      "step": 16781
+    },
+    {
+      "epoch": 33.564,
+      "grad_norm": 2.2487990856170654,
+      "learning_rate": 2e-05,
+      "loss": 0.04758396,
+      "step": 16782
+    },
+    {
+      "epoch": 33.566,
+      "grad_norm": 2.5132153034210205,
+      "learning_rate": 2e-05,
+      "loss": 0.05307368,
+      "step": 16783
+    },
+    {
+      "epoch": 33.568,
+      "grad_norm": 2.023242235183716,
+      "learning_rate": 2e-05,
+      "loss": 0.03659308,
+      "step": 16784
+    },
+    {
+      "epoch": 33.57,
+      "grad_norm": 1.0173568725585938,
+      "learning_rate": 2e-05,
+      "loss": 0.03574786,
+      "step": 16785
+    },
+    {
+      "epoch": 33.572,
+      "grad_norm": 1.4924144744873047,
+      "learning_rate": 2e-05,
+      "loss": 0.04335709,
+      "step": 16786
+    },
+    {
+      "epoch": 33.574,
+      "grad_norm": 1.2708486318588257,
+      "learning_rate": 2e-05,
+      "loss": 0.04690626,
+      "step": 16787
+    },
+    {
+      "epoch": 33.576,
+      "grad_norm": 1.9800394773483276,
+      "learning_rate": 2e-05,
+      "loss": 0.04107783,
+      "step": 16788
+    },
+    {
+      "epoch": 33.578,
+      "grad_norm": 2.4189324378967285,
+      "learning_rate": 2e-05,
+      "loss": 0.05079643,
+      "step": 16789
+    },
+    {
+      "epoch": 33.58,
+      "grad_norm": 1.4563528299331665,
+      "learning_rate": 2e-05,
+      "loss": 0.04888935,
+      "step": 16790
+    },
+    {
+      "epoch": 33.582,
+      "grad_norm": 2.460385799407959,
+      "learning_rate": 2e-05,
+      "loss": 0.04483432,
+      "step": 16791
+    },
+    {
+      "epoch": 33.584,
+      "grad_norm": 2.3535897731781006,
+      "learning_rate": 2e-05,
+      "loss": 0.0520683,
+      "step": 16792
+    },
+    {
+      "epoch": 33.586,
+      "grad_norm": 1.497182846069336,
+      "learning_rate": 2e-05,
+      "loss": 0.04292556,
+      "step": 16793
+    },
+    {
+      "epoch": 33.588,
+      "grad_norm": 2.476083993911743,
+      "learning_rate": 2e-05,
+      "loss": 0.04846964,
+      "step": 16794
+    },
+    {
+      "epoch": 33.59,
+      "grad_norm": 1.223767876625061,
+      "learning_rate": 2e-05,
+      "loss": 0.06145135,
+      "step": 16795
+    },
+    {
+      "epoch": 33.592,
+      "grad_norm": 2.5422379970550537,
+      "learning_rate": 2e-05,
+      "loss": 0.05470592,
+      "step": 16796
+    },
+    {
+      "epoch": 33.594,
+      "grad_norm": 1.644066333770752,
+      "learning_rate": 2e-05,
+      "loss": 0.03853694,
+      "step": 16797
+    },
+    {
+      "epoch": 33.596,
+      "grad_norm": 1.5902693271636963,
+      "learning_rate": 2e-05,
+      "loss": 0.04749844,
+      "step": 16798
+    },
+    {
+      "epoch": 33.598,
+      "grad_norm": 0.8459369540214539,
+      "learning_rate": 2e-05,
+      "loss": 0.02365252,
+      "step": 16799
+    },
+    {
+      "epoch": 33.6,
+      "grad_norm": 0.9091578722000122,
+      "learning_rate": 2e-05,
+      "loss": 0.03125982,
+      "step": 16800
+    },
+    {
+      "epoch": 33.602,
+      "grad_norm": 1.5347368717193604,
+      "learning_rate": 2e-05,
+      "loss": 0.04600755,
+      "step": 16801
+    },
+    {
+      "epoch": 33.604,
+      "grad_norm": 1.4432686567306519,
+      "learning_rate": 2e-05,
+      "loss": 0.04101708,
+      "step": 16802
+    },
+    {
+      "epoch": 33.606,
+      "grad_norm": 1.1293833255767822,
+      "learning_rate": 2e-05,
+      "loss": 0.04084689,
+      "step": 16803
+    },
+    {
+      "epoch": 33.608,
+      "grad_norm": 1.2917184829711914,
+      "learning_rate": 2e-05,
+      "loss": 0.04250922,
+      "step": 16804
+    },
+    {
+      "epoch": 33.61,
+      "grad_norm": 3.9780828952789307,
+      "learning_rate": 2e-05,
+      "loss": 0.03426572,
+      "step": 16805
+    },
+    {
+      "epoch": 33.612,
+      "grad_norm": 2.3805651664733887,
+      "learning_rate": 2e-05,
+      "loss": 0.03994187,
+      "step": 16806
+    },
+    {
+      "epoch": 33.614,
+      "grad_norm": 1.2909883260726929,
+      "learning_rate": 2e-05,
+      "loss": 0.0498014,
+      "step": 16807
+    },
+    {
+      "epoch": 33.616,
+      "grad_norm": 1.0932729244232178,
+      "learning_rate": 2e-05,
+      "loss": 0.03009418,
+      "step": 16808
+    },
+    {
+      "epoch": 33.618,
+      "grad_norm": 1.3504739999771118,
+      "learning_rate": 2e-05,
+      "loss": 0.04348255,
+      "step": 16809
+    },
+    {
+      "epoch": 33.62,
+      "grad_norm": 1.2880327701568604,
+      "learning_rate": 2e-05,
+      "loss": 0.05897732,
+      "step": 16810
+    },
+    {
+      "epoch": 33.622,
+      "grad_norm": 1.2188899517059326,
+      "learning_rate": 2e-05,
+      "loss": 0.04643267,
+      "step": 16811
+    },
+    {
+      "epoch": 33.624,
+      "grad_norm": 1.2222504615783691,
+      "learning_rate": 2e-05,
+      "loss": 0.03556299,
+      "step": 16812
+    },
+    {
+      "epoch": 33.626,
+      "grad_norm": 1.506365418434143,
+      "learning_rate": 2e-05,
+      "loss": 0.03758072,
+      "step": 16813
+    },
+    {
+      "epoch": 33.628,
+      "grad_norm": 1.1334285736083984,
+      "learning_rate": 2e-05,
+      "loss": 0.03652769,
+      "step": 16814
+    },
+    {
+      "epoch": 33.63,
+      "grad_norm": 1.8486438989639282,
+      "learning_rate": 2e-05,
+      "loss": 0.03768915,
+      "step": 16815
+    },
+    {
+      "epoch": 33.632,
+      "grad_norm": 2.419917345046997,
+      "learning_rate": 2e-05,
+      "loss": 0.05534883,
+      "step": 16816
+    },
+    {
+      "epoch": 33.634,
+      "grad_norm": 1.1120336055755615,
+      "learning_rate": 2e-05,
+      "loss": 0.02930129,
+      "step": 16817
+    },
+    {
+      "epoch": 33.636,
+      "grad_norm": 0.9981685280799866,
+      "learning_rate": 2e-05,
+      "loss": 0.03070155,
+      "step": 16818
+    },
+    {
+      "epoch": 33.638,
+      "grad_norm": 1.1295273303985596,
+      "learning_rate": 2e-05,
+      "loss": 0.04213031,
+      "step": 16819
+    },
+    {
+      "epoch": 33.64,
+      "grad_norm": 3.216989278793335,
+      "learning_rate": 2e-05,
+      "loss": 0.03833716,
+      "step": 16820
+    },
+    {
+      "epoch": 33.642,
+      "grad_norm": 2.445704936981201,
+      "learning_rate": 2e-05,
+      "loss": 0.04210251,
+      "step": 16821
+    },
+    {
+      "epoch": 33.644,
+      "grad_norm": 1.0248154401779175,
+      "learning_rate": 2e-05,
+      "loss": 0.03077208,
+      "step": 16822
+    },
+    {
+      "epoch": 33.646,
+      "grad_norm": 1.1069806814193726,
+      "learning_rate": 2e-05,
+      "loss": 0.04635371,
+      "step": 16823
+    },
+    {
+      "epoch": 33.648,
+      "grad_norm": 2.114872455596924,
+      "learning_rate": 2e-05,
+      "loss": 0.06503963,
+      "step": 16824
+    },
+    {
+      "epoch": 33.65,
+      "grad_norm": 1.2636666297912598,
+      "learning_rate": 2e-05,
+      "loss": 0.02907618,
+      "step": 16825
+    },
+    {
+      "epoch": 33.652,
+      "grad_norm": 1.651583194732666,
+      "learning_rate": 2e-05,
+      "loss": 0.03353261,
+      "step": 16826
+    },
+    {
+      "epoch": 33.654,
+      "grad_norm": 1.2803378105163574,
+      "learning_rate": 2e-05,
+      "loss": 0.03870942,
+      "step": 16827
+    },
+    {
+      "epoch": 33.656,
+      "grad_norm": 1.2577418088912964,
+      "learning_rate": 2e-05,
+      "loss": 0.04237111,
+      "step": 16828
+    },
+    {
+      "epoch": 33.658,
+      "grad_norm": 1.1378142833709717,
+      "learning_rate": 2e-05,
+      "loss": 0.04286456,
+      "step": 16829
+    },
+    {
+      "epoch": 33.66,
+      "grad_norm": 2.4869840145111084,
+      "learning_rate": 2e-05,
+      "loss": 0.03604544,
+      "step": 16830
+    },
+    {
+      "epoch": 33.662,
+      "grad_norm": 1.4990627765655518,
+      "learning_rate": 2e-05,
+      "loss": 0.03898953,
+      "step": 16831
+    },
+    {
+      "epoch": 33.664,
+      "grad_norm": 1.4893847703933716,
+      "learning_rate": 2e-05,
+      "loss": 0.040364,
+      "step": 16832
+    },
+    {
+      "epoch": 33.666,
+      "grad_norm": 1.1118310689926147,
+      "learning_rate": 2e-05,
+      "loss": 0.03826702,
+      "step": 16833
+    },
+    {
+      "epoch": 33.668,
+      "grad_norm": 1.1041016578674316,
+      "learning_rate": 2e-05,
+      "loss": 0.04368906,
+      "step": 16834
+    },
+    {
+      "epoch": 33.67,
+      "grad_norm": 1.1376192569732666,
+      "learning_rate": 2e-05,
+      "loss": 0.04184592,
+      "step": 16835
+    },
+    {
+      "epoch": 33.672,
+      "grad_norm": 1.2078814506530762,
+      "learning_rate": 2e-05,
+      "loss": 0.04568657,
+      "step": 16836
+    },
+    {
+      "epoch": 33.674,
+      "grad_norm": 1.4587289094924927,
+      "learning_rate": 2e-05,
+      "loss": 0.04338841,
+      "step": 16837
+    },
+    {
+      "epoch": 33.676,
+      "grad_norm": 1.011198878288269,
+      "learning_rate": 2e-05,
+      "loss": 0.03405341,
+      "step": 16838
+    },
+    {
+      "epoch": 33.678,
+      "grad_norm": 1.286076307296753,
+      "learning_rate": 2e-05,
+      "loss": 0.04624006,
+      "step": 16839
+    },
+    {
+      "epoch": 33.68,
+      "grad_norm": 1.1408072710037231,
+      "learning_rate": 2e-05,
+      "loss": 0.04370711,
+      "step": 16840
+    },
+    {
+      "epoch": 33.682,
+      "grad_norm": 1.0561197996139526,
+      "learning_rate": 2e-05,
+      "loss": 0.04226065,
+      "step": 16841
+    },
+    {
+      "epoch": 33.684,
+      "grad_norm": 1.5302330255508423,
+      "learning_rate": 2e-05,
+      "loss": 0.05760441,
+      "step": 16842
+    },
+    {
+      "epoch": 33.686,
+      "grad_norm": 1.3950103521347046,
+      "learning_rate": 2e-05,
+      "loss": 0.06257196,
+      "step": 16843
+    },
+    {
+      "epoch": 33.688,
+      "grad_norm": 1.3161698579788208,
+      "learning_rate": 2e-05,
+      "loss": 0.04578729,
+      "step": 16844
+    },
+    {
+      "epoch": 33.69,
+      "grad_norm": 1.1795623302459717,
+      "learning_rate": 2e-05,
+      "loss": 0.02204716,
+      "step": 16845
+    },
+    {
+      "epoch": 33.692,
+      "grad_norm": 1.0264464616775513,
+      "learning_rate": 2e-05,
+      "loss": 0.04933001,
+      "step": 16846
+    },
+    {
+      "epoch": 33.694,
+      "grad_norm": 1.3931221961975098,
+      "learning_rate": 2e-05,
+      "loss": 0.04392222,
+      "step": 16847
+    },
+    {
+      "epoch": 33.696,
+      "grad_norm": 0.9317388534545898,
+      "learning_rate": 2e-05,
+      "loss": 0.02685932,
+      "step": 16848
+    },
+    {
+      "epoch": 33.698,
+      "grad_norm": 0.9781988263130188,
+      "learning_rate": 2e-05,
+      "loss": 0.03584233,
+      "step": 16849
+    },
+    {
+      "epoch": 33.7,
+      "grad_norm": 1.2689793109893799,
+      "learning_rate": 2e-05,
+      "loss": 0.04835085,
+      "step": 16850
+    },
+    {
+      "epoch": 33.702,
+      "grad_norm": 1.4062492847442627,
+      "learning_rate": 2e-05,
+      "loss": 0.04714158,
+      "step": 16851
+    },
+    {
+      "epoch": 33.704,
+      "grad_norm": 3.2041304111480713,
+      "learning_rate": 2e-05,
+      "loss": 0.05374725,
+      "step": 16852
+    },
+    {
+      "epoch": 33.706,
+      "grad_norm": 1.2796287536621094,
+      "learning_rate": 2e-05,
+      "loss": 0.0468824,
+      "step": 16853
+    },
+    {
+      "epoch": 33.708,
+      "grad_norm": 2.208021640777588,
+      "learning_rate": 2e-05,
+      "loss": 0.05651149,
+      "step": 16854
+    },
+    {
+      "epoch": 33.71,
+      "grad_norm": 1.8455325365066528,
+      "learning_rate": 2e-05,
+      "loss": 0.0509971,
+      "step": 16855
+    },
+    {
+      "epoch": 33.712,
+      "grad_norm": 0.9580972790718079,
+      "learning_rate": 2e-05,
+      "loss": 0.02455459,
+      "step": 16856
+    },
+    {
+      "epoch": 33.714,
+      "grad_norm": 1.2142010927200317,
+      "learning_rate": 2e-05,
+      "loss": 0.03981315,
+      "step": 16857
+    },
+    {
+      "epoch": 33.716,
+      "grad_norm": 1.0216953754425049,
+      "learning_rate": 2e-05,
+      "loss": 0.0417695,
+      "step": 16858
+    },
+    {
+      "epoch": 33.718,
+      "grad_norm": 1.1976450681686401,
+      "learning_rate": 2e-05,
+      "loss": 0.02827924,
+      "step": 16859
+    },
+    {
+      "epoch": 33.72,
+      "grad_norm": 1.1605479717254639,
+      "learning_rate": 2e-05,
+      "loss": 0.04454171,
+      "step": 16860
+    },
+    {
+      "epoch": 33.722,
+      "grad_norm": 1.382978081703186,
+      "learning_rate": 2e-05,
+      "loss": 0.06117807,
+      "step": 16861
+    },
+    {
+      "epoch": 33.724,
+      "grad_norm": 0.8501904010772705,
+      "learning_rate": 2e-05,
+      "loss": 0.0254951,
+      "step": 16862
+    },
+    {
+      "epoch": 33.726,
+      "grad_norm": 1.1386749744415283,
+      "learning_rate": 2e-05,
+      "loss": 0.04194954,
+      "step": 16863
+    },
+    {
+      "epoch": 33.728,
+      "grad_norm": 0.990307092666626,
+      "learning_rate": 2e-05,
+      "loss": 0.03089491,
+      "step": 16864
+    },
+    {
+      "epoch": 33.73,
+      "grad_norm": 0.9442136287689209,
+      "learning_rate": 2e-05,
+      "loss": 0.03526897,
+      "step": 16865
+    },
+    {
+      "epoch": 33.732,
+      "grad_norm": 1.119691014289856,
+      "learning_rate": 2e-05,
+      "loss": 0.03854529,
+      "step": 16866
+    },
+    {
+      "epoch": 33.734,
+      "grad_norm": 1.748223066329956,
+      "learning_rate": 2e-05,
+      "loss": 0.02689804,
+      "step": 16867
+    },
+    {
+      "epoch": 33.736,
+      "grad_norm": 1.6101315021514893,
+      "learning_rate": 2e-05,
+      "loss": 0.04781541,
+      "step": 16868
+    },
+    {
+      "epoch": 33.738,
+      "grad_norm": 1.0552995204925537,
+      "learning_rate": 2e-05,
+      "loss": 0.03037945,
+      "step": 16869
+    },
+    {
+      "epoch": 33.74,
+      "grad_norm": 1.2612851858139038,
+      "learning_rate": 2e-05,
+      "loss": 0.0377595,
+      "step": 16870
+    },
+    {
+      "epoch": 33.742,
+      "grad_norm": 0.9552090167999268,
+      "learning_rate": 2e-05,
+      "loss": 0.02962127,
+      "step": 16871
+    },
+    {
+      "epoch": 33.744,
+      "grad_norm": 0.9637928009033203,
+      "learning_rate": 2e-05,
+      "loss": 0.04046572,
+      "step": 16872
+    },
+    {
+      "epoch": 33.746,
+      "grad_norm": 1.0410312414169312,
+      "learning_rate": 2e-05,
+      "loss": 0.04488287,
+      "step": 16873
+    },
+    {
+      "epoch": 33.748,
+      "grad_norm": 1.2417147159576416,
+      "learning_rate": 2e-05,
+      "loss": 0.05209735,
+      "step": 16874
+    },
+    {
+      "epoch": 33.75,
+      "grad_norm": 1.3345032930374146,
+      "learning_rate": 2e-05,
+      "loss": 0.06243813,
+      "step": 16875
+    },
+    {
+      "epoch": 33.752,
+      "grad_norm": 0.8462825417518616,
+      "learning_rate": 2e-05,
+      "loss": 0.02463733,
+      "step": 16876
+    },
+    {
+      "epoch": 33.754,
+      "grad_norm": 1.9590753316879272,
+      "learning_rate": 2e-05,
+      "loss": 0.05277262,
+      "step": 16877
+    },
+    {
+      "epoch": 33.756,
+      "grad_norm": 1.141790747642517,
+      "learning_rate": 2e-05,
+      "loss": 0.04721228,
+      "step": 16878
+    },
+    {
+      "epoch": 33.758,
+      "grad_norm": 0.9426037669181824,
+      "learning_rate": 2e-05,
+      "loss": 0.03288198,
+      "step": 16879
+    },
+    {
+      "epoch": 33.76,
+      "grad_norm": 1.2627002000808716,
+      "learning_rate": 2e-05,
+      "loss": 0.05176467,
+      "step": 16880
+    },
+    {
+      "epoch": 33.762,
+      "grad_norm": 1.8264248371124268,
+      "learning_rate": 2e-05,
+      "loss": 0.04593317,
+      "step": 16881
+    },
+    {
+      "epoch": 33.764,
+      "grad_norm": 1.3055675029754639,
+      "learning_rate": 2e-05,
+      "loss": 0.06014679,
+      "step": 16882
+    },
+    {
+      "epoch": 33.766,
+      "grad_norm": 1.199167013168335,
+      "learning_rate": 2e-05,
+      "loss": 0.05326901,
+      "step": 16883
+    },
+    {
+      "epoch": 33.768,
+      "grad_norm": 1.399657964706421,
+      "learning_rate": 2e-05,
+      "loss": 0.03976363,
+      "step": 16884
+    },
+    {
+      "epoch": 33.77,
+      "grad_norm": 1.4370189905166626,
+      "learning_rate": 2e-05,
+      "loss": 0.04173758,
+      "step": 16885
+    },
+    {
+      "epoch": 33.772,
+      "grad_norm": 1.3217477798461914,
+      "learning_rate": 2e-05,
+      "loss": 0.04312097,
+      "step": 16886
+    },
+    {
+      "epoch": 33.774,
+      "grad_norm": 1.0405267477035522,
+      "learning_rate": 2e-05,
+      "loss": 0.02572107,
+      "step": 16887
+    },
+    {
+      "epoch": 33.776,
+      "grad_norm": 3.8962230682373047,
+      "learning_rate": 2e-05,
+      "loss": 0.05806974,
+      "step": 16888
+    },
+    {
+      "epoch": 33.778,
+      "grad_norm": 1.183304786682129,
+      "learning_rate": 2e-05,
+      "loss": 0.05114609,
+      "step": 16889
+    },
+    {
+      "epoch": 33.78,
+      "grad_norm": 1.0144455432891846,
+      "learning_rate": 2e-05,
+      "loss": 0.03402551,
+      "step": 16890
+    },
+    {
+      "epoch": 33.782,
+      "grad_norm": 0.912356436252594,
+      "learning_rate": 2e-05,
+      "loss": 0.03162487,
+      "step": 16891
+    },
+    {
+      "epoch": 33.784,
+      "grad_norm": 1.5084586143493652,
+      "learning_rate": 2e-05,
+      "loss": 0.04724879,
+      "step": 16892
+    },
+    {
+      "epoch": 33.786,
+      "grad_norm": 1.081552267074585,
+      "learning_rate": 2e-05,
+      "loss": 0.03258973,
+      "step": 16893
+    },
+    {
+      "epoch": 33.788,
+      "grad_norm": 1.0721511840820312,
+      "learning_rate": 2e-05,
+      "loss": 0.04917396,
+      "step": 16894
+    },
+    {
+      "epoch": 33.79,
+      "grad_norm": 1.1885251998901367,
+      "learning_rate": 2e-05,
+      "loss": 0.04815122,
+      "step": 16895
+    },
+    {
+      "epoch": 33.792,
+      "grad_norm": 1.4012700319290161,
+      "learning_rate": 2e-05,
+      "loss": 0.04503307,
+      "step": 16896
+    },
+    {
+      "epoch": 33.794,
+      "grad_norm": 1.018221378326416,
+      "learning_rate": 2e-05,
+      "loss": 0.03371017,
+      "step": 16897
+    },
+    {
+      "epoch": 33.796,
+      "grad_norm": 1.1117615699768066,
+      "learning_rate": 2e-05,
+      "loss": 0.04455762,
+      "step": 16898
+    },
+    {
+      "epoch": 33.798,
+      "grad_norm": 2.338944435119629,
+      "learning_rate": 2e-05,
+      "loss": 0.05771154,
+      "step": 16899
+    },
+    {
+      "epoch": 33.8,
+      "grad_norm": 1.3427327871322632,
+      "learning_rate": 2e-05,
+      "loss": 0.04563061,
+      "step": 16900
+    },
+    {
+      "epoch": 33.802,
+      "grad_norm": 1.598863124847412,
+      "learning_rate": 2e-05,
+      "loss": 0.04085599,
+      "step": 16901
+    },
+    {
+      "epoch": 33.804,
+      "grad_norm": 4.519804000854492,
+      "learning_rate": 2e-05,
+      "loss": 0.05506884,
+      "step": 16902
+    },
+    {
+      "epoch": 33.806,
+      "grad_norm": 1.7649790048599243,
+      "learning_rate": 2e-05,
+      "loss": 0.04375284,
+      "step": 16903
+    },
+    {
+      "epoch": 33.808,
+      "grad_norm": 0.7932525873184204,
+      "learning_rate": 2e-05,
+      "loss": 0.02062046,
+      "step": 16904
+    },
+    {
+      "epoch": 33.81,
+      "grad_norm": 1.4744629859924316,
+      "learning_rate": 2e-05,
+      "loss": 0.0412587,
+      "step": 16905
+    },
+    {
+      "epoch": 33.812,
+      "grad_norm": 1.8036439418792725,
+      "learning_rate": 2e-05,
+      "loss": 0.04557544,
+      "step": 16906
+    },
+    {
+      "epoch": 33.814,
+      "grad_norm": 1.2862471342086792,
+      "learning_rate": 2e-05,
+      "loss": 0.06118719,
+      "step": 16907
+    },
+    {
+      "epoch": 33.816,
+      "grad_norm": 1.0730700492858887,
+      "learning_rate": 2e-05,
+      "loss": 0.03417275,
+      "step": 16908
+    },
+    {
+      "epoch": 33.818,
+      "grad_norm": 1.6941379308700562,
+      "learning_rate": 2e-05,
+      "loss": 0.06889927,
+      "step": 16909
+    },
+    {
+      "epoch": 33.82,
+      "grad_norm": 1.0613124370574951,
+      "learning_rate": 2e-05,
+      "loss": 0.04204726,
+      "step": 16910
+    },
+    {
+      "epoch": 33.822,
+      "grad_norm": 1.436385154724121,
+      "learning_rate": 2e-05,
+      "loss": 0.05231769,
+      "step": 16911
+    },
+    {
+      "epoch": 33.824,
+      "grad_norm": 1.1055042743682861,
+      "learning_rate": 2e-05,
+      "loss": 0.02975886,
+      "step": 16912
+    },
+    {
+      "epoch": 33.826,
+      "grad_norm": 1.3664203882217407,
+      "learning_rate": 2e-05,
+      "loss": 0.0504609,
+      "step": 16913
+    },
+    {
+      "epoch": 33.828,
+      "grad_norm": 1.1539993286132812,
+      "learning_rate": 2e-05,
+      "loss": 0.04137579,
+      "step": 16914
+    },
+    {
+      "epoch": 33.83,
+      "grad_norm": 1.266783595085144,
+      "learning_rate": 2e-05,
+      "loss": 0.04526065,
+      "step": 16915
+    },
+    {
+      "epoch": 33.832,
+      "grad_norm": 1.3044017553329468,
+      "learning_rate": 2e-05,
+      "loss": 0.04141079,
+      "step": 16916
+    },
+    {
+      "epoch": 33.834,
+      "grad_norm": 2.1756279468536377,
+      "learning_rate": 2e-05,
+      "loss": 0.04577321,
+      "step": 16917
+    },
+    {
+      "epoch": 33.836,
+      "grad_norm": 1.8856961727142334,
+      "learning_rate": 2e-05,
+      "loss": 0.04963861,
+      "step": 16918
+    },
+    {
+      "epoch": 33.838,
+      "grad_norm": 1.0676918029785156,
+      "learning_rate": 2e-05,
+      "loss": 0.04658003,
+      "step": 16919
+    },
+    {
+      "epoch": 33.84,
+      "grad_norm": 2.096513509750366,
+      "learning_rate": 2e-05,
+      "loss": 0.04904509,
+      "step": 16920
+    },
+    {
+      "epoch": 33.842,
+      "grad_norm": 1.2753912210464478,
+      "learning_rate": 2e-05,
+      "loss": 0.05007047,
+      "step": 16921
+    },
+    {
+      "epoch": 33.844,
+      "grad_norm": 0.9849660396575928,
+      "learning_rate": 2e-05,
+      "loss": 0.03128029,
+      "step": 16922
+    },
+    {
+      "epoch": 33.846,
+      "grad_norm": 1.4278568029403687,
+      "learning_rate": 2e-05,
+      "loss": 0.04545929,
+      "step": 16923
+    },
+    {
+      "epoch": 33.848,
+      "grad_norm": 1.4661980867385864,
+      "learning_rate": 2e-05,
+      "loss": 0.05411399,
+      "step": 16924
+    },
+    {
+      "epoch": 33.85,
+      "grad_norm": 1.2466408014297485,
+      "learning_rate": 2e-05,
+      "loss": 0.04471597,
+      "step": 16925
+    },
+    {
+      "epoch": 33.852,
+      "grad_norm": 0.9380550980567932,
+      "learning_rate": 2e-05,
+      "loss": 0.04102387,
+      "step": 16926
+    },
+    {
+      "epoch": 33.854,
+      "grad_norm": 1.1571276187896729,
+      "learning_rate": 2e-05,
+      "loss": 0.05654067,
+      "step": 16927
+    },
+    {
+      "epoch": 33.856,
+      "grad_norm": 0.9969253540039062,
+      "learning_rate": 2e-05,
+      "loss": 0.03924453,
+      "step": 16928
+    },
+    {
+      "epoch": 33.858,
+      "grad_norm": 1.004270076751709,
+      "learning_rate": 2e-05,
+      "loss": 0.0299517,
+      "step": 16929
+    },
+    {
+      "epoch": 33.86,
+      "grad_norm": 1.096602439880371,
+      "learning_rate": 2e-05,
+      "loss": 0.03355722,
+      "step": 16930
+    },
+    {
+      "epoch": 33.862,
+      "grad_norm": 1.0416635274887085,
+      "learning_rate": 2e-05,
+      "loss": 0.03369167,
+      "step": 16931
+    },
+    {
+      "epoch": 33.864,
+      "grad_norm": 1.327217698097229,
+      "learning_rate": 2e-05,
+      "loss": 0.05232648,
+      "step": 16932
+    },
+    {
+      "epoch": 33.866,
+      "grad_norm": 1.9164153337478638,
+      "learning_rate": 2e-05,
+      "loss": 0.05434879,
+      "step": 16933
+    },
+    {
+      "epoch": 33.868,
+      "grad_norm": 1.9355528354644775,
+      "learning_rate": 2e-05,
+      "loss": 0.04286207,
+      "step": 16934
+    },
+    {
+      "epoch": 33.87,
+      "grad_norm": 2.0409860610961914,
+      "learning_rate": 2e-05,
+      "loss": 0.03842381,
+      "step": 16935
+    },
+    {
+      "epoch": 33.872,
+      "grad_norm": 1.4562387466430664,
+      "learning_rate": 2e-05,
+      "loss": 0.03790183,
+      "step": 16936
+    },
+    {
+      "epoch": 33.874,
+      "grad_norm": 1.2432488203048706,
+      "learning_rate": 2e-05,
+      "loss": 0.04522638,
+      "step": 16937
+    },
+    {
+      "epoch": 33.876,
+      "grad_norm": 1.5313806533813477,
+      "learning_rate": 2e-05,
+      "loss": 0.04211116,
+      "step": 16938
+    },
+    {
+      "epoch": 33.878,
+      "grad_norm": 1.330566644668579,
+      "learning_rate": 2e-05,
+      "loss": 0.04907233,
+      "step": 16939
+    },
+    {
+      "epoch": 33.88,
+      "grad_norm": 1.2614879608154297,
+      "learning_rate": 2e-05,
+      "loss": 0.06006509,
+      "step": 16940
+    },
+    {
+      "epoch": 33.882,
+      "grad_norm": 0.8992356061935425,
+      "learning_rate": 2e-05,
+      "loss": 0.03267233,
+      "step": 16941
+    },
+    {
+      "epoch": 33.884,
+      "grad_norm": 1.0046213865280151,
+      "learning_rate": 2e-05,
+      "loss": 0.04911587,
+      "step": 16942
+    },
+    {
+      "epoch": 33.886,
+      "grad_norm": 1.811383605003357,
+      "learning_rate": 2e-05,
+      "loss": 0.05342248,
+      "step": 16943
+    },
+    {
+      "epoch": 33.888,
+      "grad_norm": 1.098055362701416,
+      "learning_rate": 2e-05,
+      "loss": 0.04004695,
+      "step": 16944
+    },
+    {
+      "epoch": 33.89,
+      "grad_norm": 1.5141690969467163,
+      "learning_rate": 2e-05,
+      "loss": 0.05056385,
+      "step": 16945
+    },
+    {
+      "epoch": 33.892,
+      "grad_norm": 1.2271699905395508,
+      "learning_rate": 2e-05,
+      "loss": 0.04924796,
+      "step": 16946
+    },
+    {
+      "epoch": 33.894,
+      "grad_norm": 1.2581502199172974,
+      "learning_rate": 2e-05,
+      "loss": 0.03863451,
+      "step": 16947
+    },
+    {
+      "epoch": 33.896,
+      "grad_norm": 0.9692806005477905,
+      "learning_rate": 2e-05,
+      "loss": 0.04194497,
+      "step": 16948
+    },
+    {
+      "epoch": 33.898,
+      "grad_norm": 1.4562419652938843,
+      "learning_rate": 2e-05,
+      "loss": 0.04685233,
+      "step": 16949
+    },
+    {
+      "epoch": 33.9,
+      "grad_norm": 1.32866632938385,
+      "learning_rate": 2e-05,
+      "loss": 0.05198713,
+      "step": 16950
+    },
+    {
+      "epoch": 33.902,
+      "grad_norm": 1.1685373783111572,
+      "learning_rate": 2e-05,
+      "loss": 0.04833569,
+      "step": 16951
+    },
+    {
+      "epoch": 33.904,
+      "grad_norm": 1.1611336469650269,
+      "learning_rate": 2e-05,
+      "loss": 0.04151731,
+      "step": 16952
+    },
+    {
+      "epoch": 33.906,
+      "grad_norm": 1.1714519262313843,
+      "learning_rate": 2e-05,
+      "loss": 0.04272041,
+      "step": 16953
+    },
+    {
+      "epoch": 33.908,
+      "grad_norm": 1.6427088975906372,
+      "learning_rate": 2e-05,
+      "loss": 0.05739013,
+      "step": 16954
+    },
+    {
+      "epoch": 33.91,
+      "grad_norm": 1.11444890499115,
+      "learning_rate": 2e-05,
+      "loss": 0.03927352,
+      "step": 16955
+    },
+    {
+      "epoch": 33.912,
+      "grad_norm": 1.4082242250442505,
+      "learning_rate": 2e-05,
+      "loss": 0.06680731,
+      "step": 16956
+    },
+    {
+      "epoch": 33.914,
+      "grad_norm": 1.295926809310913,
+      "learning_rate": 2e-05,
+      "loss": 0.04506046,
+      "step": 16957
+    },
+    {
+      "epoch": 33.916,
+      "grad_norm": 0.9405683279037476,
+      "learning_rate": 2e-05,
+      "loss": 0.02393782,
+      "step": 16958
+    },
+    {
+      "epoch": 33.918,
+      "grad_norm": 1.58418607711792,
+      "learning_rate": 2e-05,
+      "loss": 0.04732141,
+      "step": 16959
+    },
+    {
+      "epoch": 33.92,
+      "grad_norm": 1.6658512353897095,
+      "learning_rate": 2e-05,
+      "loss": 0.05234197,
+      "step": 16960
+    },
+    {
+      "epoch": 33.922,
+      "grad_norm": 1.5868287086486816,
+      "learning_rate": 2e-05,
+      "loss": 0.05572698,
+      "step": 16961
+    },
+    {
+      "epoch": 33.924,
+      "grad_norm": 1.2670631408691406,
+      "learning_rate": 2e-05,
+      "loss": 0.03493927,
+      "step": 16962
+    },
+    {
+      "epoch": 33.926,
+      "grad_norm": 1.2648658752441406,
+      "learning_rate": 2e-05,
+      "loss": 0.04391746,
+      "step": 16963
+    },
+    {
+      "epoch": 33.928,
+      "grad_norm": 1.5117779970169067,
+      "learning_rate": 2e-05,
+      "loss": 0.04314294,
+      "step": 16964
+    },
+    {
+      "epoch": 33.93,
+      "grad_norm": 1.159271001815796,
+      "learning_rate": 2e-05,
+      "loss": 0.03362155,
+      "step": 16965
+    },
+    {
+      "epoch": 33.932,
+      "grad_norm": 1.2142740488052368,
+      "learning_rate": 2e-05,
+      "loss": 0.03751262,
+      "step": 16966
+    },
+    {
+      "epoch": 33.934,
+      "grad_norm": 1.2300341129302979,
+      "learning_rate": 2e-05,
+      "loss": 0.04216234,
+      "step": 16967
+    },
+    {
+      "epoch": 33.936,
+      "grad_norm": 1.0165789127349854,
+      "learning_rate": 2e-05,
+      "loss": 0.03385535,
+      "step": 16968
+    },
+    {
+      "epoch": 33.938,
+      "grad_norm": 1.1254189014434814,
+      "learning_rate": 2e-05,
+      "loss": 0.03804931,
+      "step": 16969
+    },
+    {
+      "epoch": 33.94,
+      "grad_norm": 1.0231702327728271,
+      "learning_rate": 2e-05,
+      "loss": 0.03211974,
+      "step": 16970
+    },
+    {
+      "epoch": 33.942,
+      "grad_norm": 1.1226533651351929,
+      "learning_rate": 2e-05,
+      "loss": 0.04243637,
+      "step": 16971
+    },
+    {
+      "epoch": 33.944,
+      "grad_norm": 1.0575497150421143,
+      "learning_rate": 2e-05,
+      "loss": 0.03891562,
+      "step": 16972
+    },
+    {
+      "epoch": 33.946,
+      "grad_norm": 1.3300598859786987,
+      "learning_rate": 2e-05,
+      "loss": 0.03813462,
+      "step": 16973
+    },
+    {
+      "epoch": 33.948,
+      "grad_norm": 1.1231666803359985,
+      "learning_rate": 2e-05,
+      "loss": 0.04401132,
+      "step": 16974
+    },
+    {
+      "epoch": 33.95,
+      "grad_norm": 1.0384533405303955,
+      "learning_rate": 2e-05,
+      "loss": 0.04073372,
+      "step": 16975
+    },
+    {
+      "epoch": 33.952,
+      "grad_norm": 0.9705667495727539,
+      "learning_rate": 2e-05,
+      "loss": 0.0333796,
+      "step": 16976
+    },
+    {
+      "epoch": 33.954,
+      "grad_norm": 3.3953044414520264,
+      "learning_rate": 2e-05,
+      "loss": 0.04748266,
+      "step": 16977
+    },
+    {
+      "epoch": 33.956,
+      "grad_norm": 1.0151921510696411,
+      "learning_rate": 2e-05,
+      "loss": 0.03253239,
+      "step": 16978
+    },
+    {
+      "epoch": 33.958,
+      "grad_norm": 1.612838864326477,
+      "learning_rate": 2e-05,
+      "loss": 0.04560025,
+      "step": 16979
+    },
+    {
+      "epoch": 33.96,
+      "grad_norm": 1.2825559377670288,
+      "learning_rate": 2e-05,
+      "loss": 0.03292632,
+      "step": 16980
+    },
+    {
+      "epoch": 33.962,
+      "grad_norm": 1.1484404802322388,
+      "learning_rate": 2e-05,
+      "loss": 0.03267542,
+      "step": 16981
+    },
+    {
+      "epoch": 33.964,
+      "grad_norm": 1.3197839260101318,
+      "learning_rate": 2e-05,
+      "loss": 0.0385856,
+      "step": 16982
+    },
+    {
+      "epoch": 33.966,
+      "grad_norm": 1.66212797164917,
+      "learning_rate": 2e-05,
+      "loss": 0.03775371,
+      "step": 16983
+    },
+    {
+      "epoch": 33.968,
+      "grad_norm": 2.77333664894104,
+      "learning_rate": 2e-05,
+      "loss": 0.04835136,
+      "step": 16984
+    },
+    {
+      "epoch": 33.97,
+      "grad_norm": 1.2204259634017944,
+      "learning_rate": 2e-05,
+      "loss": 0.03903702,
+      "step": 16985
+    },
+    {
+      "epoch": 33.972,
+      "grad_norm": 1.9606164693832397,
+      "learning_rate": 2e-05,
+      "loss": 0.03565089,
+      "step": 16986
+    },
+    {
+      "epoch": 33.974,
+      "grad_norm": 1.029175043106079,
+      "learning_rate": 2e-05,
+      "loss": 0.03104983,
+      "step": 16987
+    },
+    {
+      "epoch": 33.976,
+      "grad_norm": 1.9764819145202637,
+      "learning_rate": 2e-05,
+      "loss": 0.04124163,
+      "step": 16988
+    },
+    {
+      "epoch": 33.978,
+      "grad_norm": 1.2919042110443115,
+      "learning_rate": 2e-05,
+      "loss": 0.04482358,
+      "step": 16989
+    },
+    {
+      "epoch": 33.98,
+      "grad_norm": 1.8748044967651367,
+      "learning_rate": 2e-05,
+      "loss": 0.05687953,
+      "step": 16990
+    },
+    {
+      "epoch": 33.982,
+      "grad_norm": 1.5990813970565796,
+      "learning_rate": 2e-05,
+      "loss": 0.06885971,
+      "step": 16991
+    },
+    {
+      "epoch": 33.984,
+      "grad_norm": 1.6569875478744507,
+      "learning_rate": 2e-05,
+      "loss": 0.02224008,
+      "step": 16992
+    },
+    {
+      "epoch": 33.986,
+      "grad_norm": 1.5059123039245605,
+      "learning_rate": 2e-05,
+      "loss": 0.03797079,
+      "step": 16993
+    },
+    {
+      "epoch": 33.988,
+      "grad_norm": 1.1320881843566895,
+      "learning_rate": 2e-05,
+      "loss": 0.04442799,
+      "step": 16994
+    },
+    {
+      "epoch": 33.99,
+      "grad_norm": 1.8858788013458252,
+      "learning_rate": 2e-05,
+      "loss": 0.05333455,
+      "step": 16995
+    },
+    {
+      "epoch": 33.992,
+      "grad_norm": 1.0770121812820435,
+      "learning_rate": 2e-05,
+      "loss": 0.04818095,
+      "step": 16996
+    },
+    {
+      "epoch": 33.994,
+      "grad_norm": 2.505625009536743,
+      "learning_rate": 2e-05,
+      "loss": 0.04461709,
+      "step": 16997
+    },
+    {
+      "epoch": 33.996,
+      "grad_norm": 1.5307080745697021,
+      "learning_rate": 2e-05,
+      "loss": 0.04006788,
+      "step": 16998
+    },
+    {
+      "epoch": 33.998,
+      "grad_norm": 1.1944080591201782,
+      "learning_rate": 2e-05,
+      "loss": 0.05010471,
+      "step": 16999
+    },
+    {
+      "epoch": 34.0,
+      "grad_norm": 1.0636547803878784,
+      "learning_rate": 2e-05,
+      "loss": 0.04512432,
+      "step": 17000
+    },
+    {
+      "epoch": 34.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9820359281437125,
+        "Equal_1": 1.0,
+        "Equal_2": 0.9860279441117764,
+        "Equal_3": 0.9860279441117764,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.998003992015968,
+        "Parallel_1": 0.9919839679358717,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.992,
+        "Perpendicular_1": 0.996,
+        "Perpendicular_2": 0.98,
+        "Perpendicular_3": 0.905811623246493,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 0.996,
+        "PointLiesOnCircle_3": 0.994,
+        "PointLiesOnLine_1": 0.9959919839679359,
+        "PointLiesOnLine_2": 0.9939879759519038,
+        "PointLiesOnLine_3": 0.9880239520958084
+      },
+      "eval_runtime": 319.634,
+      "eval_samples_per_second": 32.85,
+      "eval_steps_per_second": 0.657,
+      "step": 17000
+    },
+    {
+      "epoch": 34.002,
+      "grad_norm": 1.0812188386917114,
+      "learning_rate": 2e-05,
+      "loss": 0.03332932,
+      "step": 17001
+    },
+    {
+      "epoch": 34.004,
+      "grad_norm": 1.6815991401672363,
+      "learning_rate": 2e-05,
+      "loss": 0.04507479,
+      "step": 17002
+    },
+    {
+      "epoch": 34.006,
+      "grad_norm": 1.3439853191375732,
+      "learning_rate": 2e-05,
+      "loss": 0.04140931,
+      "step": 17003
+    },
+    {
+      "epoch": 34.008,
+      "grad_norm": 1.2833200693130493,
+      "learning_rate": 2e-05,
+      "loss": 0.03846888,
+      "step": 17004
+    },
+    {
+      "epoch": 34.01,
+      "grad_norm": 1.3974056243896484,
+      "learning_rate": 2e-05,
+      "loss": 0.0499987,
+      "step": 17005
+    },
+    {
+      "epoch": 34.012,
+      "grad_norm": 2.1579604148864746,
+      "learning_rate": 2e-05,
+      "loss": 0.04521484,
+      "step": 17006
+    },
+    {
+      "epoch": 34.014,
+      "grad_norm": 1.3013967275619507,
+      "learning_rate": 2e-05,
+      "loss": 0.04818133,
+      "step": 17007
+    },
+    {
+      "epoch": 34.016,
+      "grad_norm": 1.740728497505188,
+      "learning_rate": 2e-05,
+      "loss": 0.04121596,
+      "step": 17008
+    },
+    {
+      "epoch": 34.018,
+      "grad_norm": 0.9428570866584778,
+      "learning_rate": 2e-05,
+      "loss": 0.03057436,
+      "step": 17009
+    },
+    {
+      "epoch": 34.02,
+      "grad_norm": 1.1139823198318481,
+      "learning_rate": 2e-05,
+      "loss": 0.04498542,
+      "step": 17010
+    },
+    {
+      "epoch": 34.022,
+      "grad_norm": 1.3437737226486206,
+      "learning_rate": 2e-05,
+      "loss": 0.04737519,
+      "step": 17011
+    },
+    {
+      "epoch": 34.024,
+      "grad_norm": 1.2312670946121216,
+      "learning_rate": 2e-05,
+      "loss": 0.04723905,
+      "step": 17012
+    },
+    {
+      "epoch": 34.026,
+      "grad_norm": 2.406672954559326,
+      "learning_rate": 2e-05,
+      "loss": 0.04161016,
+      "step": 17013
+    },
+    {
+      "epoch": 34.028,
+      "grad_norm": 1.13112211227417,
+      "learning_rate": 2e-05,
+      "loss": 0.05677346,
+      "step": 17014
+    },
+    {
+      "epoch": 34.03,
+      "grad_norm": 1.2599197626113892,
+      "learning_rate": 2e-05,
+      "loss": 0.05227666,
+      "step": 17015
+    },
+    {
+      "epoch": 34.032,
+      "grad_norm": 1.2989979982376099,
+      "learning_rate": 2e-05,
+      "loss": 0.0471041,
+      "step": 17016
+    },
+    {
+      "epoch": 34.034,
+      "grad_norm": 1.432118535041809,
+      "learning_rate": 2e-05,
+      "loss": 0.05415585,
+      "step": 17017
+    },
+    {
+      "epoch": 34.036,
+      "grad_norm": 1.2104634046554565,
+      "learning_rate": 2e-05,
+      "loss": 0.04434905,
+      "step": 17018
+    },
+    {
+      "epoch": 34.038,
+      "grad_norm": 1.2272220849990845,
+      "learning_rate": 2e-05,
+      "loss": 0.04309473,
+      "step": 17019
+    },
+    {
+      "epoch": 34.04,
+      "grad_norm": 1.2818056344985962,
+      "learning_rate": 2e-05,
+      "loss": 0.04750851,
+      "step": 17020
+    },
+    {
+      "epoch": 34.042,
+      "grad_norm": 1.795697808265686,
+      "learning_rate": 2e-05,
+      "loss": 0.04016794,
+      "step": 17021
+    },
+    {
+      "epoch": 34.044,
+      "grad_norm": 1.261646032333374,
+      "learning_rate": 2e-05,
+      "loss": 0.05218561,
+      "step": 17022
+    },
+    {
+      "epoch": 34.046,
+      "grad_norm": 1.2014031410217285,
+      "learning_rate": 2e-05,
+      "loss": 0.03123101,
+      "step": 17023
+    },
+    {
+      "epoch": 34.048,
+      "grad_norm": 1.3816989660263062,
+      "learning_rate": 2e-05,
+      "loss": 0.05157385,
+      "step": 17024
+    },
+    {
+      "epoch": 34.05,
+      "grad_norm": 0.9320183992385864,
+      "learning_rate": 2e-05,
+      "loss": 0.02841931,
+      "step": 17025
+    },
+    {
+      "epoch": 34.052,
+      "grad_norm": 1.5845396518707275,
+      "learning_rate": 2e-05,
+      "loss": 0.0396069,
+      "step": 17026
+    },
+    {
+      "epoch": 34.054,
+      "grad_norm": 1.7424157857894897,
+      "learning_rate": 2e-05,
+      "loss": 0.05160773,
+      "step": 17027
+    },
+    {
+      "epoch": 34.056,
+      "grad_norm": 1.3250648975372314,
+      "learning_rate": 2e-05,
+      "loss": 0.03230084,
+      "step": 17028
+    },
+    {
+      "epoch": 34.058,
+      "grad_norm": 1.3948570489883423,
+      "learning_rate": 2e-05,
+      "loss": 0.04338755,
+      "step": 17029
+    },
+    {
+      "epoch": 34.06,
+      "grad_norm": 1.9198899269104004,
+      "learning_rate": 2e-05,
+      "loss": 0.03582556,
+      "step": 17030
+    },
+    {
+      "epoch": 34.062,
+      "grad_norm": 1.0664349794387817,
+      "learning_rate": 2e-05,
+      "loss": 0.02794036,
+      "step": 17031
+    },
+    {
+      "epoch": 34.064,
+      "grad_norm": 1.0144078731536865,
+      "learning_rate": 2e-05,
+      "loss": 0.03366199,
+      "step": 17032
+    },
+    {
+      "epoch": 34.066,
+      "grad_norm": 1.4609384536743164,
+      "learning_rate": 2e-05,
+      "loss": 0.0822296,
+      "step": 17033
+    },
+    {
+      "epoch": 34.068,
+      "grad_norm": 1.6045973300933838,
+      "learning_rate": 2e-05,
+      "loss": 0.04467433,
+      "step": 17034
+    },
+    {
+      "epoch": 34.07,
+      "grad_norm": 2.1055264472961426,
+      "learning_rate": 2e-05,
+      "loss": 0.04265523,
+      "step": 17035
+    },
+    {
+      "epoch": 34.072,
+      "grad_norm": 1.1740909814834595,
+      "learning_rate": 2e-05,
+      "loss": 0.04725765,
+      "step": 17036
+    },
+    {
+      "epoch": 34.074,
+      "grad_norm": 1.264272928237915,
+      "learning_rate": 2e-05,
+      "loss": 0.04407164,
+      "step": 17037
+    },
+    {
+      "epoch": 34.076,
+      "grad_norm": 1.4800697565078735,
+      "learning_rate": 2e-05,
+      "loss": 0.05361645,
+      "step": 17038
+    },
+    {
+      "epoch": 34.078,
+      "grad_norm": 0.9612439870834351,
+      "learning_rate": 2e-05,
+      "loss": 0.02787845,
+      "step": 17039
+    },
+    {
+      "epoch": 34.08,
+      "grad_norm": 1.2758971452713013,
+      "learning_rate": 2e-05,
+      "loss": 0.03535602,
+      "step": 17040
+    },
+    {
+      "epoch": 34.082,
+      "grad_norm": 1.4546493291854858,
+      "learning_rate": 2e-05,
+      "loss": 0.03896,
+      "step": 17041
+    },
+    {
+      "epoch": 34.084,
+      "grad_norm": 1.3334996700286865,
+      "learning_rate": 2e-05,
+      "loss": 0.0402444,
+      "step": 17042
+    },
+    {
+      "epoch": 34.086,
+      "grad_norm": 0.9739708304405212,
+      "learning_rate": 2e-05,
+      "loss": 0.03635871,
+      "step": 17043
+    },
+    {
+      "epoch": 34.088,
+      "grad_norm": 1.5088661909103394,
+      "learning_rate": 2e-05,
+      "loss": 0.03371816,
+      "step": 17044
+    },
+    {
+      "epoch": 34.09,
+      "grad_norm": 1.130311131477356,
+      "learning_rate": 2e-05,
+      "loss": 0.03255319,
+      "step": 17045
+    },
+    {
+      "epoch": 34.092,
+      "grad_norm": 1.0935840606689453,
+      "learning_rate": 2e-05,
+      "loss": 0.03417723,
+      "step": 17046
+    },
+    {
+      "epoch": 34.094,
+      "grad_norm": 1.748842716217041,
+      "learning_rate": 2e-05,
+      "loss": 0.04565499,
+      "step": 17047
+    },
+    {
+      "epoch": 34.096,
+      "grad_norm": 0.9785546064376831,
+      "learning_rate": 2e-05,
+      "loss": 0.02791492,
+      "step": 17048
+    },
+    {
+      "epoch": 34.098,
+      "grad_norm": 1.1637617349624634,
+      "learning_rate": 2e-05,
+      "loss": 0.0296043,
+      "step": 17049
+    },
+    {
+      "epoch": 34.1,
+      "grad_norm": 1.0762447118759155,
+      "learning_rate": 2e-05,
+      "loss": 0.02891765,
+      "step": 17050
+    },
+    {
+      "epoch": 34.102,
+      "grad_norm": 0.9392781257629395,
+      "learning_rate": 2e-05,
+      "loss": 0.02610712,
+      "step": 17051
+    },
+    {
+      "epoch": 34.104,
+      "grad_norm": 1.363243579864502,
+      "learning_rate": 2e-05,
+      "loss": 0.04912724,
+      "step": 17052
+    },
+    {
+      "epoch": 34.106,
+      "grad_norm": 1.1277014017105103,
+      "learning_rate": 2e-05,
+      "loss": 0.0407737,
+      "step": 17053
+    },
+    {
+      "epoch": 34.108,
+      "grad_norm": 1.3390322923660278,
+      "learning_rate": 2e-05,
+      "loss": 0.04672408,
+      "step": 17054
+    },
+    {
+      "epoch": 34.11,
+      "grad_norm": 1.0482418537139893,
+      "learning_rate": 2e-05,
+      "loss": 0.03395636,
+      "step": 17055
+    },
+    {
+      "epoch": 34.112,
+      "grad_norm": 1.1809817552566528,
+      "learning_rate": 2e-05,
+      "loss": 0.04168408,
+      "step": 17056
+    },
+    {
+      "epoch": 34.114,
+      "grad_norm": 1.6997599601745605,
+      "learning_rate": 2e-05,
+      "loss": 0.07434915,
+      "step": 17057
+    },
+    {
+      "epoch": 34.116,
+      "grad_norm": 1.3499462604522705,
+      "learning_rate": 2e-05,
+      "loss": 0.04132972,
+      "step": 17058
+    },
+    {
+      "epoch": 34.118,
+      "grad_norm": 1.1026899814605713,
+      "learning_rate": 2e-05,
+      "loss": 0.04192564,
+      "step": 17059
+    },
+    {
+      "epoch": 34.12,
+      "grad_norm": 1.6073766946792603,
+      "learning_rate": 2e-05,
+      "loss": 0.02773944,
+      "step": 17060
+    },
+    {
+      "epoch": 34.122,
+      "grad_norm": 1.5061614513397217,
+      "learning_rate": 2e-05,
+      "loss": 0.04761346,
+      "step": 17061
+    },
+    {
+      "epoch": 34.124,
+      "grad_norm": 4.682290077209473,
+      "learning_rate": 2e-05,
+      "loss": 0.02792961,
+      "step": 17062
+    },
+    {
+      "epoch": 34.126,
+      "grad_norm": 2.404047966003418,
+      "learning_rate": 2e-05,
+      "loss": 0.03837546,
+      "step": 17063
+    },
+    {
+      "epoch": 34.128,
+      "grad_norm": 1.1497547626495361,
+      "learning_rate": 2e-05,
+      "loss": 0.05234189,
+      "step": 17064
+    },
+    {
+      "epoch": 34.13,
+      "grad_norm": 0.9120892286300659,
+      "learning_rate": 2e-05,
+      "loss": 0.02887437,
+      "step": 17065
+    },
+    {
+      "epoch": 34.132,
+      "grad_norm": 1.2333347797393799,
+      "learning_rate": 2e-05,
+      "loss": 0.04958087,
+      "step": 17066
+    },
+    {
+      "epoch": 34.134,
+      "grad_norm": 1.1911989450454712,
+      "learning_rate": 2e-05,
+      "loss": 0.03579342,
+      "step": 17067
+    },
+    {
+      "epoch": 34.136,
+      "grad_norm": 1.0716452598571777,
+      "learning_rate": 2e-05,
+      "loss": 0.04589354,
+      "step": 17068
+    },
+    {
+      "epoch": 34.138,
+      "grad_norm": 1.1978440284729004,
+      "learning_rate": 2e-05,
+      "loss": 0.04309279,
+      "step": 17069
+    },
+    {
+      "epoch": 34.14,
+      "grad_norm": 1.454819679260254,
+      "learning_rate": 2e-05,
+      "loss": 0.03111471,
+      "step": 17070
+    },
+    {
+      "epoch": 34.142,
+      "grad_norm": 1.3682692050933838,
+      "learning_rate": 2e-05,
+      "loss": 0.0418794,
+      "step": 17071
+    },
+    {
+      "epoch": 34.144,
+      "grad_norm": 1.9202009439468384,
+      "learning_rate": 2e-05,
+      "loss": 0.04135188,
+      "step": 17072
+    },
+    {
+      "epoch": 34.146,
+      "grad_norm": 1.1273765563964844,
+      "learning_rate": 2e-05,
+      "loss": 0.03442348,
+      "step": 17073
+    },
+    {
+      "epoch": 34.148,
+      "grad_norm": 1.3619593381881714,
+      "learning_rate": 2e-05,
+      "loss": 0.04825344,
+      "step": 17074
+    },
+    {
+      "epoch": 34.15,
+      "grad_norm": 1.4348422288894653,
+      "learning_rate": 2e-05,
+      "loss": 0.05776092,
+      "step": 17075
+    },
+    {
+      "epoch": 34.152,
+      "grad_norm": 0.9510125517845154,
+      "learning_rate": 2e-05,
+      "loss": 0.02918218,
+      "step": 17076
+    },
+    {
+      "epoch": 34.154,
+      "grad_norm": 1.260632872581482,
+      "learning_rate": 2e-05,
+      "loss": 0.04130819,
+      "step": 17077
+    },
+    {
+      "epoch": 34.156,
+      "grad_norm": 1.5947790145874023,
+      "learning_rate": 2e-05,
+      "loss": 0.05725566,
+      "step": 17078
+    },
+    {
+      "epoch": 34.158,
+      "grad_norm": 1.1542794704437256,
+      "learning_rate": 2e-05,
+      "loss": 0.04837495,
+      "step": 17079
+    },
+    {
+      "epoch": 34.16,
+      "grad_norm": 1.712775707244873,
+      "learning_rate": 2e-05,
+      "loss": 0.07433301,
+      "step": 17080
+    },
+    {
+      "epoch": 34.162,
+      "grad_norm": 1.1782052516937256,
+      "learning_rate": 2e-05,
+      "loss": 0.0498495,
+      "step": 17081
+    },
+    {
+      "epoch": 34.164,
+      "grad_norm": 1.077620267868042,
+      "learning_rate": 2e-05,
+      "loss": 0.040213,
+      "step": 17082
+    },
+    {
+      "epoch": 34.166,
+      "grad_norm": 1.2808398008346558,
+      "learning_rate": 2e-05,
+      "loss": 0.05944657,
+      "step": 17083
+    },
+    {
+      "epoch": 34.168,
+      "grad_norm": 1.1864197254180908,
+      "learning_rate": 2e-05,
+      "loss": 0.05283065,
+      "step": 17084
+    },
+    {
+      "epoch": 34.17,
+      "grad_norm": 1.0761622190475464,
+      "learning_rate": 2e-05,
+      "loss": 0.03802584,
+      "step": 17085
+    },
+    {
+      "epoch": 34.172,
+      "grad_norm": 1.062628149986267,
+      "learning_rate": 2e-05,
+      "loss": 0.04319198,
+      "step": 17086
+    },
+    {
+      "epoch": 34.174,
+      "grad_norm": 1.46955144405365,
+      "learning_rate": 2e-05,
+      "loss": 0.04133756,
+      "step": 17087
+    },
+    {
+      "epoch": 34.176,
+      "grad_norm": 0.9547522068023682,
+      "learning_rate": 2e-05,
+      "loss": 0.035832,
+      "step": 17088
+    },
+    {
+      "epoch": 34.178,
+      "grad_norm": 4.085507869720459,
+      "learning_rate": 2e-05,
+      "loss": 0.06185192,
+      "step": 17089
+    },
+    {
+      "epoch": 34.18,
+      "grad_norm": 2.0465495586395264,
+      "learning_rate": 2e-05,
+      "loss": 0.04510322,
+      "step": 17090
+    },
+    {
+      "epoch": 34.182,
+      "grad_norm": 1.657041311264038,
+      "learning_rate": 2e-05,
+      "loss": 0.04181718,
+      "step": 17091
+    },
+    {
+      "epoch": 34.184,
+      "grad_norm": 1.1236737966537476,
+      "learning_rate": 2e-05,
+      "loss": 0.0362424,
+      "step": 17092
+    },
+    {
+      "epoch": 34.186,
+      "grad_norm": 1.528167724609375,
+      "learning_rate": 2e-05,
+      "loss": 0.05361965,
+      "step": 17093
+    },
+    {
+      "epoch": 34.188,
+      "grad_norm": 0.9988680481910706,
+      "learning_rate": 2e-05,
+      "loss": 0.03881805,
+      "step": 17094
+    },
+    {
+      "epoch": 34.19,
+      "grad_norm": 2.2689661979675293,
+      "learning_rate": 2e-05,
+      "loss": 0.04616371,
+      "step": 17095
+    },
+    {
+      "epoch": 34.192,
+      "grad_norm": 1.069471836090088,
+      "learning_rate": 2e-05,
+      "loss": 0.0355745,
+      "step": 17096
+    },
+    {
+      "epoch": 34.194,
+      "grad_norm": 1.2013444900512695,
+      "learning_rate": 2e-05,
+      "loss": 0.04738539,
+      "step": 17097
+    },
+    {
+      "epoch": 34.196,
+      "grad_norm": 1.4872270822525024,
+      "learning_rate": 2e-05,
+      "loss": 0.04246665,
+      "step": 17098
+    },
+    {
+      "epoch": 34.198,
+      "grad_norm": 1.2898576259613037,
+      "learning_rate": 2e-05,
+      "loss": 0.03277687,
+      "step": 17099
+    },
+    {
+      "epoch": 34.2,
+      "grad_norm": 0.9258760213851929,
+      "learning_rate": 2e-05,
+      "loss": 0.02966371,
+      "step": 17100
+    },
+    {
+      "epoch": 34.202,
+      "grad_norm": 1.5347909927368164,
+      "learning_rate": 2e-05,
+      "loss": 0.03460214,
+      "step": 17101
+    },
+    {
+      "epoch": 34.204,
+      "grad_norm": 1.5886799097061157,
+      "learning_rate": 2e-05,
+      "loss": 0.04721684,
+      "step": 17102
+    },
+    {
+      "epoch": 34.206,
+      "grad_norm": 0.8684136271476746,
+      "learning_rate": 2e-05,
+      "loss": 0.03214806,
+      "step": 17103
+    },
+    {
+      "epoch": 34.208,
+      "grad_norm": 1.1119003295898438,
+      "learning_rate": 2e-05,
+      "loss": 0.04176969,
+      "step": 17104
+    },
+    {
+      "epoch": 34.21,
+      "grad_norm": 1.462074637413025,
+      "learning_rate": 2e-05,
+      "loss": 0.04227774,
+      "step": 17105
+    },
+    {
+      "epoch": 34.212,
+      "grad_norm": 1.4271862506866455,
+      "learning_rate": 2e-05,
+      "loss": 0.05438799,
+      "step": 17106
+    },
+    {
+      "epoch": 34.214,
+      "grad_norm": 1.3217233419418335,
+      "learning_rate": 2e-05,
+      "loss": 0.05445388,
+      "step": 17107
+    },
+    {
+      "epoch": 34.216,
+      "grad_norm": 0.9007492661476135,
+      "learning_rate": 2e-05,
+      "loss": 0.02505394,
+      "step": 17108
+    },
+    {
+      "epoch": 34.218,
+      "grad_norm": 1.0942732095718384,
+      "learning_rate": 2e-05,
+      "loss": 0.04544266,
+      "step": 17109
+    },
+    {
+      "epoch": 34.22,
+      "grad_norm": 1.4460761547088623,
+      "learning_rate": 2e-05,
+      "loss": 0.04229202,
+      "step": 17110
+    },
+    {
+      "epoch": 34.222,
+      "grad_norm": 1.192115306854248,
+      "learning_rate": 2e-05,
+      "loss": 0.05149196,
+      "step": 17111
+    },
+    {
+      "epoch": 34.224,
+      "grad_norm": 1.0671484470367432,
+      "learning_rate": 2e-05,
+      "loss": 0.03309963,
+      "step": 17112
+    },
+    {
+      "epoch": 34.226,
+      "grad_norm": 1.1515556573867798,
+      "learning_rate": 2e-05,
+      "loss": 0.04172114,
+      "step": 17113
+    },
+    {
+      "epoch": 34.228,
+      "grad_norm": 1.0468469858169556,
+      "learning_rate": 2e-05,
+      "loss": 0.03374918,
+      "step": 17114
+    },
+    {
+      "epoch": 34.23,
+      "grad_norm": 1.1221612691879272,
+      "learning_rate": 2e-05,
+      "loss": 0.05307683,
+      "step": 17115
+    },
+    {
+      "epoch": 34.232,
+      "grad_norm": 1.047747015953064,
+      "learning_rate": 2e-05,
+      "loss": 0.03866299,
+      "step": 17116
+    },
+    {
+      "epoch": 34.234,
+      "grad_norm": 0.9740225076675415,
+      "learning_rate": 2e-05,
+      "loss": 0.02800692,
+      "step": 17117
+    },
+    {
+      "epoch": 34.236,
+      "grad_norm": 1.0027884244918823,
+      "learning_rate": 2e-05,
+      "loss": 0.01938191,
+      "step": 17118
+    },
+    {
+      "epoch": 34.238,
+      "grad_norm": 1.1319332122802734,
+      "learning_rate": 2e-05,
+      "loss": 0.03879067,
+      "step": 17119
+    },
+    {
+      "epoch": 34.24,
+      "grad_norm": 1.2376909255981445,
+      "learning_rate": 2e-05,
+      "loss": 0.04197739,
+      "step": 17120
+    },
+    {
+      "epoch": 34.242,
+      "grad_norm": 1.7739689350128174,
+      "learning_rate": 2e-05,
+      "loss": 0.0463138,
+      "step": 17121
+    },
+    {
+      "epoch": 34.244,
+      "grad_norm": 1.1264393329620361,
+      "learning_rate": 2e-05,
+      "loss": 0.03141479,
+      "step": 17122
+    },
+    {
+      "epoch": 34.246,
+      "grad_norm": 1.261618733406067,
+      "learning_rate": 2e-05,
+      "loss": 0.03376903,
+      "step": 17123
+    },
+    {
+      "epoch": 34.248,
+      "grad_norm": 1.123366355895996,
+      "learning_rate": 2e-05,
+      "loss": 0.04086906,
+      "step": 17124
+    },
+    {
+      "epoch": 34.25,
+      "grad_norm": 2.098238229751587,
+      "learning_rate": 2e-05,
+      "loss": 0.03711101,
+      "step": 17125
+    },
+    {
+      "epoch": 34.252,
+      "grad_norm": 1.1595600843429565,
+      "learning_rate": 2e-05,
+      "loss": 0.04579591,
+      "step": 17126
+    },
+    {
+      "epoch": 34.254,
+      "grad_norm": 1.2296733856201172,
+      "learning_rate": 2e-05,
+      "loss": 0.05098995,
+      "step": 17127
+    },
+    {
+      "epoch": 34.256,
+      "grad_norm": 1.07576322555542,
+      "learning_rate": 2e-05,
+      "loss": 0.04090093,
+      "step": 17128
+    },
+    {
+      "epoch": 34.258,
+      "grad_norm": 1.4022080898284912,
+      "learning_rate": 2e-05,
+      "loss": 0.05891519,
+      "step": 17129
+    },
+    {
+      "epoch": 34.26,
+      "grad_norm": 1.2884039878845215,
+      "learning_rate": 2e-05,
+      "loss": 0.06018764,
+      "step": 17130
+    },
+    {
+      "epoch": 34.262,
+      "grad_norm": 1.513996958732605,
+      "learning_rate": 2e-05,
+      "loss": 0.05563644,
+      "step": 17131
+    },
+    {
+      "epoch": 34.264,
+      "grad_norm": 1.0170012712478638,
+      "learning_rate": 2e-05,
+      "loss": 0.0308474,
+      "step": 17132
+    },
+    {
+      "epoch": 34.266,
+      "grad_norm": 1.7678961753845215,
+      "learning_rate": 2e-05,
+      "loss": 0.03012414,
+      "step": 17133
+    },
+    {
+      "epoch": 34.268,
+      "grad_norm": 1.930487036705017,
+      "learning_rate": 2e-05,
+      "loss": 0.03921436,
+      "step": 17134
+    },
+    {
+      "epoch": 34.27,
+      "grad_norm": 1.1381620168685913,
+      "learning_rate": 2e-05,
+      "loss": 0.04433856,
+      "step": 17135
+    },
+    {
+      "epoch": 34.272,
+      "grad_norm": 1.3920421600341797,
+      "learning_rate": 2e-05,
+      "loss": 0.04576306,
+      "step": 17136
+    },
+    {
+      "epoch": 34.274,
+      "grad_norm": 1.132322907447815,
+      "learning_rate": 2e-05,
+      "loss": 0.04681329,
+      "step": 17137
+    },
+    {
+      "epoch": 34.276,
+      "grad_norm": 1.0427021980285645,
+      "learning_rate": 2e-05,
+      "loss": 0.03912016,
+      "step": 17138
+    },
+    {
+      "epoch": 34.278,
+      "grad_norm": 1.3596690893173218,
+      "learning_rate": 2e-05,
+      "loss": 0.0395805,
+      "step": 17139
+    },
+    {
+      "epoch": 34.28,
+      "grad_norm": 1.1399190425872803,
+      "learning_rate": 2e-05,
+      "loss": 0.03435064,
+      "step": 17140
+    },
+    {
+      "epoch": 34.282,
+      "grad_norm": 1.2819249629974365,
+      "learning_rate": 2e-05,
+      "loss": 0.0416284,
+      "step": 17141
+    },
+    {
+      "epoch": 34.284,
+      "grad_norm": 1.2883572578430176,
+      "learning_rate": 2e-05,
+      "loss": 0.02817941,
+      "step": 17142
+    },
+    {
+      "epoch": 34.286,
+      "grad_norm": 1.449006199836731,
+      "learning_rate": 2e-05,
+      "loss": 0.05150205,
+      "step": 17143
+    },
+    {
+      "epoch": 34.288,
+      "grad_norm": 1.3080250024795532,
+      "learning_rate": 2e-05,
+      "loss": 0.03051752,
+      "step": 17144
+    },
+    {
+      "epoch": 34.29,
+      "grad_norm": 1.1870465278625488,
+      "learning_rate": 2e-05,
+      "loss": 0.04249543,
+      "step": 17145
+    },
+    {
+      "epoch": 34.292,
+      "grad_norm": 1.2269293069839478,
+      "learning_rate": 2e-05,
+      "loss": 0.0272982,
+      "step": 17146
+    },
+    {
+      "epoch": 34.294,
+      "grad_norm": 1.2062244415283203,
+      "learning_rate": 2e-05,
+      "loss": 0.04380896,
+      "step": 17147
+    },
+    {
+      "epoch": 34.296,
+      "grad_norm": 1.1690454483032227,
+      "learning_rate": 2e-05,
+      "loss": 0.04224586,
+      "step": 17148
+    },
+    {
+      "epoch": 34.298,
+      "grad_norm": 1.076251745223999,
+      "learning_rate": 2e-05,
+      "loss": 0.04731111,
+      "step": 17149
+    },
+    {
+      "epoch": 34.3,
+      "grad_norm": 0.9125476479530334,
+      "learning_rate": 2e-05,
+      "loss": 0.03293589,
+      "step": 17150
+    },
+    {
+      "epoch": 34.302,
+      "grad_norm": 1.1172863245010376,
+      "learning_rate": 2e-05,
+      "loss": 0.04434405,
+      "step": 17151
+    },
+    {
+      "epoch": 34.304,
+      "grad_norm": 1.4537452459335327,
+      "learning_rate": 2e-05,
+      "loss": 0.05699692,
+      "step": 17152
+    },
+    {
+      "epoch": 34.306,
+      "grad_norm": 1.2859165668487549,
+      "learning_rate": 2e-05,
+      "loss": 0.05222432,
+      "step": 17153
+    },
+    {
+      "epoch": 34.308,
+      "grad_norm": 1.1391727924346924,
+      "learning_rate": 2e-05,
+      "loss": 0.04175641,
+      "step": 17154
+    },
+    {
+      "epoch": 34.31,
+      "grad_norm": 1.3534083366394043,
+      "learning_rate": 2e-05,
+      "loss": 0.04102936,
+      "step": 17155
+    },
+    {
+      "epoch": 34.312,
+      "grad_norm": 2.2053699493408203,
+      "learning_rate": 2e-05,
+      "loss": 0.04537173,
+      "step": 17156
+    },
+    {
+      "epoch": 34.314,
+      "grad_norm": 1.1330358982086182,
+      "learning_rate": 2e-05,
+      "loss": 0.05229243,
+      "step": 17157
+    },
+    {
+      "epoch": 34.316,
+      "grad_norm": 1.8947052955627441,
+      "learning_rate": 2e-05,
+      "loss": 0.02641325,
+      "step": 17158
+    },
+    {
+      "epoch": 34.318,
+      "grad_norm": 1.3450164794921875,
+      "learning_rate": 2e-05,
+      "loss": 0.04416771,
+      "step": 17159
+    },
+    {
+      "epoch": 34.32,
+      "grad_norm": 2.277355909347534,
+      "learning_rate": 2e-05,
+      "loss": 0.03164542,
+      "step": 17160
+    },
+    {
+      "epoch": 34.322,
+      "grad_norm": 1.2452149391174316,
+      "learning_rate": 2e-05,
+      "loss": 0.03742987,
+      "step": 17161
+    },
+    {
+      "epoch": 34.324,
+      "grad_norm": 1.3950529098510742,
+      "learning_rate": 2e-05,
+      "loss": 0.04363611,
+      "step": 17162
+    },
+    {
+      "epoch": 34.326,
+      "grad_norm": 1.2250491380691528,
+      "learning_rate": 2e-05,
+      "loss": 0.04496596,
+      "step": 17163
+    },
+    {
+      "epoch": 34.328,
+      "grad_norm": 1.1004265546798706,
+      "learning_rate": 2e-05,
+      "loss": 0.04556531,
+      "step": 17164
+    },
+    {
+      "epoch": 34.33,
+      "grad_norm": 1.185380458831787,
+      "learning_rate": 2e-05,
+      "loss": 0.0397131,
+      "step": 17165
+    },
+    {
+      "epoch": 34.332,
+      "grad_norm": 1.1757383346557617,
+      "learning_rate": 2e-05,
+      "loss": 0.04187442,
+      "step": 17166
+    },
+    {
+      "epoch": 34.334,
+      "grad_norm": 1.0646100044250488,
+      "learning_rate": 2e-05,
+      "loss": 0.04621556,
+      "step": 17167
+    },
+    {
+      "epoch": 34.336,
+      "grad_norm": 1.1299614906311035,
+      "learning_rate": 2e-05,
+      "loss": 0.05131941,
+      "step": 17168
+    },
+    {
+      "epoch": 34.338,
+      "grad_norm": 1.2977094650268555,
+      "learning_rate": 2e-05,
+      "loss": 0.04573498,
+      "step": 17169
+    },
+    {
+      "epoch": 34.34,
+      "grad_norm": 0.9921538829803467,
+      "learning_rate": 2e-05,
+      "loss": 0.03972767,
+      "step": 17170
+    },
+    {
+      "epoch": 34.342,
+      "grad_norm": 1.0485543012619019,
+      "learning_rate": 2e-05,
+      "loss": 0.03102573,
+      "step": 17171
+    },
+    {
+      "epoch": 34.344,
+      "grad_norm": 3.1085667610168457,
+      "learning_rate": 2e-05,
+      "loss": 0.05401374,
+      "step": 17172
+    },
+    {
+      "epoch": 34.346,
+      "grad_norm": 1.159866213798523,
+      "learning_rate": 2e-05,
+      "loss": 0.03029288,
+      "step": 17173
+    },
+    {
+      "epoch": 34.348,
+      "grad_norm": 1.4898662567138672,
+      "learning_rate": 2e-05,
+      "loss": 0.037357,
+      "step": 17174
+    },
+    {
+      "epoch": 34.35,
+      "grad_norm": 1.475274920463562,
+      "learning_rate": 2e-05,
+      "loss": 0.06047164,
+      "step": 17175
+    },
+    {
+      "epoch": 34.352,
+      "grad_norm": 1.2670931816101074,
+      "learning_rate": 2e-05,
+      "loss": 0.04098585,
+      "step": 17176
+    },
+    {
+      "epoch": 34.354,
+      "grad_norm": 1.1112204790115356,
+      "learning_rate": 2e-05,
+      "loss": 0.04078077,
+      "step": 17177
+    },
+    {
+      "epoch": 34.356,
+      "grad_norm": 1.1330214738845825,
+      "learning_rate": 2e-05,
+      "loss": 0.04091879,
+      "step": 17178
+    },
+    {
+      "epoch": 34.358,
+      "grad_norm": 1.0863505601882935,
+      "learning_rate": 2e-05,
+      "loss": 0.04044025,
+      "step": 17179
+    },
+    {
+      "epoch": 34.36,
+      "grad_norm": 1.4287241697311401,
+      "learning_rate": 2e-05,
+      "loss": 0.05269104,
+      "step": 17180
+    },
+    {
+      "epoch": 34.362,
+      "grad_norm": 1.0420600175857544,
+      "learning_rate": 2e-05,
+      "loss": 0.03304556,
+      "step": 17181
+    },
+    {
+      "epoch": 34.364,
+      "grad_norm": 1.1292872428894043,
+      "learning_rate": 2e-05,
+      "loss": 0.04075188,
+      "step": 17182
+    },
+    {
+      "epoch": 34.366,
+      "grad_norm": 1.257199764251709,
+      "learning_rate": 2e-05,
+      "loss": 0.06179626,
+      "step": 17183
+    },
+    {
+      "epoch": 34.368,
+      "grad_norm": 1.5898911952972412,
+      "learning_rate": 2e-05,
+      "loss": 0.06544568,
+      "step": 17184
+    },
+    {
+      "epoch": 34.37,
+      "grad_norm": 1.1866689920425415,
+      "learning_rate": 2e-05,
+      "loss": 0.03154187,
+      "step": 17185
+    },
+    {
+      "epoch": 34.372,
+      "grad_norm": 1.1588656902313232,
+      "learning_rate": 2e-05,
+      "loss": 0.04455046,
+      "step": 17186
+    },
+    {
+      "epoch": 34.374,
+      "grad_norm": 1.8778082132339478,
+      "learning_rate": 2e-05,
+      "loss": 0.05956361,
+      "step": 17187
+    },
+    {
+      "epoch": 34.376,
+      "grad_norm": 1.0748698711395264,
+      "learning_rate": 2e-05,
+      "loss": 0.02959104,
+      "step": 17188
+    },
+    {
+      "epoch": 34.378,
+      "grad_norm": 1.2204198837280273,
+      "learning_rate": 2e-05,
+      "loss": 0.0491489,
+      "step": 17189
+    },
+    {
+      "epoch": 34.38,
+      "grad_norm": 1.1138110160827637,
+      "learning_rate": 2e-05,
+      "loss": 0.04551839,
+      "step": 17190
+    },
+    {
+      "epoch": 34.382,
+      "grad_norm": 1.057921051979065,
+      "learning_rate": 2e-05,
+      "loss": 0.04800026,
+      "step": 17191
+    },
+    {
+      "epoch": 34.384,
+      "grad_norm": 0.97761470079422,
+      "learning_rate": 2e-05,
+      "loss": 0.03815104,
+      "step": 17192
+    },
+    {
+      "epoch": 34.386,
+      "grad_norm": 1.2259156703948975,
+      "learning_rate": 2e-05,
+      "loss": 0.03457344,
+      "step": 17193
+    },
+    {
+      "epoch": 34.388,
+      "grad_norm": 1.6335809230804443,
+      "learning_rate": 2e-05,
+      "loss": 0.05803145,
+      "step": 17194
+    },
+    {
+      "epoch": 34.39,
+      "grad_norm": 1.134580135345459,
+      "learning_rate": 2e-05,
+      "loss": 0.03817577,
+      "step": 17195
+    },
+    {
+      "epoch": 34.392,
+      "grad_norm": 1.2042521238327026,
+      "learning_rate": 2e-05,
+      "loss": 0.04776943,
+      "step": 17196
+    },
+    {
+      "epoch": 34.394,
+      "grad_norm": 0.9557961225509644,
+      "learning_rate": 2e-05,
+      "loss": 0.02998234,
+      "step": 17197
+    },
+    {
+      "epoch": 34.396,
+      "grad_norm": 1.669502854347229,
+      "learning_rate": 2e-05,
+      "loss": 0.05116883,
+      "step": 17198
+    },
+    {
+      "epoch": 34.398,
+      "grad_norm": 1.938193917274475,
+      "learning_rate": 2e-05,
+      "loss": 0.04275124,
+      "step": 17199
+    },
+    {
+      "epoch": 34.4,
+      "grad_norm": 0.95672208070755,
+      "learning_rate": 2e-05,
+      "loss": 0.02722719,
+      "step": 17200
+    },
+    {
+      "epoch": 34.402,
+      "grad_norm": 1.6504557132720947,
+      "learning_rate": 2e-05,
+      "loss": 0.04932757,
+      "step": 17201
+    },
+    {
+      "epoch": 34.404,
+      "grad_norm": 1.3142409324645996,
+      "learning_rate": 2e-05,
+      "loss": 0.05078411,
+      "step": 17202
+    },
+    {
+      "epoch": 34.406,
+      "grad_norm": 1.8775172233581543,
+      "learning_rate": 2e-05,
+      "loss": 0.0415501,
+      "step": 17203
+    },
+    {
+      "epoch": 34.408,
+      "grad_norm": 1.108086347579956,
+      "learning_rate": 2e-05,
+      "loss": 0.03562623,
+      "step": 17204
+    },
+    {
+      "epoch": 34.41,
+      "grad_norm": 1.1936184167861938,
+      "learning_rate": 2e-05,
+      "loss": 0.06731561,
+      "step": 17205
+    },
+    {
+      "epoch": 34.412,
+      "grad_norm": 1.4935089349746704,
+      "learning_rate": 2e-05,
+      "loss": 0.04526928,
+      "step": 17206
+    },
+    {
+      "epoch": 34.414,
+      "grad_norm": 0.9472814798355103,
+      "learning_rate": 2e-05,
+      "loss": 0.03531752,
+      "step": 17207
+    },
+    {
+      "epoch": 34.416,
+      "grad_norm": 1.1096245050430298,
+      "learning_rate": 2e-05,
+      "loss": 0.03912827,
+      "step": 17208
+    },
+    {
+      "epoch": 34.418,
+      "grad_norm": 1.2029078006744385,
+      "learning_rate": 2e-05,
+      "loss": 0.05240467,
+      "step": 17209
+    },
+    {
+      "epoch": 34.42,
+      "grad_norm": 1.08064603805542,
+      "learning_rate": 2e-05,
+      "loss": 0.04623947,
+      "step": 17210
+    },
+    {
+      "epoch": 34.422,
+      "grad_norm": 1.3892536163330078,
+      "learning_rate": 2e-05,
+      "loss": 0.04560886,
+      "step": 17211
+    },
+    {
+      "epoch": 34.424,
+      "grad_norm": 1.2040435075759888,
+      "learning_rate": 2e-05,
+      "loss": 0.03983107,
+      "step": 17212
+    },
+    {
+      "epoch": 34.426,
+      "grad_norm": 1.0492370128631592,
+      "learning_rate": 2e-05,
+      "loss": 0.03636064,
+      "step": 17213
+    },
+    {
+      "epoch": 34.428,
+      "grad_norm": 1.1657490730285645,
+      "learning_rate": 2e-05,
+      "loss": 0.04331504,
+      "step": 17214
+    },
+    {
+      "epoch": 34.43,
+      "grad_norm": 1.2659703493118286,
+      "learning_rate": 2e-05,
+      "loss": 0.05003392,
+      "step": 17215
+    },
+    {
+      "epoch": 34.432,
+      "grad_norm": 1.3494330644607544,
+      "learning_rate": 2e-05,
+      "loss": 0.02932413,
+      "step": 17216
+    },
+    {
+      "epoch": 34.434,
+      "grad_norm": 1.175642490386963,
+      "learning_rate": 2e-05,
+      "loss": 0.04383037,
+      "step": 17217
+    },
+    {
+      "epoch": 34.436,
+      "grad_norm": 1.0933257341384888,
+      "learning_rate": 2e-05,
+      "loss": 0.03921406,
+      "step": 17218
+    },
+    {
+      "epoch": 34.438,
+      "grad_norm": 1.1617385149002075,
+      "learning_rate": 2e-05,
+      "loss": 0.04727744,
+      "step": 17219
+    },
+    {
+      "epoch": 34.44,
+      "grad_norm": 1.1515668630599976,
+      "learning_rate": 2e-05,
+      "loss": 0.03612133,
+      "step": 17220
+    },
+    {
+      "epoch": 34.442,
+      "grad_norm": 1.1153593063354492,
+      "learning_rate": 2e-05,
+      "loss": 0.03581006,
+      "step": 17221
+    },
+    {
+      "epoch": 34.444,
+      "grad_norm": 1.1779838800430298,
+      "learning_rate": 2e-05,
+      "loss": 0.03837469,
+      "step": 17222
+    },
+    {
+      "epoch": 34.446,
+      "grad_norm": 1.257264256477356,
+      "learning_rate": 2e-05,
+      "loss": 0.04428399,
+      "step": 17223
+    },
+    {
+      "epoch": 34.448,
+      "grad_norm": 1.1216603517532349,
+      "learning_rate": 2e-05,
+      "loss": 0.03815703,
+      "step": 17224
+    },
+    {
+      "epoch": 34.45,
+      "grad_norm": 1.513350248336792,
+      "learning_rate": 2e-05,
+      "loss": 0.05262036,
+      "step": 17225
+    },
+    {
+      "epoch": 34.452,
+      "grad_norm": 1.668152093887329,
+      "learning_rate": 2e-05,
+      "loss": 0.04921922,
+      "step": 17226
+    },
+    {
+      "epoch": 34.454,
+      "grad_norm": 1.259291648864746,
+      "learning_rate": 2e-05,
+      "loss": 0.0468533,
+      "step": 17227
+    },
+    {
+      "epoch": 34.456,
+      "grad_norm": 2.0547330379486084,
+      "learning_rate": 2e-05,
+      "loss": 0.04505831,
+      "step": 17228
+    },
+    {
+      "epoch": 34.458,
+      "grad_norm": 1.0175834894180298,
+      "learning_rate": 2e-05,
+      "loss": 0.03889389,
+      "step": 17229
+    },
+    {
+      "epoch": 34.46,
+      "grad_norm": 1.205269694328308,
+      "learning_rate": 2e-05,
+      "loss": 0.05405382,
+      "step": 17230
+    },
+    {
+      "epoch": 34.462,
+      "grad_norm": 1.1954044103622437,
+      "learning_rate": 2e-05,
+      "loss": 0.06115945,
+      "step": 17231
+    },
+    {
+      "epoch": 34.464,
+      "grad_norm": 1.5559508800506592,
+      "learning_rate": 2e-05,
+      "loss": 0.04823997,
+      "step": 17232
+    },
+    {
+      "epoch": 34.466,
+      "grad_norm": 1.123988151550293,
+      "learning_rate": 2e-05,
+      "loss": 0.04282939,
+      "step": 17233
+    },
+    {
+      "epoch": 34.468,
+      "grad_norm": 1.5351160764694214,
+      "learning_rate": 2e-05,
+      "loss": 0.04434042,
+      "step": 17234
+    },
+    {
+      "epoch": 34.47,
+      "grad_norm": 1.1878026723861694,
+      "learning_rate": 2e-05,
+      "loss": 0.04576051,
+      "step": 17235
+    },
+    {
+      "epoch": 34.472,
+      "grad_norm": 1.456591248512268,
+      "learning_rate": 2e-05,
+      "loss": 0.04699751,
+      "step": 17236
+    },
+    {
+      "epoch": 34.474,
+      "grad_norm": 1.0269306898117065,
+      "learning_rate": 2e-05,
+      "loss": 0.04063164,
+      "step": 17237
+    },
+    {
+      "epoch": 34.476,
+      "grad_norm": 1.0819087028503418,
+      "learning_rate": 2e-05,
+      "loss": 0.04128548,
+      "step": 17238
+    },
+    {
+      "epoch": 34.478,
+      "grad_norm": 1.8286333084106445,
+      "learning_rate": 2e-05,
+      "loss": 0.04910714,
+      "step": 17239
+    },
+    {
+      "epoch": 34.48,
+      "grad_norm": 3.0177881717681885,
+      "learning_rate": 2e-05,
+      "loss": 0.05242274,
+      "step": 17240
+    },
+    {
+      "epoch": 34.482,
+      "grad_norm": 1.391753911972046,
+      "learning_rate": 2e-05,
+      "loss": 0.04662409,
+      "step": 17241
+    },
+    {
+      "epoch": 34.484,
+      "grad_norm": 1.5044915676116943,
+      "learning_rate": 2e-05,
+      "loss": 0.03636048,
+      "step": 17242
+    },
+    {
+      "epoch": 34.486,
+      "grad_norm": 1.38352370262146,
+      "learning_rate": 2e-05,
+      "loss": 0.04130649,
+      "step": 17243
+    },
+    {
+      "epoch": 34.488,
+      "grad_norm": 1.027741551399231,
+      "learning_rate": 2e-05,
+      "loss": 0.04034343,
+      "step": 17244
+    },
+    {
+      "epoch": 34.49,
+      "grad_norm": 1.7311643362045288,
+      "learning_rate": 2e-05,
+      "loss": 0.03252632,
+      "step": 17245
+    },
+    {
+      "epoch": 34.492,
+      "grad_norm": 0.9958137273788452,
+      "learning_rate": 2e-05,
+      "loss": 0.03432687,
+      "step": 17246
+    },
+    {
+      "epoch": 34.494,
+      "grad_norm": 0.9350914359092712,
+      "learning_rate": 2e-05,
+      "loss": 0.04081878,
+      "step": 17247
+    },
+    {
+      "epoch": 34.496,
+      "grad_norm": 1.0058718919754028,
+      "learning_rate": 2e-05,
+      "loss": 0.03128472,
+      "step": 17248
+    },
+    {
+      "epoch": 34.498,
+      "grad_norm": 1.1736693382263184,
+      "learning_rate": 2e-05,
+      "loss": 0.0507132,
+      "step": 17249
+    },
+    {
+      "epoch": 34.5,
+      "grad_norm": 2.9509620666503906,
+      "learning_rate": 2e-05,
+      "loss": 0.04513638,
+      "step": 17250
+    },
+    {
+      "epoch": 34.502,
+      "grad_norm": 1.1163856983184814,
+      "learning_rate": 2e-05,
+      "loss": 0.02961075,
+      "step": 17251
+    },
+    {
+      "epoch": 34.504,
+      "grad_norm": 1.2255582809448242,
+      "learning_rate": 2e-05,
+      "loss": 0.05628613,
+      "step": 17252
+    },
+    {
+      "epoch": 34.506,
+      "grad_norm": 1.1547157764434814,
+      "learning_rate": 2e-05,
+      "loss": 0.04185841,
+      "step": 17253
+    },
+    {
+      "epoch": 34.508,
+      "grad_norm": 1.0533969402313232,
+      "learning_rate": 2e-05,
+      "loss": 0.02807963,
+      "step": 17254
+    },
+    {
+      "epoch": 34.51,
+      "grad_norm": 1.1036581993103027,
+      "learning_rate": 2e-05,
+      "loss": 0.04609573,
+      "step": 17255
+    },
+    {
+      "epoch": 34.512,
+      "grad_norm": 1.2565536499023438,
+      "learning_rate": 2e-05,
+      "loss": 0.0547157,
+      "step": 17256
+    },
+    {
+      "epoch": 34.514,
+      "grad_norm": 1.1017217636108398,
+      "learning_rate": 2e-05,
+      "loss": 0.05001441,
+      "step": 17257
+    },
+    {
+      "epoch": 34.516,
+      "grad_norm": 1.0183624029159546,
+      "learning_rate": 2e-05,
+      "loss": 0.03306765,
+      "step": 17258
+    },
+    {
+      "epoch": 34.518,
+      "grad_norm": 1.5634461641311646,
+      "learning_rate": 2e-05,
+      "loss": 0.05358353,
+      "step": 17259
+    },
+    {
+      "epoch": 34.52,
+      "grad_norm": 0.8560883402824402,
+      "learning_rate": 2e-05,
+      "loss": 0.02674874,
+      "step": 17260
+    },
+    {
+      "epoch": 34.522,
+      "grad_norm": 1.7791016101837158,
+      "learning_rate": 2e-05,
+      "loss": 0.05959473,
+      "step": 17261
+    },
+    {
+      "epoch": 34.524,
+      "grad_norm": 1.059211015701294,
+      "learning_rate": 2e-05,
+      "loss": 0.04623821,
+      "step": 17262
+    },
+    {
+      "epoch": 34.526,
+      "grad_norm": 2.1442856788635254,
+      "learning_rate": 2e-05,
+      "loss": 0.03975558,
+      "step": 17263
+    },
+    {
+      "epoch": 34.528,
+      "grad_norm": 1.1825506687164307,
+      "learning_rate": 2e-05,
+      "loss": 0.04681484,
+      "step": 17264
+    },
+    {
+      "epoch": 34.53,
+      "grad_norm": 1.1355949640274048,
+      "learning_rate": 2e-05,
+      "loss": 0.04300448,
+      "step": 17265
+    },
+    {
+      "epoch": 34.532,
+      "grad_norm": 1.151884913444519,
+      "learning_rate": 2e-05,
+      "loss": 0.05281693,
+      "step": 17266
+    },
+    {
+      "epoch": 34.534,
+      "grad_norm": 1.244160532951355,
+      "learning_rate": 2e-05,
+      "loss": 0.05270495,
+      "step": 17267
+    },
+    {
+      "epoch": 34.536,
+      "grad_norm": 1.332995057106018,
+      "learning_rate": 2e-05,
+      "loss": 0.04768827,
+      "step": 17268
+    },
+    {
+      "epoch": 34.538,
+      "grad_norm": 1.1446064710617065,
+      "learning_rate": 2e-05,
+      "loss": 0.03133503,
+      "step": 17269
+    },
+    {
+      "epoch": 34.54,
+      "grad_norm": 1.1863194704055786,
+      "learning_rate": 2e-05,
+      "loss": 0.05849977,
+      "step": 17270
+    },
+    {
+      "epoch": 34.542,
+      "grad_norm": 1.2308465242385864,
+      "learning_rate": 2e-05,
+      "loss": 0.06080481,
+      "step": 17271
+    },
+    {
+      "epoch": 34.544,
+      "grad_norm": 2.3571889400482178,
+      "learning_rate": 2e-05,
+      "loss": 0.05764741,
+      "step": 17272
+    },
+    {
+      "epoch": 34.546,
+      "grad_norm": 0.713600218296051,
+      "learning_rate": 2e-05,
+      "loss": 0.0190373,
+      "step": 17273
+    },
+    {
+      "epoch": 34.548,
+      "grad_norm": 3.5225939750671387,
+      "learning_rate": 2e-05,
+      "loss": 0.05001704,
+      "step": 17274
+    },
+    {
+      "epoch": 34.55,
+      "grad_norm": 2.5222723484039307,
+      "learning_rate": 2e-05,
+      "loss": 0.03726556,
+      "step": 17275
+    },
+    {
+      "epoch": 34.552,
+      "grad_norm": 1.0584102869033813,
+      "learning_rate": 2e-05,
+      "loss": 0.04141295,
+      "step": 17276
+    },
+    {
+      "epoch": 34.554,
+      "grad_norm": 1.0614216327667236,
+      "learning_rate": 2e-05,
+      "loss": 0.04041729,
+      "step": 17277
+    },
+    {
+      "epoch": 34.556,
+      "grad_norm": 0.8729689121246338,
+      "learning_rate": 2e-05,
+      "loss": 0.03110387,
+      "step": 17278
+    },
+    {
+      "epoch": 34.558,
+      "grad_norm": 1.013295292854309,
+      "learning_rate": 2e-05,
+      "loss": 0.03849518,
+      "step": 17279
+    },
+    {
+      "epoch": 34.56,
+      "grad_norm": 1.2050495147705078,
+      "learning_rate": 2e-05,
+      "loss": 0.04337199,
+      "step": 17280
+    },
+    {
+      "epoch": 34.562,
+      "grad_norm": 1.1820077896118164,
+      "learning_rate": 2e-05,
+      "loss": 0.04693332,
+      "step": 17281
+    },
+    {
+      "epoch": 34.564,
+      "grad_norm": 1.015336036682129,
+      "learning_rate": 2e-05,
+      "loss": 0.0314662,
+      "step": 17282
+    },
+    {
+      "epoch": 34.566,
+      "grad_norm": 1.1241815090179443,
+      "learning_rate": 2e-05,
+      "loss": 0.04584487,
+      "step": 17283
+    },
+    {
+      "epoch": 34.568,
+      "grad_norm": 1.4243446588516235,
+      "learning_rate": 2e-05,
+      "loss": 0.04570295,
+      "step": 17284
+    },
+    {
+      "epoch": 34.57,
+      "grad_norm": 2.7489190101623535,
+      "learning_rate": 2e-05,
+      "loss": 0.04503506,
+      "step": 17285
+    },
+    {
+      "epoch": 34.572,
+      "grad_norm": 1.1095118522644043,
+      "learning_rate": 2e-05,
+      "loss": 0.04129759,
+      "step": 17286
+    },
+    {
+      "epoch": 34.574,
+      "grad_norm": 1.1222656965255737,
+      "learning_rate": 2e-05,
+      "loss": 0.04352672,
+      "step": 17287
+    },
+    {
+      "epoch": 34.576,
+      "grad_norm": 1.050182580947876,
+      "learning_rate": 2e-05,
+      "loss": 0.04432563,
+      "step": 17288
+    },
+    {
+      "epoch": 34.578,
+      "grad_norm": 1.1200228929519653,
+      "learning_rate": 2e-05,
+      "loss": 0.03937971,
+      "step": 17289
+    },
+    {
+      "epoch": 34.58,
+      "grad_norm": 1.0767937898635864,
+      "learning_rate": 2e-05,
+      "loss": 0.04535668,
+      "step": 17290
+    },
+    {
+      "epoch": 34.582,
+      "grad_norm": 1.556679129600525,
+      "learning_rate": 2e-05,
+      "loss": 0.05495925,
+      "step": 17291
+    },
+    {
+      "epoch": 34.584,
+      "grad_norm": 0.9800344109535217,
+      "learning_rate": 2e-05,
+      "loss": 0.03302306,
+      "step": 17292
+    },
+    {
+      "epoch": 34.586,
+      "grad_norm": 1.1676750183105469,
+      "learning_rate": 2e-05,
+      "loss": 0.04421997,
+      "step": 17293
+    },
+    {
+      "epoch": 34.588,
+      "grad_norm": 1.9742448329925537,
+      "learning_rate": 2e-05,
+      "loss": 0.05324573,
+      "step": 17294
+    },
+    {
+      "epoch": 34.59,
+      "grad_norm": 1.217177152633667,
+      "learning_rate": 2e-05,
+      "loss": 0.03502582,
+      "step": 17295
+    },
+    {
+      "epoch": 34.592,
+      "grad_norm": 1.535559058189392,
+      "learning_rate": 2e-05,
+      "loss": 0.05487555,
+      "step": 17296
+    },
+    {
+      "epoch": 34.594,
+      "grad_norm": 1.073294758796692,
+      "learning_rate": 2e-05,
+      "loss": 0.04047548,
+      "step": 17297
+    },
+    {
+      "epoch": 34.596,
+      "grad_norm": 1.468457579612732,
+      "learning_rate": 2e-05,
+      "loss": 0.03949284,
+      "step": 17298
+    },
+    {
+      "epoch": 34.598,
+      "grad_norm": 1.1987510919570923,
+      "learning_rate": 2e-05,
+      "loss": 0.038359,
+      "step": 17299
+    },
+    {
+      "epoch": 34.6,
+      "grad_norm": 1.8533692359924316,
+      "learning_rate": 2e-05,
+      "loss": 0.04139744,
+      "step": 17300
+    },
+    {
+      "epoch": 34.602,
+      "grad_norm": 0.9457276463508606,
+      "learning_rate": 2e-05,
+      "loss": 0.03080427,
+      "step": 17301
+    },
+    {
+      "epoch": 34.604,
+      "grad_norm": 1.3518832921981812,
+      "learning_rate": 2e-05,
+      "loss": 0.04552808,
+      "step": 17302
+    },
+    {
+      "epoch": 34.606,
+      "grad_norm": 1.7731199264526367,
+      "learning_rate": 2e-05,
+      "loss": 0.03314227,
+      "step": 17303
+    },
+    {
+      "epoch": 34.608,
+      "grad_norm": 1.6330634355545044,
+      "learning_rate": 2e-05,
+      "loss": 0.04139177,
+      "step": 17304
+    },
+    {
+      "epoch": 34.61,
+      "grad_norm": 1.111138939857483,
+      "learning_rate": 2e-05,
+      "loss": 0.03105775,
+      "step": 17305
+    },
+    {
+      "epoch": 34.612,
+      "grad_norm": 1.2260620594024658,
+      "learning_rate": 2e-05,
+      "loss": 0.0478155,
+      "step": 17306
+    },
+    {
+      "epoch": 34.614,
+      "grad_norm": 1.1791749000549316,
+      "learning_rate": 2e-05,
+      "loss": 0.04237508,
+      "step": 17307
+    },
+    {
+      "epoch": 34.616,
+      "grad_norm": 0.9272366166114807,
+      "learning_rate": 2e-05,
+      "loss": 0.03472099,
+      "step": 17308
+    },
+    {
+      "epoch": 34.618,
+      "grad_norm": 1.0389100313186646,
+      "learning_rate": 2e-05,
+      "loss": 0.03994408,
+      "step": 17309
+    },
+    {
+      "epoch": 34.62,
+      "grad_norm": 1.1838860511779785,
+      "learning_rate": 2e-05,
+      "loss": 0.03435428,
+      "step": 17310
+    },
+    {
+      "epoch": 34.622,
+      "grad_norm": 1.240273356437683,
+      "learning_rate": 2e-05,
+      "loss": 0.04923018,
+      "step": 17311
+    },
+    {
+      "epoch": 34.624,
+      "grad_norm": 1.0485180616378784,
+      "learning_rate": 2e-05,
+      "loss": 0.02882761,
+      "step": 17312
+    },
+    {
+      "epoch": 34.626,
+      "grad_norm": 1.092466950416565,
+      "learning_rate": 2e-05,
+      "loss": 0.02998949,
+      "step": 17313
+    },
+    {
+      "epoch": 34.628,
+      "grad_norm": 2.4525039196014404,
+      "learning_rate": 2e-05,
+      "loss": 0.05779274,
+      "step": 17314
+    },
+    {
+      "epoch": 34.63,
+      "grad_norm": 1.1219778060913086,
+      "learning_rate": 2e-05,
+      "loss": 0.03857387,
+      "step": 17315
+    },
+    {
+      "epoch": 34.632,
+      "grad_norm": 1.2241005897521973,
+      "learning_rate": 2e-05,
+      "loss": 0.04050406,
+      "step": 17316
+    },
+    {
+      "epoch": 34.634,
+      "grad_norm": 1.1506431102752686,
+      "learning_rate": 2e-05,
+      "loss": 0.04586707,
+      "step": 17317
+    },
+    {
+      "epoch": 34.636,
+      "grad_norm": 1.113982915878296,
+      "learning_rate": 2e-05,
+      "loss": 0.04473947,
+      "step": 17318
+    },
+    {
+      "epoch": 34.638,
+      "grad_norm": 1.2651467323303223,
+      "learning_rate": 2e-05,
+      "loss": 0.04464104,
+      "step": 17319
+    },
+    {
+      "epoch": 34.64,
+      "grad_norm": 1.0130887031555176,
+      "learning_rate": 2e-05,
+      "loss": 0.03748289,
+      "step": 17320
+    },
+    {
+      "epoch": 34.642,
+      "grad_norm": 1.977115273475647,
+      "learning_rate": 2e-05,
+      "loss": 0.0529108,
+      "step": 17321
+    },
+    {
+      "epoch": 34.644,
+      "grad_norm": 1.0311118364334106,
+      "learning_rate": 2e-05,
+      "loss": 0.05024704,
+      "step": 17322
+    },
+    {
+      "epoch": 34.646,
+      "grad_norm": 1.7532882690429688,
+      "learning_rate": 2e-05,
+      "loss": 0.05759249,
+      "step": 17323
+    },
+    {
+      "epoch": 34.648,
+      "grad_norm": 2.038602828979492,
+      "learning_rate": 2e-05,
+      "loss": 0.03318835,
+      "step": 17324
+    },
+    {
+      "epoch": 34.65,
+      "grad_norm": 1.232222557067871,
+      "learning_rate": 2e-05,
+      "loss": 0.04337482,
+      "step": 17325
+    },
+    {
+      "epoch": 34.652,
+      "grad_norm": 1.3680845499038696,
+      "learning_rate": 2e-05,
+      "loss": 0.02841615,
+      "step": 17326
+    },
+    {
+      "epoch": 34.654,
+      "grad_norm": 1.8066096305847168,
+      "learning_rate": 2e-05,
+      "loss": 0.03865731,
+      "step": 17327
+    },
+    {
+      "epoch": 34.656,
+      "grad_norm": 1.0195194482803345,
+      "learning_rate": 2e-05,
+      "loss": 0.03887212,
+      "step": 17328
+    },
+    {
+      "epoch": 34.658,
+      "grad_norm": 1.2496850490570068,
+      "learning_rate": 2e-05,
+      "loss": 0.04003082,
+      "step": 17329
+    },
+    {
+      "epoch": 34.66,
+      "grad_norm": 1.1898289918899536,
+      "learning_rate": 2e-05,
+      "loss": 0.03792207,
+      "step": 17330
+    },
+    {
+      "epoch": 34.662,
+      "grad_norm": 2.7598717212677,
+      "learning_rate": 2e-05,
+      "loss": 0.04439123,
+      "step": 17331
+    },
+    {
+      "epoch": 34.664,
+      "grad_norm": 1.5496615171432495,
+      "learning_rate": 2e-05,
+      "loss": 0.05002048,
+      "step": 17332
+    },
+    {
+      "epoch": 34.666,
+      "grad_norm": 1.36111581325531,
+      "learning_rate": 2e-05,
+      "loss": 0.03431807,
+      "step": 17333
+    },
+    {
+      "epoch": 34.668,
+      "grad_norm": 1.1194837093353271,
+      "learning_rate": 2e-05,
+      "loss": 0.04525601,
+      "step": 17334
+    },
+    {
+      "epoch": 34.67,
+      "grad_norm": 2.9787824153900146,
+      "learning_rate": 2e-05,
+      "loss": 0.05164247,
+      "step": 17335
+    },
+    {
+      "epoch": 34.672,
+      "grad_norm": 1.0257699489593506,
+      "learning_rate": 2e-05,
+      "loss": 0.03524974,
+      "step": 17336
+    },
+    {
+      "epoch": 34.674,
+      "grad_norm": 1.438684105873108,
+      "learning_rate": 2e-05,
+      "loss": 0.03349183,
+      "step": 17337
+    },
+    {
+      "epoch": 34.676,
+      "grad_norm": 1.544499397277832,
+      "learning_rate": 2e-05,
+      "loss": 0.04704629,
+      "step": 17338
+    },
+    {
+      "epoch": 34.678,
+      "grad_norm": 1.3472787141799927,
+      "learning_rate": 2e-05,
+      "loss": 0.04309006,
+      "step": 17339
+    },
+    {
+      "epoch": 34.68,
+      "grad_norm": 1.8793655633926392,
+      "learning_rate": 2e-05,
+      "loss": 0.05061891,
+      "step": 17340
+    },
+    {
+      "epoch": 34.682,
+      "grad_norm": 1.2611889839172363,
+      "learning_rate": 2e-05,
+      "loss": 0.05567399,
+      "step": 17341
+    },
+    {
+      "epoch": 34.684,
+      "grad_norm": 1.5432864427566528,
+      "learning_rate": 2e-05,
+      "loss": 0.03108479,
+      "step": 17342
+    },
+    {
+      "epoch": 34.686,
+      "grad_norm": 1.2350441217422485,
+      "learning_rate": 2e-05,
+      "loss": 0.03987078,
+      "step": 17343
+    },
+    {
+      "epoch": 34.688,
+      "grad_norm": 1.0708898305892944,
+      "learning_rate": 2e-05,
+      "loss": 0.03587945,
+      "step": 17344
+    },
+    {
+      "epoch": 34.69,
+      "grad_norm": 1.1190636157989502,
+      "learning_rate": 2e-05,
+      "loss": 0.04188859,
+      "step": 17345
+    },
+    {
+      "epoch": 34.692,
+      "grad_norm": 1.3909366130828857,
+      "learning_rate": 2e-05,
+      "loss": 0.0505821,
+      "step": 17346
+    },
+    {
+      "epoch": 34.694,
+      "grad_norm": 1.1136740446090698,
+      "learning_rate": 2e-05,
+      "loss": 0.04561438,
+      "step": 17347
+    },
+    {
+      "epoch": 34.696,
+      "grad_norm": 2.012559413909912,
+      "learning_rate": 2e-05,
+      "loss": 0.04886289,
+      "step": 17348
+    },
+    {
+      "epoch": 34.698,
+      "grad_norm": 1.2638075351715088,
+      "learning_rate": 2e-05,
+      "loss": 0.04957847,
+      "step": 17349
+    },
+    {
+      "epoch": 34.7,
+      "grad_norm": 1.1630687713623047,
+      "learning_rate": 2e-05,
+      "loss": 0.04921006,
+      "step": 17350
+    },
+    {
+      "epoch": 34.702,
+      "grad_norm": 1.2960374355316162,
+      "learning_rate": 2e-05,
+      "loss": 0.03544647,
+      "step": 17351
+    },
+    {
+      "epoch": 34.704,
+      "grad_norm": 1.0430259704589844,
+      "learning_rate": 2e-05,
+      "loss": 0.04330702,
+      "step": 17352
+    },
+    {
+      "epoch": 34.706,
+      "grad_norm": 1.0777888298034668,
+      "learning_rate": 2e-05,
+      "loss": 0.03540637,
+      "step": 17353
+    },
+    {
+      "epoch": 34.708,
+      "grad_norm": 1.176332712173462,
+      "learning_rate": 2e-05,
+      "loss": 0.03709304,
+      "step": 17354
+    },
+    {
+      "epoch": 34.71,
+      "grad_norm": 1.5132179260253906,
+      "learning_rate": 2e-05,
+      "loss": 0.05928625,
+      "step": 17355
+    },
+    {
+      "epoch": 34.712,
+      "grad_norm": 1.0462723970413208,
+      "learning_rate": 2e-05,
+      "loss": 0.04137703,
+      "step": 17356
+    },
+    {
+      "epoch": 34.714,
+      "grad_norm": 1.0485808849334717,
+      "learning_rate": 2e-05,
+      "loss": 0.04173344,
+      "step": 17357
+    },
+    {
+      "epoch": 34.716,
+      "grad_norm": 0.9149652719497681,
+      "learning_rate": 2e-05,
+      "loss": 0.03566572,
+      "step": 17358
+    },
+    {
+      "epoch": 34.718,
+      "grad_norm": 2.1620941162109375,
+      "learning_rate": 2e-05,
+      "loss": 0.05247457,
+      "step": 17359
+    },
+    {
+      "epoch": 34.72,
+      "grad_norm": 1.1164774894714355,
+      "learning_rate": 2e-05,
+      "loss": 0.04975937,
+      "step": 17360
+    },
+    {
+      "epoch": 34.722,
+      "grad_norm": 1.0158329010009766,
+      "learning_rate": 2e-05,
+      "loss": 0.04216751,
+      "step": 17361
+    },
+    {
+      "epoch": 34.724,
+      "grad_norm": 1.7261289358139038,
+      "learning_rate": 2e-05,
+      "loss": 0.06181233,
+      "step": 17362
+    },
+    {
+      "epoch": 34.726,
+      "grad_norm": 0.7953335046768188,
+      "learning_rate": 2e-05,
+      "loss": 0.02627176,
+      "step": 17363
+    },
+    {
+      "epoch": 34.728,
+      "grad_norm": 1.0371417999267578,
+      "learning_rate": 2e-05,
+      "loss": 0.03395245,
+      "step": 17364
+    },
+    {
+      "epoch": 34.73,
+      "grad_norm": 1.1978789567947388,
+      "learning_rate": 2e-05,
+      "loss": 0.03813355,
+      "step": 17365
+    },
+    {
+      "epoch": 34.732,
+      "grad_norm": 1.0728747844696045,
+      "learning_rate": 2e-05,
+      "loss": 0.0317129,
+      "step": 17366
+    },
+    {
+      "epoch": 34.734,
+      "grad_norm": 1.3377137184143066,
+      "learning_rate": 2e-05,
+      "loss": 0.04749308,
+      "step": 17367
+    },
+    {
+      "epoch": 34.736,
+      "grad_norm": 1.5253896713256836,
+      "learning_rate": 2e-05,
+      "loss": 0.04517586,
+      "step": 17368
+    },
+    {
+      "epoch": 34.738,
+      "grad_norm": 1.125253677368164,
+      "learning_rate": 2e-05,
+      "loss": 0.04429846,
+      "step": 17369
+    },
+    {
+      "epoch": 34.74,
+      "grad_norm": 1.279826283454895,
+      "learning_rate": 2e-05,
+      "loss": 0.04692722,
+      "step": 17370
+    },
+    {
+      "epoch": 34.742,
+      "grad_norm": 1.15207040309906,
+      "learning_rate": 2e-05,
+      "loss": 0.03784764,
+      "step": 17371
+    },
+    {
+      "epoch": 34.744,
+      "grad_norm": 1.4558398723602295,
+      "learning_rate": 2e-05,
+      "loss": 0.05303413,
+      "step": 17372
+    },
+    {
+      "epoch": 34.746,
+      "grad_norm": 2.0725667476654053,
+      "learning_rate": 2e-05,
+      "loss": 0.03780486,
+      "step": 17373
+    },
+    {
+      "epoch": 34.748,
+      "grad_norm": 1.101965069770813,
+      "learning_rate": 2e-05,
+      "loss": 0.0325265,
+      "step": 17374
+    },
+    {
+      "epoch": 34.75,
+      "grad_norm": 1.2319386005401611,
+      "learning_rate": 2e-05,
+      "loss": 0.05028831,
+      "step": 17375
+    },
+    {
+      "epoch": 34.752,
+      "grad_norm": 0.9592969417572021,
+      "learning_rate": 2e-05,
+      "loss": 0.03598889,
+      "step": 17376
+    },
+    {
+      "epoch": 34.754,
+      "grad_norm": 1.2435072660446167,
+      "learning_rate": 2e-05,
+      "loss": 0.04904407,
+      "step": 17377
+    },
+    {
+      "epoch": 34.756,
+      "grad_norm": 1.0948567390441895,
+      "learning_rate": 2e-05,
+      "loss": 0.03987746,
+      "step": 17378
+    },
+    {
+      "epoch": 34.758,
+      "grad_norm": 1.072523593902588,
+      "learning_rate": 2e-05,
+      "loss": 0.04469089,
+      "step": 17379
+    },
+    {
+      "epoch": 34.76,
+      "grad_norm": 4.93184757232666,
+      "learning_rate": 2e-05,
+      "loss": 0.05690203,
+      "step": 17380
+    },
+    {
+      "epoch": 34.762,
+      "grad_norm": 1.764129638671875,
+      "learning_rate": 2e-05,
+      "loss": 0.06511163,
+      "step": 17381
+    },
+    {
+      "epoch": 34.764,
+      "grad_norm": 4.052314758300781,
+      "learning_rate": 2e-05,
+      "loss": 0.05666729,
+      "step": 17382
+    },
+    {
+      "epoch": 34.766,
+      "grad_norm": 1.1972640752792358,
+      "learning_rate": 2e-05,
+      "loss": 0.04556664,
+      "step": 17383
+    },
+    {
+      "epoch": 34.768,
+      "grad_norm": 1.0919245481491089,
+      "learning_rate": 2e-05,
+      "loss": 0.03124575,
+      "step": 17384
+    },
+    {
+      "epoch": 34.77,
+      "grad_norm": 1.072306513786316,
+      "learning_rate": 2e-05,
+      "loss": 0.03128473,
+      "step": 17385
+    },
+    {
+      "epoch": 34.772,
+      "grad_norm": 1.295246958732605,
+      "learning_rate": 2e-05,
+      "loss": 0.05659188,
+      "step": 17386
+    },
+    {
+      "epoch": 34.774,
+      "grad_norm": 0.5764150619506836,
+      "learning_rate": 2e-05,
+      "loss": 0.01222921,
+      "step": 17387
+    },
+    {
+      "epoch": 34.776,
+      "grad_norm": 2.142376661300659,
+      "learning_rate": 2e-05,
+      "loss": 0.05650788,
+      "step": 17388
+    },
+    {
+      "epoch": 34.778,
+      "grad_norm": 1.199610710144043,
+      "learning_rate": 2e-05,
+      "loss": 0.03026199,
+      "step": 17389
+    },
+    {
+      "epoch": 34.78,
+      "grad_norm": 4.657124996185303,
+      "learning_rate": 2e-05,
+      "loss": 0.05358446,
+      "step": 17390
+    },
+    {
+      "epoch": 34.782,
+      "grad_norm": 1.857397198677063,
+      "learning_rate": 2e-05,
+      "loss": 0.04083439,
+      "step": 17391
+    },
+    {
+      "epoch": 34.784,
+      "grad_norm": 1.1742832660675049,
+      "learning_rate": 2e-05,
+      "loss": 0.04639963,
+      "step": 17392
+    },
+    {
+      "epoch": 34.786,
+      "grad_norm": 1.3055615425109863,
+      "learning_rate": 2e-05,
+      "loss": 0.03700022,
+      "step": 17393
+    },
+    {
+      "epoch": 34.788,
+      "grad_norm": 1.1897599697113037,
+      "learning_rate": 2e-05,
+      "loss": 0.03541516,
+      "step": 17394
+    },
+    {
+      "epoch": 34.79,
+      "grad_norm": 0.9236932992935181,
+      "learning_rate": 2e-05,
+      "loss": 0.02803895,
+      "step": 17395
+    },
+    {
+      "epoch": 34.792,
+      "grad_norm": 1.2352205514907837,
+      "learning_rate": 2e-05,
+      "loss": 0.05203598,
+      "step": 17396
+    },
+    {
+      "epoch": 34.794,
+      "grad_norm": 1.137605905532837,
+      "learning_rate": 2e-05,
+      "loss": 0.04387431,
+      "step": 17397
+    },
+    {
+      "epoch": 34.796,
+      "grad_norm": 1.4612951278686523,
+      "learning_rate": 2e-05,
+      "loss": 0.03690141,
+      "step": 17398
+    },
+    {
+      "epoch": 34.798,
+      "grad_norm": 1.1920640468597412,
+      "learning_rate": 2e-05,
+      "loss": 0.03032911,
+      "step": 17399
+    },
+    {
+      "epoch": 34.8,
+      "grad_norm": 1.4039093255996704,
+      "learning_rate": 2e-05,
+      "loss": 0.04082662,
+      "step": 17400
+    },
+    {
+      "epoch": 34.802,
+      "grad_norm": 1.3465620279312134,
+      "learning_rate": 2e-05,
+      "loss": 0.04480909,
+      "step": 17401
+    },
+    {
+      "epoch": 34.804,
+      "grad_norm": 1.015446424484253,
+      "learning_rate": 2e-05,
+      "loss": 0.04655547,
+      "step": 17402
+    },
+    {
+      "epoch": 34.806,
+      "grad_norm": 1.3519433736801147,
+      "learning_rate": 2e-05,
+      "loss": 0.0522088,
+      "step": 17403
+    },
+    {
+      "epoch": 34.808,
+      "grad_norm": 1.9576835632324219,
+      "learning_rate": 2e-05,
+      "loss": 0.04513229,
+      "step": 17404
+    },
+    {
+      "epoch": 34.81,
+      "grad_norm": 1.3243753910064697,
+      "learning_rate": 2e-05,
+      "loss": 0.0440766,
+      "step": 17405
+    },
+    {
+      "epoch": 34.812,
+      "grad_norm": 2.486659526824951,
+      "learning_rate": 2e-05,
+      "loss": 0.04024377,
+      "step": 17406
+    },
+    {
+      "epoch": 34.814,
+      "grad_norm": 1.11258065700531,
+      "learning_rate": 2e-05,
+      "loss": 0.03342846,
+      "step": 17407
+    },
+    {
+      "epoch": 34.816,
+      "grad_norm": 1.34394109249115,
+      "learning_rate": 2e-05,
+      "loss": 0.0359684,
+      "step": 17408
+    },
+    {
+      "epoch": 34.818,
+      "grad_norm": 1.8080177307128906,
+      "learning_rate": 2e-05,
+      "loss": 0.04266899,
+      "step": 17409
+    },
+    {
+      "epoch": 34.82,
+      "grad_norm": 0.9578447937965393,
+      "learning_rate": 2e-05,
+      "loss": 0.02569779,
+      "step": 17410
+    },
+    {
+      "epoch": 34.822,
+      "grad_norm": 1.6380552053451538,
+      "learning_rate": 2e-05,
+      "loss": 0.04872344,
+      "step": 17411
+    },
+    {
+      "epoch": 34.824,
+      "grad_norm": 1.7012923955917358,
+      "learning_rate": 2e-05,
+      "loss": 0.03796242,
+      "step": 17412
+    },
+    {
+      "epoch": 34.826,
+      "grad_norm": 1.7097322940826416,
+      "learning_rate": 2e-05,
+      "loss": 0.04473326,
+      "step": 17413
+    },
+    {
+      "epoch": 34.828,
+      "grad_norm": 0.8996273279190063,
+      "learning_rate": 2e-05,
+      "loss": 0.02824127,
+      "step": 17414
+    },
+    {
+      "epoch": 34.83,
+      "grad_norm": 1.1458615064620972,
+      "learning_rate": 2e-05,
+      "loss": 0.04492954,
+      "step": 17415
+    },
+    {
+      "epoch": 34.832,
+      "grad_norm": 1.754016637802124,
+      "learning_rate": 2e-05,
+      "loss": 0.04475306,
+      "step": 17416
+    },
+    {
+      "epoch": 34.834,
+      "grad_norm": 1.4149738550186157,
+      "learning_rate": 2e-05,
+      "loss": 0.05934685,
+      "step": 17417
+    },
+    {
+      "epoch": 34.836,
+      "grad_norm": 1.1486657857894897,
+      "learning_rate": 2e-05,
+      "loss": 0.03290381,
+      "step": 17418
+    },
+    {
+      "epoch": 34.838,
+      "grad_norm": 1.1294008493423462,
+      "learning_rate": 2e-05,
+      "loss": 0.0412181,
+      "step": 17419
+    },
+    {
+      "epoch": 34.84,
+      "grad_norm": 1.3856641054153442,
+      "learning_rate": 2e-05,
+      "loss": 0.04381644,
+      "step": 17420
+    },
+    {
+      "epoch": 34.842,
+      "grad_norm": 1.5705233812332153,
+      "learning_rate": 2e-05,
+      "loss": 0.0567019,
+      "step": 17421
+    },
+    {
+      "epoch": 34.844,
+      "grad_norm": 2.1528782844543457,
+      "learning_rate": 2e-05,
+      "loss": 0.03446703,
+      "step": 17422
+    },
+    {
+      "epoch": 34.846,
+      "grad_norm": 3.2811214923858643,
+      "learning_rate": 2e-05,
+      "loss": 0.05094865,
+      "step": 17423
+    },
+    {
+      "epoch": 34.848,
+      "grad_norm": 1.390089988708496,
+      "learning_rate": 2e-05,
+      "loss": 0.06304034,
+      "step": 17424
+    },
+    {
+      "epoch": 34.85,
+      "grad_norm": 1.1673579216003418,
+      "learning_rate": 2e-05,
+      "loss": 0.03815352,
+      "step": 17425
+    },
+    {
+      "epoch": 34.852,
+      "grad_norm": 1.155975103378296,
+      "learning_rate": 2e-05,
+      "loss": 0.04023435,
+      "step": 17426
+    },
+    {
+      "epoch": 34.854,
+      "grad_norm": 2.51444673538208,
+      "learning_rate": 2e-05,
+      "loss": 0.03916804,
+      "step": 17427
+    },
+    {
+      "epoch": 34.856,
+      "grad_norm": 4.942361354827881,
+      "learning_rate": 2e-05,
+      "loss": 0.06820999,
+      "step": 17428
+    },
+    {
+      "epoch": 34.858,
+      "grad_norm": 1.0035266876220703,
+      "learning_rate": 2e-05,
+      "loss": 0.03401071,
+      "step": 17429
+    },
+    {
+      "epoch": 34.86,
+      "grad_norm": 1.0705126523971558,
+      "learning_rate": 2e-05,
+      "loss": 0.02997341,
+      "step": 17430
+    },
+    {
+      "epoch": 34.862,
+      "grad_norm": 1.0698009729385376,
+      "learning_rate": 2e-05,
+      "loss": 0.03294614,
+      "step": 17431
+    },
+    {
+      "epoch": 34.864,
+      "grad_norm": 1.062578558921814,
+      "learning_rate": 2e-05,
+      "loss": 0.04528739,
+      "step": 17432
+    },
+    {
+      "epoch": 34.866,
+      "grad_norm": 1.1648660898208618,
+      "learning_rate": 2e-05,
+      "loss": 0.04851899,
+      "step": 17433
+    },
+    {
+      "epoch": 34.868,
+      "grad_norm": 1.1196273565292358,
+      "learning_rate": 2e-05,
+      "loss": 0.04238444,
+      "step": 17434
+    },
+    {
+      "epoch": 34.87,
+      "grad_norm": 1.043180227279663,
+      "learning_rate": 2e-05,
+      "loss": 0.04557565,
+      "step": 17435
+    },
+    {
+      "epoch": 34.872,
+      "grad_norm": 1.079825520515442,
+      "learning_rate": 2e-05,
+      "loss": 0.04296175,
+      "step": 17436
+    },
+    {
+      "epoch": 34.874,
+      "grad_norm": 1.0005967617034912,
+      "learning_rate": 2e-05,
+      "loss": 0.02957804,
+      "step": 17437
+    },
+    {
+      "epoch": 34.876,
+      "grad_norm": 1.1389814615249634,
+      "learning_rate": 2e-05,
+      "loss": 0.0373849,
+      "step": 17438
+    },
+    {
+      "epoch": 34.878,
+      "grad_norm": 1.3328295946121216,
+      "learning_rate": 2e-05,
+      "loss": 0.03476042,
+      "step": 17439
+    },
+    {
+      "epoch": 34.88,
+      "grad_norm": 1.4265692234039307,
+      "learning_rate": 2e-05,
+      "loss": 0.05438258,
+      "step": 17440
+    },
+    {
+      "epoch": 34.882,
+      "grad_norm": 1.3445887565612793,
+      "learning_rate": 2e-05,
+      "loss": 0.04738908,
+      "step": 17441
+    },
+    {
+      "epoch": 34.884,
+      "grad_norm": 1.100516676902771,
+      "learning_rate": 2e-05,
+      "loss": 0.03541394,
+      "step": 17442
+    },
+    {
+      "epoch": 34.886,
+      "grad_norm": 1.2599114179611206,
+      "learning_rate": 2e-05,
+      "loss": 0.05563427,
+      "step": 17443
+    },
+    {
+      "epoch": 34.888,
+      "grad_norm": 1.3111646175384521,
+      "learning_rate": 2e-05,
+      "loss": 0.04711374,
+      "step": 17444
+    },
+    {
+      "epoch": 34.89,
+      "grad_norm": 1.0831414461135864,
+      "learning_rate": 2e-05,
+      "loss": 0.03449087,
+      "step": 17445
+    },
+    {
+      "epoch": 34.892,
+      "grad_norm": 1.390146255493164,
+      "learning_rate": 2e-05,
+      "loss": 0.04913954,
+      "step": 17446
+    },
+    {
+      "epoch": 34.894,
+      "grad_norm": 1.4582270383834839,
+      "learning_rate": 2e-05,
+      "loss": 0.04977434,
+      "step": 17447
+    },
+    {
+      "epoch": 34.896,
+      "grad_norm": 1.040456771850586,
+      "learning_rate": 2e-05,
+      "loss": 0.04064398,
+      "step": 17448
+    },
+    {
+      "epoch": 34.898,
+      "grad_norm": 1.125290036201477,
+      "learning_rate": 2e-05,
+      "loss": 0.0368836,
+      "step": 17449
+    },
+    {
+      "epoch": 34.9,
+      "grad_norm": 1.6109248399734497,
+      "learning_rate": 2e-05,
+      "loss": 0.05745726,
+      "step": 17450
+    },
+    {
+      "epoch": 34.902,
+      "grad_norm": 1.6481939554214478,
+      "learning_rate": 2e-05,
+      "loss": 0.05567736,
+      "step": 17451
+    },
+    {
+      "epoch": 34.904,
+      "grad_norm": 1.1878528594970703,
+      "learning_rate": 2e-05,
+      "loss": 0.03877228,
+      "step": 17452
+    },
+    {
+      "epoch": 34.906,
+      "grad_norm": 1.0018279552459717,
+      "learning_rate": 2e-05,
+      "loss": 0.03138274,
+      "step": 17453
+    },
+    {
+      "epoch": 34.908,
+      "grad_norm": 1.4977102279663086,
+      "learning_rate": 2e-05,
+      "loss": 0.03562863,
+      "step": 17454
+    },
+    {
+      "epoch": 34.91,
+      "grad_norm": 1.0117096900939941,
+      "learning_rate": 2e-05,
+      "loss": 0.03509045,
+      "step": 17455
+    },
+    {
+      "epoch": 34.912,
+      "grad_norm": 1.3096954822540283,
+      "learning_rate": 2e-05,
+      "loss": 0.04645215,
+      "step": 17456
+    },
+    {
+      "epoch": 34.914,
+      "grad_norm": 1.7684214115142822,
+      "learning_rate": 2e-05,
+      "loss": 0.04929952,
+      "step": 17457
+    },
+    {
+      "epoch": 34.916,
+      "grad_norm": 1.0269994735717773,
+      "learning_rate": 2e-05,
+      "loss": 0.03455571,
+      "step": 17458
+    },
+    {
+      "epoch": 34.918,
+      "grad_norm": 0.9055759310722351,
+      "learning_rate": 2e-05,
+      "loss": 0.03060112,
+      "step": 17459
+    },
+    {
+      "epoch": 34.92,
+      "grad_norm": 1.3847931623458862,
+      "learning_rate": 2e-05,
+      "loss": 0.05905629,
+      "step": 17460
+    },
+    {
+      "epoch": 34.922,
+      "grad_norm": 0.8554072380065918,
+      "learning_rate": 2e-05,
+      "loss": 0.02603616,
+      "step": 17461
+    },
+    {
+      "epoch": 34.924,
+      "grad_norm": 1.3800654411315918,
+      "learning_rate": 2e-05,
+      "loss": 0.04450605,
+      "step": 17462
+    },
+    {
+      "epoch": 34.926,
+      "grad_norm": 1.0499212741851807,
+      "learning_rate": 2e-05,
+      "loss": 0.03139212,
+      "step": 17463
+    },
+    {
+      "epoch": 34.928,
+      "grad_norm": 1.5442363023757935,
+      "learning_rate": 2e-05,
+      "loss": 0.05026272,
+      "step": 17464
+    },
+    {
+      "epoch": 34.93,
+      "grad_norm": 1.0467393398284912,
+      "learning_rate": 2e-05,
+      "loss": 0.03508221,
+      "step": 17465
+    },
+    {
+      "epoch": 34.932,
+      "grad_norm": 1.9836903810501099,
+      "learning_rate": 2e-05,
+      "loss": 0.03904892,
+      "step": 17466
+    },
+    {
+      "epoch": 34.934,
+      "grad_norm": 1.667801022529602,
+      "learning_rate": 2e-05,
+      "loss": 0.06595255,
+      "step": 17467
+    },
+    {
+      "epoch": 34.936,
+      "grad_norm": 1.248084545135498,
+      "learning_rate": 2e-05,
+      "loss": 0.04658974,
+      "step": 17468
+    },
+    {
+      "epoch": 34.938,
+      "grad_norm": 1.1718586683273315,
+      "learning_rate": 2e-05,
+      "loss": 0.04607401,
+      "step": 17469
+    },
+    {
+      "epoch": 34.94,
+      "grad_norm": 1.562265396118164,
+      "learning_rate": 2e-05,
+      "loss": 0.05189687,
+      "step": 17470
+    },
+    {
+      "epoch": 34.942,
+      "grad_norm": 2.5033798217773438,
+      "learning_rate": 2e-05,
+      "loss": 0.04024173,
+      "step": 17471
+    },
+    {
+      "epoch": 34.944,
+      "grad_norm": 1.3833789825439453,
+      "learning_rate": 2e-05,
+      "loss": 0.04375815,
+      "step": 17472
+    },
+    {
+      "epoch": 34.946,
+      "grad_norm": 1.2791060209274292,
+      "learning_rate": 2e-05,
+      "loss": 0.03404561,
+      "step": 17473
+    },
+    {
+      "epoch": 34.948,
+      "grad_norm": 4.338647365570068,
+      "learning_rate": 2e-05,
+      "loss": 0.05873897,
+      "step": 17474
+    },
+    {
+      "epoch": 34.95,
+      "grad_norm": 1.4956046342849731,
+      "learning_rate": 2e-05,
+      "loss": 0.04620642,
+      "step": 17475
+    },
+    {
+      "epoch": 34.952,
+      "grad_norm": 1.1326979398727417,
+      "learning_rate": 2e-05,
+      "loss": 0.04122002,
+      "step": 17476
+    },
+    {
+      "epoch": 34.954,
+      "grad_norm": 1.0940099954605103,
+      "learning_rate": 2e-05,
+      "loss": 0.04148469,
+      "step": 17477
+    },
+    {
+      "epoch": 34.956,
+      "grad_norm": 1.1113457679748535,
+      "learning_rate": 2e-05,
+      "loss": 0.04638319,
+      "step": 17478
+    },
+    {
+      "epoch": 34.958,
+      "grad_norm": 1.088361144065857,
+      "learning_rate": 2e-05,
+      "loss": 0.04051323,
+      "step": 17479
+    },
+    {
+      "epoch": 34.96,
+      "grad_norm": 1.1193358898162842,
+      "learning_rate": 2e-05,
+      "loss": 0.04485134,
+      "step": 17480
+    },
+    {
+      "epoch": 34.962,
+      "grad_norm": 1.2051522731781006,
+      "learning_rate": 2e-05,
+      "loss": 0.03799698,
+      "step": 17481
+    },
+    {
+      "epoch": 34.964,
+      "grad_norm": 1.1342461109161377,
+      "learning_rate": 2e-05,
+      "loss": 0.03318293,
+      "step": 17482
+    },
+    {
+      "epoch": 34.966,
+      "grad_norm": 1.2009114027023315,
+      "learning_rate": 2e-05,
+      "loss": 0.04339875,
+      "step": 17483
+    },
+    {
+      "epoch": 34.968,
+      "grad_norm": 1.1767265796661377,
+      "learning_rate": 2e-05,
+      "loss": 0.04513696,
+      "step": 17484
+    },
+    {
+      "epoch": 34.97,
+      "grad_norm": 1.503244400024414,
+      "learning_rate": 2e-05,
+      "loss": 0.07590361,
+      "step": 17485
+    },
+    {
+      "epoch": 34.972,
+      "grad_norm": 0.9458696246147156,
+      "learning_rate": 2e-05,
+      "loss": 0.03492884,
+      "step": 17486
+    },
+    {
+      "epoch": 34.974,
+      "grad_norm": 1.2209609746932983,
+      "learning_rate": 2e-05,
+      "loss": 0.02998404,
+      "step": 17487
+    },
+    {
+      "epoch": 34.976,
+      "grad_norm": 1.2666972875595093,
+      "learning_rate": 2e-05,
+      "loss": 0.05165854,
+      "step": 17488
+    },
+    {
+      "epoch": 34.978,
+      "grad_norm": 1.2078595161437988,
+      "learning_rate": 2e-05,
+      "loss": 0.04767955,
+      "step": 17489
+    },
+    {
+      "epoch": 34.98,
+      "grad_norm": 1.205334186553955,
+      "learning_rate": 2e-05,
+      "loss": 0.05208175,
+      "step": 17490
+    },
+    {
+      "epoch": 34.982,
+      "grad_norm": 1.0179728269577026,
+      "learning_rate": 2e-05,
+      "loss": 0.03114939,
+      "step": 17491
+    },
+    {
+      "epoch": 34.984,
+      "grad_norm": 1.8329960107803345,
+      "learning_rate": 2e-05,
+      "loss": 0.06110509,
+      "step": 17492
+    },
+    {
+      "epoch": 34.986,
+      "grad_norm": 1.045487880706787,
+      "learning_rate": 2e-05,
+      "loss": 0.02562892,
+      "step": 17493
+    },
+    {
+      "epoch": 34.988,
+      "grad_norm": 1.160578966140747,
+      "learning_rate": 2e-05,
+      "loss": 0.04529508,
+      "step": 17494
+    },
+    {
+      "epoch": 34.99,
+      "grad_norm": 1.1257965564727783,
+      "learning_rate": 2e-05,
+      "loss": 0.04581528,
+      "step": 17495
+    },
+    {
+      "epoch": 34.992,
+      "grad_norm": 1.1430251598358154,
+      "learning_rate": 2e-05,
+      "loss": 0.05036756,
+      "step": 17496
+    },
+    {
+      "epoch": 34.994,
+      "grad_norm": 1.6295039653778076,
+      "learning_rate": 2e-05,
+      "loss": 0.04688197,
+      "step": 17497
+    },
+    {
+      "epoch": 34.996,
+      "grad_norm": 1.4103097915649414,
+      "learning_rate": 2e-05,
+      "loss": 0.05346009,
+      "step": 17498
+    },
+    {
+      "epoch": 34.998,
+      "grad_norm": 1.2611072063446045,
+      "learning_rate": 2e-05,
+      "loss": 0.0430942,
+      "step": 17499
+    },
+    {
+      "epoch": 35.0,
+      "grad_norm": 2.2868216037750244,
+      "learning_rate": 2e-05,
+      "loss": 0.04770452,
+      "step": 17500
+    },
+    {
+      "epoch": 35.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.998,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9820359281437125,
+        "Equal_1": 1.0,
+        "Equal_2": 0.9780439121756487,
+        "Equal_3": 0.9880239520958084,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.998003992015968,
+        "Parallel_1": 0.9939879759519038,
+        "Parallel_2": 0.9959919839679359,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.996,
+        "Perpendicular_3": 0.9038076152304609,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 0.996,
+        "PointLiesOnCircle_3": 0.988,
+        "PointLiesOnLine_1": 0.9959919839679359,
+        "PointLiesOnLine_2": 0.9919839679358717,
+        "PointLiesOnLine_3": 0.9900199600798403
+      },
+      "eval_runtime": 319.8873,
+      "eval_samples_per_second": 32.824,
+      "eval_steps_per_second": 0.656,
+      "step": 17500
+    },
+    {
+      "epoch": 35.002,
+      "grad_norm": 1.9222464561462402,
+      "learning_rate": 2e-05,
+      "loss": 0.03432529,
+      "step": 17501
+    },
+    {
+      "epoch": 35.004,
+      "grad_norm": 1.4019160270690918,
+      "learning_rate": 2e-05,
+      "loss": 0.03604925,
+      "step": 17502
+    },
+    {
+      "epoch": 35.006,
+      "grad_norm": 0.9827688336372375,
+      "learning_rate": 2e-05,
+      "loss": 0.02998075,
+      "step": 17503
+    },
+    {
+      "epoch": 35.008,
+      "grad_norm": 1.1721317768096924,
+      "learning_rate": 2e-05,
+      "loss": 0.04828912,
+      "step": 17504
+    },
+    {
+      "epoch": 35.01,
+      "grad_norm": 1.120781421661377,
+      "learning_rate": 2e-05,
+      "loss": 0.04639394,
+      "step": 17505
+    },
+    {
+      "epoch": 35.012,
+      "grad_norm": 1.3322588205337524,
+      "learning_rate": 2e-05,
+      "loss": 0.03849679,
+      "step": 17506
+    },
+    {
+      "epoch": 35.014,
+      "grad_norm": 1.1528602838516235,
+      "learning_rate": 2e-05,
+      "loss": 0.04227344,
+      "step": 17507
+    },
+    {
+      "epoch": 35.016,
+      "grad_norm": 1.0254783630371094,
+      "learning_rate": 2e-05,
+      "loss": 0.02713573,
+      "step": 17508
+    },
+    {
+      "epoch": 35.018,
+      "grad_norm": 2.194639205932617,
+      "learning_rate": 2e-05,
+      "loss": 0.06066288,
+      "step": 17509
+    },
+    {
+      "epoch": 35.02,
+      "grad_norm": 1.1950559616088867,
+      "learning_rate": 2e-05,
+      "loss": 0.05654139,
+      "step": 17510
+    },
+    {
+      "epoch": 35.022,
+      "grad_norm": 1.4963222742080688,
+      "learning_rate": 2e-05,
+      "loss": 0.03999151,
+      "step": 17511
+    },
+    {
+      "epoch": 35.024,
+      "grad_norm": 1.4710837602615356,
+      "learning_rate": 2e-05,
+      "loss": 0.05531586,
+      "step": 17512
+    },
+    {
+      "epoch": 35.026,
+      "grad_norm": 1.7270408868789673,
+      "learning_rate": 2e-05,
+      "loss": 0.04574881,
+      "step": 17513
+    },
+    {
+      "epoch": 35.028,
+      "grad_norm": 1.1254507303237915,
+      "learning_rate": 2e-05,
+      "loss": 0.04714286,
+      "step": 17514
+    },
+    {
+      "epoch": 35.03,
+      "grad_norm": 1.2321765422821045,
+      "learning_rate": 2e-05,
+      "loss": 0.0381449,
+      "step": 17515
+    },
+    {
+      "epoch": 35.032,
+      "grad_norm": 1.3832772970199585,
+      "learning_rate": 2e-05,
+      "loss": 0.04403721,
+      "step": 17516
+    },
+    {
+      "epoch": 35.034,
+      "grad_norm": 2.5237765312194824,
+      "learning_rate": 2e-05,
+      "loss": 0.0705086,
+      "step": 17517
+    },
+    {
+      "epoch": 35.036,
+      "grad_norm": 1.3510403633117676,
+      "learning_rate": 2e-05,
+      "loss": 0.04805104,
+      "step": 17518
+    },
+    {
+      "epoch": 35.038,
+      "grad_norm": 1.1089110374450684,
+      "learning_rate": 2e-05,
+      "loss": 0.04748667,
+      "step": 17519
+    },
+    {
+      "epoch": 35.04,
+      "grad_norm": 1.0043740272521973,
+      "learning_rate": 2e-05,
+      "loss": 0.03443906,
+      "step": 17520
+    },
+    {
+      "epoch": 35.042,
+      "grad_norm": 1.2505918741226196,
+      "learning_rate": 2e-05,
+      "loss": 0.0434991,
+      "step": 17521
+    },
+    {
+      "epoch": 35.044,
+      "grad_norm": 1.0381025075912476,
+      "learning_rate": 2e-05,
+      "loss": 0.04269955,
+      "step": 17522
+    },
+    {
+      "epoch": 35.046,
+      "grad_norm": 1.2612941265106201,
+      "learning_rate": 2e-05,
+      "loss": 0.04119605,
+      "step": 17523
+    },
+    {
+      "epoch": 35.048,
+      "grad_norm": 2.3623406887054443,
+      "learning_rate": 2e-05,
+      "loss": 0.03740188,
+      "step": 17524
+    },
+    {
+      "epoch": 35.05,
+      "grad_norm": 1.2907884120941162,
+      "learning_rate": 2e-05,
+      "loss": 0.04235482,
+      "step": 17525
+    },
+    {
+      "epoch": 35.052,
+      "grad_norm": 0.83847975730896,
+      "learning_rate": 2e-05,
+      "loss": 0.02778682,
+      "step": 17526
+    },
+    {
+      "epoch": 35.054,
+      "grad_norm": 1.2352306842803955,
+      "learning_rate": 2e-05,
+      "loss": 0.04302741,
+      "step": 17527
+    },
+    {
+      "epoch": 35.056,
+      "grad_norm": 1.0656414031982422,
+      "learning_rate": 2e-05,
+      "loss": 0.04405284,
+      "step": 17528
+    },
+    {
+      "epoch": 35.058,
+      "grad_norm": 2.7260079383850098,
+      "learning_rate": 2e-05,
+      "loss": 0.04868072,
+      "step": 17529
+    },
+    {
+      "epoch": 35.06,
+      "grad_norm": 0.9766820073127747,
+      "learning_rate": 2e-05,
+      "loss": 0.03807481,
+      "step": 17530
+    },
+    {
+      "epoch": 35.062,
+      "grad_norm": 0.8948190808296204,
+      "learning_rate": 2e-05,
+      "loss": 0.02815506,
+      "step": 17531
+    },
+    {
+      "epoch": 35.064,
+      "grad_norm": 2.2773566246032715,
+      "learning_rate": 2e-05,
+      "loss": 0.05238456,
+      "step": 17532
+    },
+    {
+      "epoch": 35.066,
+      "grad_norm": 1.1301003694534302,
+      "learning_rate": 2e-05,
+      "loss": 0.04259311,
+      "step": 17533
+    },
+    {
+      "epoch": 35.068,
+      "grad_norm": 1.2741830348968506,
+      "learning_rate": 2e-05,
+      "loss": 0.06291425,
+      "step": 17534
+    },
+    {
+      "epoch": 35.07,
+      "grad_norm": 1.1298288106918335,
+      "learning_rate": 2e-05,
+      "loss": 0.04579066,
+      "step": 17535
+    },
+    {
+      "epoch": 35.072,
+      "grad_norm": 1.111412525177002,
+      "learning_rate": 2e-05,
+      "loss": 0.02930526,
+      "step": 17536
+    },
+    {
+      "epoch": 35.074,
+      "grad_norm": 1.3507663011550903,
+      "learning_rate": 2e-05,
+      "loss": 0.06096983,
+      "step": 17537
+    },
+    {
+      "epoch": 35.076,
+      "grad_norm": 1.3331928253173828,
+      "learning_rate": 2e-05,
+      "loss": 0.06763585,
+      "step": 17538
+    },
+    {
+      "epoch": 35.078,
+      "grad_norm": 1.7155866622924805,
+      "learning_rate": 2e-05,
+      "loss": 0.04312644,
+      "step": 17539
+    },
+    {
+      "epoch": 35.08,
+      "grad_norm": 1.359623908996582,
+      "learning_rate": 2e-05,
+      "loss": 0.04029921,
+      "step": 17540
+    },
+    {
+      "epoch": 35.082,
+      "grad_norm": 1.0476104021072388,
+      "learning_rate": 2e-05,
+      "loss": 0.04592532,
+      "step": 17541
+    },
+    {
+      "epoch": 35.084,
+      "grad_norm": 2.1303536891937256,
+      "learning_rate": 2e-05,
+      "loss": 0.03890026,
+      "step": 17542
+    },
+    {
+      "epoch": 35.086,
+      "grad_norm": 1.1645375490188599,
+      "learning_rate": 2e-05,
+      "loss": 0.04217415,
+      "step": 17543
+    },
+    {
+      "epoch": 35.088,
+      "grad_norm": 1.1522724628448486,
+      "learning_rate": 2e-05,
+      "loss": 0.05060149,
+      "step": 17544
+    },
+    {
+      "epoch": 35.09,
+      "grad_norm": 0.8973413705825806,
+      "learning_rate": 2e-05,
+      "loss": 0.02097614,
+      "step": 17545
+    },
+    {
+      "epoch": 35.092,
+      "grad_norm": 1.3803879022598267,
+      "learning_rate": 2e-05,
+      "loss": 0.04214935,
+      "step": 17546
+    },
+    {
+      "epoch": 35.094,
+      "grad_norm": 1.174539566040039,
+      "learning_rate": 2e-05,
+      "loss": 0.04683707,
+      "step": 17547
+    },
+    {
+      "epoch": 35.096,
+      "grad_norm": 1.838232398033142,
+      "learning_rate": 2e-05,
+      "loss": 0.04326537,
+      "step": 17548
+    },
+    {
+      "epoch": 35.098,
+      "grad_norm": 1.5689642429351807,
+      "learning_rate": 2e-05,
+      "loss": 0.04072879,
+      "step": 17549
+    },
+    {
+      "epoch": 35.1,
+      "grad_norm": 1.0995928049087524,
+      "learning_rate": 2e-05,
+      "loss": 0.03880709,
+      "step": 17550
+    },
+    {
+      "epoch": 35.102,
+      "grad_norm": 0.8378002047538757,
+      "learning_rate": 2e-05,
+      "loss": 0.02281972,
+      "step": 17551
+    },
+    {
+      "epoch": 35.104,
+      "grad_norm": 1.1497136354446411,
+      "learning_rate": 2e-05,
+      "loss": 0.04842537,
+      "step": 17552
+    },
+    {
+      "epoch": 35.106,
+      "grad_norm": 1.2610912322998047,
+      "learning_rate": 2e-05,
+      "loss": 0.04262506,
+      "step": 17553
+    },
+    {
+      "epoch": 35.108,
+      "grad_norm": 0.969350278377533,
+      "learning_rate": 2e-05,
+      "loss": 0.03585976,
+      "step": 17554
+    },
+    {
+      "epoch": 35.11,
+      "grad_norm": 1.0584502220153809,
+      "learning_rate": 2e-05,
+      "loss": 0.03664526,
+      "step": 17555
+    },
+    {
+      "epoch": 35.112,
+      "grad_norm": 1.3232938051223755,
+      "learning_rate": 2e-05,
+      "loss": 0.05472728,
+      "step": 17556
+    },
+    {
+      "epoch": 35.114,
+      "grad_norm": 1.1075985431671143,
+      "learning_rate": 2e-05,
+      "loss": 0.05077582,
+      "step": 17557
+    },
+    {
+      "epoch": 35.116,
+      "grad_norm": 1.2517019510269165,
+      "learning_rate": 2e-05,
+      "loss": 0.05858276,
+      "step": 17558
+    },
+    {
+      "epoch": 35.118,
+      "grad_norm": 1.2317686080932617,
+      "learning_rate": 2e-05,
+      "loss": 0.04953536,
+      "step": 17559
+    },
+    {
+      "epoch": 35.12,
+      "grad_norm": 2.1341912746429443,
+      "learning_rate": 2e-05,
+      "loss": 0.03007166,
+      "step": 17560
+    },
+    {
+      "epoch": 35.122,
+      "grad_norm": 2.465643882751465,
+      "learning_rate": 2e-05,
+      "loss": 0.03830947,
+      "step": 17561
+    },
+    {
+      "epoch": 35.124,
+      "grad_norm": 1.5876415967941284,
+      "learning_rate": 2e-05,
+      "loss": 0.03763131,
+      "step": 17562
+    },
+    {
+      "epoch": 35.126,
+      "grad_norm": 1.1403216123580933,
+      "learning_rate": 2e-05,
+      "loss": 0.04078696,
+      "step": 17563
+    },
+    {
+      "epoch": 35.128,
+      "grad_norm": 1.7711175680160522,
+      "learning_rate": 2e-05,
+      "loss": 0.05040234,
+      "step": 17564
+    },
+    {
+      "epoch": 35.13,
+      "grad_norm": 1.334191918373108,
+      "learning_rate": 2e-05,
+      "loss": 0.04977368,
+      "step": 17565
+    },
+    {
+      "epoch": 35.132,
+      "grad_norm": 1.9622347354888916,
+      "learning_rate": 2e-05,
+      "loss": 0.04543035,
+      "step": 17566
+    },
+    {
+      "epoch": 35.134,
+      "grad_norm": 1.4173260927200317,
+      "learning_rate": 2e-05,
+      "loss": 0.05295032,
+      "step": 17567
+    },
+    {
+      "epoch": 35.136,
+      "grad_norm": 1.4834239482879639,
+      "learning_rate": 2e-05,
+      "loss": 0.04917359,
+      "step": 17568
+    },
+    {
+      "epoch": 35.138,
+      "grad_norm": 1.31399405002594,
+      "learning_rate": 2e-05,
+      "loss": 0.03308884,
+      "step": 17569
+    },
+    {
+      "epoch": 35.14,
+      "grad_norm": 1.307672142982483,
+      "learning_rate": 2e-05,
+      "loss": 0.04067755,
+      "step": 17570
+    },
+    {
+      "epoch": 35.142,
+      "grad_norm": 1.3411407470703125,
+      "learning_rate": 2e-05,
+      "loss": 0.041333,
+      "step": 17571
+    },
+    {
+      "epoch": 35.144,
+      "grad_norm": 5.57957649230957,
+      "learning_rate": 2e-05,
+      "loss": 0.04797975,
+      "step": 17572
+    },
+    {
+      "epoch": 35.146,
+      "grad_norm": 1.3772554397583008,
+      "learning_rate": 2e-05,
+      "loss": 0.05607633,
+      "step": 17573
+    },
+    {
+      "epoch": 35.148,
+      "grad_norm": 2.2462568283081055,
+      "learning_rate": 2e-05,
+      "loss": 0.06492513,
+      "step": 17574
+    },
+    {
+      "epoch": 35.15,
+      "grad_norm": 1.0662626028060913,
+      "learning_rate": 2e-05,
+      "loss": 0.05135501,
+      "step": 17575
+    },
+    {
+      "epoch": 35.152,
+      "grad_norm": 1.4305570125579834,
+      "learning_rate": 2e-05,
+      "loss": 0.04058772,
+      "step": 17576
+    },
+    {
+      "epoch": 35.154,
+      "grad_norm": 1.031005620956421,
+      "learning_rate": 2e-05,
+      "loss": 0.0381646,
+      "step": 17577
+    },
+    {
+      "epoch": 35.156,
+      "grad_norm": 1.168686866760254,
+      "learning_rate": 2e-05,
+      "loss": 0.04385469,
+      "step": 17578
+    },
+    {
+      "epoch": 35.158,
+      "grad_norm": 0.9006215333938599,
+      "learning_rate": 2e-05,
+      "loss": 0.03294571,
+      "step": 17579
+    },
+    {
+      "epoch": 35.16,
+      "grad_norm": 1.1258612871170044,
+      "learning_rate": 2e-05,
+      "loss": 0.05022587,
+      "step": 17580
+    },
+    {
+      "epoch": 35.162,
+      "grad_norm": 1.2660801410675049,
+      "learning_rate": 2e-05,
+      "loss": 0.06481211,
+      "step": 17581
+    },
+    {
+      "epoch": 35.164,
+      "grad_norm": 1.8085393905639648,
+      "learning_rate": 2e-05,
+      "loss": 0.03045495,
+      "step": 17582
+    },
+    {
+      "epoch": 35.166,
+      "grad_norm": 1.5225801467895508,
+      "learning_rate": 2e-05,
+      "loss": 0.04185986,
+      "step": 17583
+    },
+    {
+      "epoch": 35.168,
+      "grad_norm": 1.3285081386566162,
+      "learning_rate": 2e-05,
+      "loss": 0.03883301,
+      "step": 17584
+    },
+    {
+      "epoch": 35.17,
+      "grad_norm": 0.9945167899131775,
+      "learning_rate": 2e-05,
+      "loss": 0.03533852,
+      "step": 17585
+    },
+    {
+      "epoch": 35.172,
+      "grad_norm": 1.0777450799942017,
+      "learning_rate": 2e-05,
+      "loss": 0.03911363,
+      "step": 17586
+    },
+    {
+      "epoch": 35.174,
+      "grad_norm": 0.8671603202819824,
+      "learning_rate": 2e-05,
+      "loss": 0.02011818,
+      "step": 17587
+    },
+    {
+      "epoch": 35.176,
+      "grad_norm": 1.2364228963851929,
+      "learning_rate": 2e-05,
+      "loss": 0.04776933,
+      "step": 17588
+    },
+    {
+      "epoch": 35.178,
+      "grad_norm": 1.3007594347000122,
+      "learning_rate": 2e-05,
+      "loss": 0.04221889,
+      "step": 17589
+    },
+    {
+      "epoch": 35.18,
+      "grad_norm": 1.1037037372589111,
+      "learning_rate": 2e-05,
+      "loss": 0.03446237,
+      "step": 17590
+    },
+    {
+      "epoch": 35.182,
+      "grad_norm": 1.2064841985702515,
+      "learning_rate": 2e-05,
+      "loss": 0.04089108,
+      "step": 17591
+    },
+    {
+      "epoch": 35.184,
+      "grad_norm": 0.9847924113273621,
+      "learning_rate": 2e-05,
+      "loss": 0.0352126,
+      "step": 17592
+    },
+    {
+      "epoch": 35.186,
+      "grad_norm": 1.4962389469146729,
+      "learning_rate": 2e-05,
+      "loss": 0.06021068,
+      "step": 17593
+    },
+    {
+      "epoch": 35.188,
+      "grad_norm": 1.4527535438537598,
+      "learning_rate": 2e-05,
+      "loss": 0.0581253,
+      "step": 17594
+    },
+    {
+      "epoch": 35.19,
+      "grad_norm": 1.0838546752929688,
+      "learning_rate": 2e-05,
+      "loss": 0.03708592,
+      "step": 17595
+    },
+    {
+      "epoch": 35.192,
+      "grad_norm": 1.216435432434082,
+      "learning_rate": 2e-05,
+      "loss": 0.0478345,
+      "step": 17596
+    },
+    {
+      "epoch": 35.194,
+      "grad_norm": 3.266355514526367,
+      "learning_rate": 2e-05,
+      "loss": 0.04139251,
+      "step": 17597
+    },
+    {
+      "epoch": 35.196,
+      "grad_norm": 2.6940810680389404,
+      "learning_rate": 2e-05,
+      "loss": 0.04337883,
+      "step": 17598
+    },
+    {
+      "epoch": 35.198,
+      "grad_norm": 1.02515709400177,
+      "learning_rate": 2e-05,
+      "loss": 0.0304161,
+      "step": 17599
+    },
+    {
+      "epoch": 35.2,
+      "grad_norm": 1.0762580633163452,
+      "learning_rate": 2e-05,
+      "loss": 0.04152739,
+      "step": 17600
+    },
+    {
+      "epoch": 35.202,
+      "grad_norm": 2.462306499481201,
+      "learning_rate": 2e-05,
+      "loss": 0.03642452,
+      "step": 17601
+    },
+    {
+      "epoch": 35.204,
+      "grad_norm": 1.2067070007324219,
+      "learning_rate": 2e-05,
+      "loss": 0.04705513,
+      "step": 17602
+    },
+    {
+      "epoch": 35.206,
+      "grad_norm": 1.4548958539962769,
+      "learning_rate": 2e-05,
+      "loss": 0.06413062,
+      "step": 17603
+    },
+    {
+      "epoch": 35.208,
+      "grad_norm": 1.3760452270507812,
+      "learning_rate": 2e-05,
+      "loss": 0.05110206,
+      "step": 17604
+    },
+    {
+      "epoch": 35.21,
+      "grad_norm": 1.0767890214920044,
+      "learning_rate": 2e-05,
+      "loss": 0.03728911,
+      "step": 17605
+    },
+    {
+      "epoch": 35.212,
+      "grad_norm": 1.1798242330551147,
+      "learning_rate": 2e-05,
+      "loss": 0.04494874,
+      "step": 17606
+    },
+    {
+      "epoch": 35.214,
+      "grad_norm": 1.1159026622772217,
+      "learning_rate": 2e-05,
+      "loss": 0.04022665,
+      "step": 17607
+    },
+    {
+      "epoch": 35.216,
+      "grad_norm": 1.239780306816101,
+      "learning_rate": 2e-05,
+      "loss": 0.04985084,
+      "step": 17608
+    },
+    {
+      "epoch": 35.218,
+      "grad_norm": 1.381771445274353,
+      "learning_rate": 2e-05,
+      "loss": 0.02961168,
+      "step": 17609
+    },
+    {
+      "epoch": 35.22,
+      "grad_norm": 1.4819316864013672,
+      "learning_rate": 2e-05,
+      "loss": 0.04220682,
+      "step": 17610
+    },
+    {
+      "epoch": 35.222,
+      "grad_norm": 1.0734423398971558,
+      "learning_rate": 2e-05,
+      "loss": 0.03761606,
+      "step": 17611
+    },
+    {
+      "epoch": 35.224,
+      "grad_norm": 1.9286874532699585,
+      "learning_rate": 2e-05,
+      "loss": 0.04495767,
+      "step": 17612
+    },
+    {
+      "epoch": 35.226,
+      "grad_norm": 1.3085813522338867,
+      "learning_rate": 2e-05,
+      "loss": 0.04169934,
+      "step": 17613
+    },
+    {
+      "epoch": 35.228,
+      "grad_norm": 1.250126838684082,
+      "learning_rate": 2e-05,
+      "loss": 0.04465639,
+      "step": 17614
+    },
+    {
+      "epoch": 35.23,
+      "grad_norm": 1.335493803024292,
+      "learning_rate": 2e-05,
+      "loss": 0.04012146,
+      "step": 17615
+    },
+    {
+      "epoch": 35.232,
+      "grad_norm": 1.5321751832962036,
+      "learning_rate": 2e-05,
+      "loss": 0.06640027,
+      "step": 17616
+    },
+    {
+      "epoch": 35.234,
+      "grad_norm": 2.0908942222595215,
+      "learning_rate": 2e-05,
+      "loss": 0.05824264,
+      "step": 17617
+    },
+    {
+      "epoch": 35.236,
+      "grad_norm": 1.1407338380813599,
+      "learning_rate": 2e-05,
+      "loss": 0.04133571,
+      "step": 17618
+    },
+    {
+      "epoch": 35.238,
+      "grad_norm": 1.3628578186035156,
+      "learning_rate": 2e-05,
+      "loss": 0.03877375,
+      "step": 17619
+    },
+    {
+      "epoch": 35.24,
+      "grad_norm": 3.9366886615753174,
+      "learning_rate": 2e-05,
+      "loss": 0.0478052,
+      "step": 17620
+    },
+    {
+      "epoch": 35.242,
+      "grad_norm": 2.441206216812134,
+      "learning_rate": 2e-05,
+      "loss": 0.03485952,
+      "step": 17621
+    },
+    {
+      "epoch": 35.244,
+      "grad_norm": 2.0725955963134766,
+      "learning_rate": 2e-05,
+      "loss": 0.03543587,
+      "step": 17622
+    },
+    {
+      "epoch": 35.246,
+      "grad_norm": 1.0343211889266968,
+      "learning_rate": 2e-05,
+      "loss": 0.03453851,
+      "step": 17623
+    },
+    {
+      "epoch": 35.248,
+      "grad_norm": 0.9420909881591797,
+      "learning_rate": 2e-05,
+      "loss": 0.02860404,
+      "step": 17624
+    },
+    {
+      "epoch": 35.25,
+      "grad_norm": 1.3072288036346436,
+      "learning_rate": 2e-05,
+      "loss": 0.04773654,
+      "step": 17625
+    },
+    {
+      "epoch": 35.252,
+      "grad_norm": 1.3463304042816162,
+      "learning_rate": 2e-05,
+      "loss": 0.03698765,
+      "step": 17626
+    },
+    {
+      "epoch": 35.254,
+      "grad_norm": 2.281853199005127,
+      "learning_rate": 2e-05,
+      "loss": 0.05275848,
+      "step": 17627
+    },
+    {
+      "epoch": 35.256,
+      "grad_norm": 1.1776528358459473,
+      "learning_rate": 2e-05,
+      "loss": 0.02095211,
+      "step": 17628
+    },
+    {
+      "epoch": 35.258,
+      "grad_norm": 1.3505492210388184,
+      "learning_rate": 2e-05,
+      "loss": 0.04562983,
+      "step": 17629
+    },
+    {
+      "epoch": 35.26,
+      "grad_norm": 1.349131464958191,
+      "learning_rate": 2e-05,
+      "loss": 0.03608799,
+      "step": 17630
+    },
+    {
+      "epoch": 35.262,
+      "grad_norm": 1.218159794807434,
+      "learning_rate": 2e-05,
+      "loss": 0.0397279,
+      "step": 17631
+    },
+    {
+      "epoch": 35.264,
+      "grad_norm": 2.076681137084961,
+      "learning_rate": 2e-05,
+      "loss": 0.06574233,
+      "step": 17632
+    },
+    {
+      "epoch": 35.266,
+      "grad_norm": 1.0747252702713013,
+      "learning_rate": 2e-05,
+      "loss": 0.04180413,
+      "step": 17633
+    },
+    {
+      "epoch": 35.268,
+      "grad_norm": 1.5515995025634766,
+      "learning_rate": 2e-05,
+      "loss": 0.05342175,
+      "step": 17634
+    },
+    {
+      "epoch": 35.27,
+      "grad_norm": 1.4581546783447266,
+      "learning_rate": 2e-05,
+      "loss": 0.06602959,
+      "step": 17635
+    },
+    {
+      "epoch": 35.272,
+      "grad_norm": 3.7796173095703125,
+      "learning_rate": 2e-05,
+      "loss": 0.0564354,
+      "step": 17636
+    },
+    {
+      "epoch": 35.274,
+      "grad_norm": 3.0336711406707764,
+      "learning_rate": 2e-05,
+      "loss": 0.05401599,
+      "step": 17637
+    },
+    {
+      "epoch": 35.276,
+      "grad_norm": 1.2968264818191528,
+      "learning_rate": 2e-05,
+      "loss": 0.0414682,
+      "step": 17638
+    },
+    {
+      "epoch": 35.278,
+      "grad_norm": 2.132004737854004,
+      "learning_rate": 2e-05,
+      "loss": 0.0513651,
+      "step": 17639
+    },
+    {
+      "epoch": 35.28,
+      "grad_norm": 2.880918025970459,
+      "learning_rate": 2e-05,
+      "loss": 0.04456554,
+      "step": 17640
+    },
+    {
+      "epoch": 35.282,
+      "grad_norm": 1.0972630977630615,
+      "learning_rate": 2e-05,
+      "loss": 0.04144361,
+      "step": 17641
+    },
+    {
+      "epoch": 35.284,
+      "grad_norm": 1.0943173170089722,
+      "learning_rate": 2e-05,
+      "loss": 0.04478058,
+      "step": 17642
+    },
+    {
+      "epoch": 35.286,
+      "grad_norm": 1.1014378070831299,
+      "learning_rate": 2e-05,
+      "loss": 0.04401636,
+      "step": 17643
+    },
+    {
+      "epoch": 35.288,
+      "grad_norm": 1.2285698652267456,
+      "learning_rate": 2e-05,
+      "loss": 0.04485837,
+      "step": 17644
+    },
+    {
+      "epoch": 35.29,
+      "grad_norm": 1.2702783346176147,
+      "learning_rate": 2e-05,
+      "loss": 0.0508531,
+      "step": 17645
+    },
+    {
+      "epoch": 35.292,
+      "grad_norm": 1.179574728012085,
+      "learning_rate": 2e-05,
+      "loss": 0.03975598,
+      "step": 17646
+    },
+    {
+      "epoch": 35.294,
+      "grad_norm": 0.9671401381492615,
+      "learning_rate": 2e-05,
+      "loss": 0.03313099,
+      "step": 17647
+    },
+    {
+      "epoch": 35.296,
+      "grad_norm": 1.1078392267227173,
+      "learning_rate": 2e-05,
+      "loss": 0.03759495,
+      "step": 17648
+    },
+    {
+      "epoch": 35.298,
+      "grad_norm": 1.5418148040771484,
+      "learning_rate": 2e-05,
+      "loss": 0.0561055,
+      "step": 17649
+    },
+    {
+      "epoch": 35.3,
+      "grad_norm": 1.1559892892837524,
+      "learning_rate": 2e-05,
+      "loss": 0.04232976,
+      "step": 17650
+    },
+    {
+      "epoch": 35.302,
+      "grad_norm": 1.9095799922943115,
+      "learning_rate": 2e-05,
+      "loss": 0.05348632,
+      "step": 17651
+    },
+    {
+      "epoch": 35.304,
+      "grad_norm": 0.9116816520690918,
+      "learning_rate": 2e-05,
+      "loss": 0.03719225,
+      "step": 17652
+    },
+    {
+      "epoch": 35.306,
+      "grad_norm": 1.5335626602172852,
+      "learning_rate": 2e-05,
+      "loss": 0.0475245,
+      "step": 17653
+    },
+    {
+      "epoch": 35.308,
+      "grad_norm": 1.8686447143554688,
+      "learning_rate": 2e-05,
+      "loss": 0.0369437,
+      "step": 17654
+    },
+    {
+      "epoch": 35.31,
+      "grad_norm": 1.4260104894638062,
+      "learning_rate": 2e-05,
+      "loss": 0.0563136,
+      "step": 17655
+    },
+    {
+      "epoch": 35.312,
+      "grad_norm": 1.3176708221435547,
+      "learning_rate": 2e-05,
+      "loss": 0.04525885,
+      "step": 17656
+    },
+    {
+      "epoch": 35.314,
+      "grad_norm": 1.062448263168335,
+      "learning_rate": 2e-05,
+      "loss": 0.03579092,
+      "step": 17657
+    },
+    {
+      "epoch": 35.316,
+      "grad_norm": 0.8730036020278931,
+      "learning_rate": 2e-05,
+      "loss": 0.02629447,
+      "step": 17658
+    },
+    {
+      "epoch": 35.318,
+      "grad_norm": 1.2714184522628784,
+      "learning_rate": 2e-05,
+      "loss": 0.03522924,
+      "step": 17659
+    },
+    {
+      "epoch": 35.32,
+      "grad_norm": 1.3497344255447388,
+      "learning_rate": 2e-05,
+      "loss": 0.0463984,
+      "step": 17660
+    },
+    {
+      "epoch": 35.322,
+      "grad_norm": 2.49761962890625,
+      "learning_rate": 2e-05,
+      "loss": 0.0529076,
+      "step": 17661
+    },
+    {
+      "epoch": 35.324,
+      "grad_norm": 1.4788845777511597,
+      "learning_rate": 2e-05,
+      "loss": 0.04073741,
+      "step": 17662
+    },
+    {
+      "epoch": 35.326,
+      "grad_norm": 1.1707643270492554,
+      "learning_rate": 2e-05,
+      "loss": 0.03425131,
+      "step": 17663
+    },
+    {
+      "epoch": 35.328,
+      "grad_norm": 0.9647262692451477,
+      "learning_rate": 2e-05,
+      "loss": 0.02807866,
+      "step": 17664
+    },
+    {
+      "epoch": 35.33,
+      "grad_norm": 1.2062174081802368,
+      "learning_rate": 2e-05,
+      "loss": 0.06155042,
+      "step": 17665
+    },
+    {
+      "epoch": 35.332,
+      "grad_norm": 0.9606382250785828,
+      "learning_rate": 2e-05,
+      "loss": 0.03627588,
+      "step": 17666
+    },
+    {
+      "epoch": 35.334,
+      "grad_norm": 1.414086937904358,
+      "learning_rate": 2e-05,
+      "loss": 0.04845129,
+      "step": 17667
+    },
+    {
+      "epoch": 35.336,
+      "grad_norm": 1.6463829278945923,
+      "learning_rate": 2e-05,
+      "loss": 0.03860418,
+      "step": 17668
+    },
+    {
+      "epoch": 35.338,
+      "grad_norm": 1.1206401586532593,
+      "learning_rate": 2e-05,
+      "loss": 0.04083535,
+      "step": 17669
+    },
+    {
+      "epoch": 35.34,
+      "grad_norm": 2.738128662109375,
+      "learning_rate": 2e-05,
+      "loss": 0.06861308,
+      "step": 17670
+    },
+    {
+      "epoch": 35.342,
+      "grad_norm": 2.6043827533721924,
+      "learning_rate": 2e-05,
+      "loss": 0.05943043,
+      "step": 17671
+    },
+    {
+      "epoch": 35.344,
+      "grad_norm": 1.0954211950302124,
+      "learning_rate": 2e-05,
+      "loss": 0.04462023,
+      "step": 17672
+    },
+    {
+      "epoch": 35.346,
+      "grad_norm": 1.1352951526641846,
+      "learning_rate": 2e-05,
+      "loss": 0.05026999,
+      "step": 17673
+    },
+    {
+      "epoch": 35.348,
+      "grad_norm": 1.3933370113372803,
+      "learning_rate": 2e-05,
+      "loss": 0.03066497,
+      "step": 17674
+    },
+    {
+      "epoch": 35.35,
+      "grad_norm": 1.0098553895950317,
+      "learning_rate": 2e-05,
+      "loss": 0.04665726,
+      "step": 17675
+    },
+    {
+      "epoch": 35.352,
+      "grad_norm": 1.050323486328125,
+      "learning_rate": 2e-05,
+      "loss": 0.04057711,
+      "step": 17676
+    },
+    {
+      "epoch": 35.354,
+      "grad_norm": 1.5547417402267456,
+      "learning_rate": 2e-05,
+      "loss": 0.04004311,
+      "step": 17677
+    },
+    {
+      "epoch": 35.356,
+      "grad_norm": 0.8295509219169617,
+      "learning_rate": 2e-05,
+      "loss": 0.02311123,
+      "step": 17678
+    },
+    {
+      "epoch": 35.358,
+      "grad_norm": 1.6722297668457031,
+      "learning_rate": 2e-05,
+      "loss": 0.04569897,
+      "step": 17679
+    },
+    {
+      "epoch": 35.36,
+      "grad_norm": 1.2508385181427002,
+      "learning_rate": 2e-05,
+      "loss": 0.04461116,
+      "step": 17680
+    },
+    {
+      "epoch": 35.362,
+      "grad_norm": 1.20958411693573,
+      "learning_rate": 2e-05,
+      "loss": 0.04479233,
+      "step": 17681
+    },
+    {
+      "epoch": 35.364,
+      "grad_norm": 1.1637946367263794,
+      "learning_rate": 2e-05,
+      "loss": 0.04216972,
+      "step": 17682
+    },
+    {
+      "epoch": 35.366,
+      "grad_norm": 1.2432118654251099,
+      "learning_rate": 2e-05,
+      "loss": 0.0468252,
+      "step": 17683
+    },
+    {
+      "epoch": 35.368,
+      "grad_norm": 1.830465316772461,
+      "learning_rate": 2e-05,
+      "loss": 0.04752547,
+      "step": 17684
+    },
+    {
+      "epoch": 35.37,
+      "grad_norm": 1.0570487976074219,
+      "learning_rate": 2e-05,
+      "loss": 0.03741091,
+      "step": 17685
+    },
+    {
+      "epoch": 35.372,
+      "grad_norm": 2.6473779678344727,
+      "learning_rate": 2e-05,
+      "loss": 0.04489774,
+      "step": 17686
+    },
+    {
+      "epoch": 35.374,
+      "grad_norm": 1.851365327835083,
+      "learning_rate": 2e-05,
+      "loss": 0.04287185,
+      "step": 17687
+    },
+    {
+      "epoch": 35.376,
+      "grad_norm": 0.9998928308486938,
+      "learning_rate": 2e-05,
+      "loss": 0.02766431,
+      "step": 17688
+    },
+    {
+      "epoch": 35.378,
+      "grad_norm": 0.9591315388679504,
+      "learning_rate": 2e-05,
+      "loss": 0.04038401,
+      "step": 17689
+    },
+    {
+      "epoch": 35.38,
+      "grad_norm": 1.2864335775375366,
+      "learning_rate": 2e-05,
+      "loss": 0.03989007,
+      "step": 17690
+    },
+    {
+      "epoch": 35.382,
+      "grad_norm": 0.9475865364074707,
+      "learning_rate": 2e-05,
+      "loss": 0.02444749,
+      "step": 17691
+    },
+    {
+      "epoch": 35.384,
+      "grad_norm": 1.568764090538025,
+      "learning_rate": 2e-05,
+      "loss": 0.0419982,
+      "step": 17692
+    },
+    {
+      "epoch": 35.386,
+      "grad_norm": 1.4672602415084839,
+      "learning_rate": 2e-05,
+      "loss": 0.04082417,
+      "step": 17693
+    },
+    {
+      "epoch": 35.388,
+      "grad_norm": 1.2463346719741821,
+      "learning_rate": 2e-05,
+      "loss": 0.04095122,
+      "step": 17694
+    },
+    {
+      "epoch": 35.39,
+      "grad_norm": 0.9424423575401306,
+      "learning_rate": 2e-05,
+      "loss": 0.03527466,
+      "step": 17695
+    },
+    {
+      "epoch": 35.392,
+      "grad_norm": 1.1378228664398193,
+      "learning_rate": 2e-05,
+      "loss": 0.04225166,
+      "step": 17696
+    },
+    {
+      "epoch": 35.394,
+      "grad_norm": 0.9545497298240662,
+      "learning_rate": 2e-05,
+      "loss": 0.0355546,
+      "step": 17697
+    },
+    {
+      "epoch": 35.396,
+      "grad_norm": 1.1322404146194458,
+      "learning_rate": 2e-05,
+      "loss": 0.05043054,
+      "step": 17698
+    },
+    {
+      "epoch": 35.398,
+      "grad_norm": 1.0658605098724365,
+      "learning_rate": 2e-05,
+      "loss": 0.04704443,
+      "step": 17699
+    },
+    {
+      "epoch": 35.4,
+      "grad_norm": 1.517007827758789,
+      "learning_rate": 2e-05,
+      "loss": 0.03810771,
+      "step": 17700
+    },
+    {
+      "epoch": 35.402,
+      "grad_norm": 1.1072192192077637,
+      "learning_rate": 2e-05,
+      "loss": 0.04016921,
+      "step": 17701
+    },
+    {
+      "epoch": 35.404,
+      "grad_norm": 2.109288215637207,
+      "learning_rate": 2e-05,
+      "loss": 0.04785345,
+      "step": 17702
+    },
+    {
+      "epoch": 35.406,
+      "grad_norm": 2.865124464035034,
+      "learning_rate": 2e-05,
+      "loss": 0.04972965,
+      "step": 17703
+    },
+    {
+      "epoch": 35.408,
+      "grad_norm": 1.0339571237564087,
+      "learning_rate": 2e-05,
+      "loss": 0.03648417,
+      "step": 17704
+    },
+    {
+      "epoch": 35.41,
+      "grad_norm": 1.8117121458053589,
+      "learning_rate": 2e-05,
+      "loss": 0.04509541,
+      "step": 17705
+    },
+    {
+      "epoch": 35.412,
+      "grad_norm": 1.2131115198135376,
+      "learning_rate": 2e-05,
+      "loss": 0.04989497,
+      "step": 17706
+    },
+    {
+      "epoch": 35.414,
+      "grad_norm": 1.2343026399612427,
+      "learning_rate": 2e-05,
+      "loss": 0.04979831,
+      "step": 17707
+    },
+    {
+      "epoch": 35.416,
+      "grad_norm": 1.052886724472046,
+      "learning_rate": 2e-05,
+      "loss": 0.03326161,
+      "step": 17708
+    },
+    {
+      "epoch": 35.418,
+      "grad_norm": 1.1077262163162231,
+      "learning_rate": 2e-05,
+      "loss": 0.0355419,
+      "step": 17709
+    },
+    {
+      "epoch": 35.42,
+      "grad_norm": 1.4632622003555298,
+      "learning_rate": 2e-05,
+      "loss": 0.047338,
+      "step": 17710
+    },
+    {
+      "epoch": 35.422,
+      "grad_norm": 1.1542631387710571,
+      "learning_rate": 2e-05,
+      "loss": 0.04234635,
+      "step": 17711
+    },
+    {
+      "epoch": 35.424,
+      "grad_norm": 4.269802570343018,
+      "learning_rate": 2e-05,
+      "loss": 0.03470967,
+      "step": 17712
+    },
+    {
+      "epoch": 35.426,
+      "grad_norm": 1.3135454654693604,
+      "learning_rate": 2e-05,
+      "loss": 0.04146645,
+      "step": 17713
+    },
+    {
+      "epoch": 35.428,
+      "grad_norm": 1.0634958744049072,
+      "learning_rate": 2e-05,
+      "loss": 0.02393471,
+      "step": 17714
+    },
+    {
+      "epoch": 35.43,
+      "grad_norm": 1.000059723854065,
+      "learning_rate": 2e-05,
+      "loss": 0.0369064,
+      "step": 17715
+    },
+    {
+      "epoch": 35.432,
+      "grad_norm": 2.9598135948181152,
+      "learning_rate": 2e-05,
+      "loss": 0.04118183,
+      "step": 17716
+    },
+    {
+      "epoch": 35.434,
+      "grad_norm": 1.133852481842041,
+      "learning_rate": 2e-05,
+      "loss": 0.04435404,
+      "step": 17717
+    },
+    {
+      "epoch": 35.436,
+      "grad_norm": 0.9735626578330994,
+      "learning_rate": 2e-05,
+      "loss": 0.03352756,
+      "step": 17718
+    },
+    {
+      "epoch": 35.438,
+      "grad_norm": 1.356118083000183,
+      "learning_rate": 2e-05,
+      "loss": 0.05865297,
+      "step": 17719
+    },
+    {
+      "epoch": 35.44,
+      "grad_norm": 0.9971575736999512,
+      "learning_rate": 2e-05,
+      "loss": 0.02424793,
+      "step": 17720
+    },
+    {
+      "epoch": 35.442,
+      "grad_norm": 1.2533522844314575,
+      "learning_rate": 2e-05,
+      "loss": 0.03869835,
+      "step": 17721
+    },
+    {
+      "epoch": 35.444,
+      "grad_norm": 0.8965350985527039,
+      "learning_rate": 2e-05,
+      "loss": 0.03387743,
+      "step": 17722
+    },
+    {
+      "epoch": 35.446,
+      "grad_norm": 1.1389036178588867,
+      "learning_rate": 2e-05,
+      "loss": 0.04247169,
+      "step": 17723
+    },
+    {
+      "epoch": 35.448,
+      "grad_norm": 1.369778037071228,
+      "learning_rate": 2e-05,
+      "loss": 0.06438439,
+      "step": 17724
+    },
+    {
+      "epoch": 35.45,
+      "grad_norm": 1.1697176694869995,
+      "learning_rate": 2e-05,
+      "loss": 0.04246277,
+      "step": 17725
+    },
+    {
+      "epoch": 35.452,
+      "grad_norm": 1.5296801328659058,
+      "learning_rate": 2e-05,
+      "loss": 0.03574028,
+      "step": 17726
+    },
+    {
+      "epoch": 35.454,
+      "grad_norm": 1.0308504104614258,
+      "learning_rate": 2e-05,
+      "loss": 0.03716336,
+      "step": 17727
+    },
+    {
+      "epoch": 35.456,
+      "grad_norm": 1.0190324783325195,
+      "learning_rate": 2e-05,
+      "loss": 0.03794012,
+      "step": 17728
+    },
+    {
+      "epoch": 35.458,
+      "grad_norm": 1.2319554090499878,
+      "learning_rate": 2e-05,
+      "loss": 0.04770632,
+      "step": 17729
+    },
+    {
+      "epoch": 35.46,
+      "grad_norm": 1.7517224550247192,
+      "learning_rate": 2e-05,
+      "loss": 0.04705173,
+      "step": 17730
+    },
+    {
+      "epoch": 35.462,
+      "grad_norm": 1.5621715784072876,
+      "learning_rate": 2e-05,
+      "loss": 0.04366981,
+      "step": 17731
+    },
+    {
+      "epoch": 35.464,
+      "grad_norm": 1.3193645477294922,
+      "learning_rate": 2e-05,
+      "loss": 0.05270861,
+      "step": 17732
+    },
+    {
+      "epoch": 35.466,
+      "grad_norm": 0.9536468982696533,
+      "learning_rate": 2e-05,
+      "loss": 0.0323831,
+      "step": 17733
+    },
+    {
+      "epoch": 35.468,
+      "grad_norm": 1.2698876857757568,
+      "learning_rate": 2e-05,
+      "loss": 0.04317931,
+      "step": 17734
+    },
+    {
+      "epoch": 35.47,
+      "grad_norm": 1.2431633472442627,
+      "learning_rate": 2e-05,
+      "loss": 0.05195896,
+      "step": 17735
+    },
+    {
+      "epoch": 35.472,
+      "grad_norm": 1.0701299905776978,
+      "learning_rate": 2e-05,
+      "loss": 0.03547895,
+      "step": 17736
+    },
+    {
+      "epoch": 35.474,
+      "grad_norm": 1.2134490013122559,
+      "learning_rate": 2e-05,
+      "loss": 0.05419344,
+      "step": 17737
+    },
+    {
+      "epoch": 35.476,
+      "grad_norm": 1.2833257913589478,
+      "learning_rate": 2e-05,
+      "loss": 0.03962324,
+      "step": 17738
+    },
+    {
+      "epoch": 35.478,
+      "grad_norm": 1.1532343626022339,
+      "learning_rate": 2e-05,
+      "loss": 0.03333162,
+      "step": 17739
+    },
+    {
+      "epoch": 35.48,
+      "grad_norm": 1.1357736587524414,
+      "learning_rate": 2e-05,
+      "loss": 0.03868002,
+      "step": 17740
+    },
+    {
+      "epoch": 35.482,
+      "grad_norm": 1.2286957502365112,
+      "learning_rate": 2e-05,
+      "loss": 0.04705796,
+      "step": 17741
+    },
+    {
+      "epoch": 35.484,
+      "grad_norm": 1.0505681037902832,
+      "learning_rate": 2e-05,
+      "loss": 0.04804849,
+      "step": 17742
+    },
+    {
+      "epoch": 35.486,
+      "grad_norm": 1.327806830406189,
+      "learning_rate": 2e-05,
+      "loss": 0.04823949,
+      "step": 17743
+    },
+    {
+      "epoch": 35.488,
+      "grad_norm": 1.412100076675415,
+      "learning_rate": 2e-05,
+      "loss": 0.06005654,
+      "step": 17744
+    },
+    {
+      "epoch": 35.49,
+      "grad_norm": 1.1800956726074219,
+      "learning_rate": 2e-05,
+      "loss": 0.04362959,
+      "step": 17745
+    },
+    {
+      "epoch": 35.492,
+      "grad_norm": 1.503348708152771,
+      "learning_rate": 2e-05,
+      "loss": 0.03750347,
+      "step": 17746
+    },
+    {
+      "epoch": 35.494,
+      "grad_norm": 1.1118701696395874,
+      "learning_rate": 2e-05,
+      "loss": 0.03852141,
+      "step": 17747
+    },
+    {
+      "epoch": 35.496,
+      "grad_norm": 1.1212671995162964,
+      "learning_rate": 2e-05,
+      "loss": 0.0386279,
+      "step": 17748
+    },
+    {
+      "epoch": 35.498,
+      "grad_norm": 1.625443935394287,
+      "learning_rate": 2e-05,
+      "loss": 0.0379408,
+      "step": 17749
+    },
+    {
+      "epoch": 35.5,
+      "grad_norm": 1.2306393384933472,
+      "learning_rate": 2e-05,
+      "loss": 0.04350326,
+      "step": 17750
+    },
+    {
+      "epoch": 35.502,
+      "grad_norm": 2.8825604915618896,
+      "learning_rate": 2e-05,
+      "loss": 0.05902184,
+      "step": 17751
+    },
+    {
+      "epoch": 35.504,
+      "grad_norm": 0.9984104037284851,
+      "learning_rate": 2e-05,
+      "loss": 0.0298137,
+      "step": 17752
+    },
+    {
+      "epoch": 35.506,
+      "grad_norm": 1.1372625827789307,
+      "learning_rate": 2e-05,
+      "loss": 0.04307988,
+      "step": 17753
+    },
+    {
+      "epoch": 35.508,
+      "grad_norm": 1.264777660369873,
+      "learning_rate": 2e-05,
+      "loss": 0.05577639,
+      "step": 17754
+    },
+    {
+      "epoch": 35.51,
+      "grad_norm": 1.6673481464385986,
+      "learning_rate": 2e-05,
+      "loss": 0.03963385,
+      "step": 17755
+    },
+    {
+      "epoch": 35.512,
+      "grad_norm": 1.1486952304840088,
+      "learning_rate": 2e-05,
+      "loss": 0.0351126,
+      "step": 17756
+    },
+    {
+      "epoch": 35.514,
+      "grad_norm": 1.608984351158142,
+      "learning_rate": 2e-05,
+      "loss": 0.04021899,
+      "step": 17757
+    },
+    {
+      "epoch": 35.516,
+      "grad_norm": 1.772530436515808,
+      "learning_rate": 2e-05,
+      "loss": 0.05870032,
+      "step": 17758
+    },
+    {
+      "epoch": 35.518,
+      "grad_norm": 2.1825215816497803,
+      "learning_rate": 2e-05,
+      "loss": 0.03407493,
+      "step": 17759
+    },
+    {
+      "epoch": 35.52,
+      "grad_norm": 1.3578035831451416,
+      "learning_rate": 2e-05,
+      "loss": 0.04622669,
+      "step": 17760
+    },
+    {
+      "epoch": 35.522,
+      "grad_norm": 0.9831930994987488,
+      "learning_rate": 2e-05,
+      "loss": 0.0285987,
+      "step": 17761
+    },
+    {
+      "epoch": 35.524,
+      "grad_norm": 1.3356895446777344,
+      "learning_rate": 2e-05,
+      "loss": 0.04365962,
+      "step": 17762
+    },
+    {
+      "epoch": 35.526,
+      "grad_norm": 1.0443134307861328,
+      "learning_rate": 2e-05,
+      "loss": 0.03748023,
+      "step": 17763
+    },
+    {
+      "epoch": 35.528,
+      "grad_norm": 1.1682651042938232,
+      "learning_rate": 2e-05,
+      "loss": 0.04070065,
+      "step": 17764
+    },
+    {
+      "epoch": 35.53,
+      "grad_norm": 1.5059261322021484,
+      "learning_rate": 2e-05,
+      "loss": 0.03936179,
+      "step": 17765
+    },
+    {
+      "epoch": 35.532,
+      "grad_norm": 1.2437570095062256,
+      "learning_rate": 2e-05,
+      "loss": 0.05362586,
+      "step": 17766
+    },
+    {
+      "epoch": 35.534,
+      "grad_norm": 1.5657322406768799,
+      "learning_rate": 2e-05,
+      "loss": 0.04938463,
+      "step": 17767
+    },
+    {
+      "epoch": 35.536,
+      "grad_norm": 0.9113253355026245,
+      "learning_rate": 2e-05,
+      "loss": 0.02991488,
+      "step": 17768
+    },
+    {
+      "epoch": 35.538,
+      "grad_norm": 1.499922752380371,
+      "learning_rate": 2e-05,
+      "loss": 0.04125605,
+      "step": 17769
+    },
+    {
+      "epoch": 35.54,
+      "grad_norm": 1.880131721496582,
+      "learning_rate": 2e-05,
+      "loss": 0.04844569,
+      "step": 17770
+    },
+    {
+      "epoch": 35.542,
+      "grad_norm": 1.1775963306427002,
+      "learning_rate": 2e-05,
+      "loss": 0.0403573,
+      "step": 17771
+    },
+    {
+      "epoch": 35.544,
+      "grad_norm": 1.082275629043579,
+      "learning_rate": 2e-05,
+      "loss": 0.05027766,
+      "step": 17772
+    },
+    {
+      "epoch": 35.546,
+      "grad_norm": 1.3758782148361206,
+      "learning_rate": 2e-05,
+      "loss": 0.03608634,
+      "step": 17773
+    },
+    {
+      "epoch": 35.548,
+      "grad_norm": 1.6104998588562012,
+      "learning_rate": 2e-05,
+      "loss": 0.05908882,
+      "step": 17774
+    },
+    {
+      "epoch": 35.55,
+      "grad_norm": 1.236854076385498,
+      "learning_rate": 2e-05,
+      "loss": 0.0350708,
+      "step": 17775
+    },
+    {
+      "epoch": 35.552,
+      "grad_norm": 1.2313014268875122,
+      "learning_rate": 2e-05,
+      "loss": 0.0381471,
+      "step": 17776
+    },
+    {
+      "epoch": 35.554,
+      "grad_norm": 1.328316569328308,
+      "learning_rate": 2e-05,
+      "loss": 0.03309543,
+      "step": 17777
+    },
+    {
+      "epoch": 35.556,
+      "grad_norm": 0.9129838943481445,
+      "learning_rate": 2e-05,
+      "loss": 0.02543531,
+      "step": 17778
+    },
+    {
+      "epoch": 35.558,
+      "grad_norm": 1.7649102210998535,
+      "learning_rate": 2e-05,
+      "loss": 0.06082066,
+      "step": 17779
+    },
+    {
+      "epoch": 35.56,
+      "grad_norm": 1.1966280937194824,
+      "learning_rate": 2e-05,
+      "loss": 0.03694362,
+      "step": 17780
+    },
+    {
+      "epoch": 35.562,
+      "grad_norm": 1.4627177715301514,
+      "learning_rate": 2e-05,
+      "loss": 0.05721138,
+      "step": 17781
+    },
+    {
+      "epoch": 35.564,
+      "grad_norm": 1.157109260559082,
+      "learning_rate": 2e-05,
+      "loss": 0.03740972,
+      "step": 17782
+    },
+    {
+      "epoch": 35.566,
+      "grad_norm": 1.4257314205169678,
+      "learning_rate": 2e-05,
+      "loss": 0.03248692,
+      "step": 17783
+    },
+    {
+      "epoch": 35.568,
+      "grad_norm": 0.8779945969581604,
+      "learning_rate": 2e-05,
+      "loss": 0.02795019,
+      "step": 17784
+    },
+    {
+      "epoch": 35.57,
+      "grad_norm": 1.033923625946045,
+      "learning_rate": 2e-05,
+      "loss": 0.0340129,
+      "step": 17785
+    },
+    {
+      "epoch": 35.572,
+      "grad_norm": 1.221801996231079,
+      "learning_rate": 2e-05,
+      "loss": 0.04395582,
+      "step": 17786
+    },
+    {
+      "epoch": 35.574,
+      "grad_norm": 1.7551618814468384,
+      "learning_rate": 2e-05,
+      "loss": 0.03690699,
+      "step": 17787
+    },
+    {
+      "epoch": 35.576,
+      "grad_norm": 1.3354934453964233,
+      "learning_rate": 2e-05,
+      "loss": 0.05585198,
+      "step": 17788
+    },
+    {
+      "epoch": 35.578,
+      "grad_norm": 1.180930495262146,
+      "learning_rate": 2e-05,
+      "loss": 0.04893908,
+      "step": 17789
+    },
+    {
+      "epoch": 35.58,
+      "grad_norm": 1.4613933563232422,
+      "learning_rate": 2e-05,
+      "loss": 0.02561338,
+      "step": 17790
+    },
+    {
+      "epoch": 35.582,
+      "grad_norm": 1.3072844743728638,
+      "learning_rate": 2e-05,
+      "loss": 0.05350181,
+      "step": 17791
+    },
+    {
+      "epoch": 35.584,
+      "grad_norm": 1.5525765419006348,
+      "learning_rate": 2e-05,
+      "loss": 0.05785905,
+      "step": 17792
+    },
+    {
+      "epoch": 35.586,
+      "grad_norm": 1.8124005794525146,
+      "learning_rate": 2e-05,
+      "loss": 0.04928161,
+      "step": 17793
+    },
+    {
+      "epoch": 35.588,
+      "grad_norm": 1.0615509748458862,
+      "learning_rate": 2e-05,
+      "loss": 0.04574425,
+      "step": 17794
+    },
+    {
+      "epoch": 35.59,
+      "grad_norm": 1.153123140335083,
+      "learning_rate": 2e-05,
+      "loss": 0.04651198,
+      "step": 17795
+    },
+    {
+      "epoch": 35.592,
+      "grad_norm": 1.0932104587554932,
+      "learning_rate": 2e-05,
+      "loss": 0.03670096,
+      "step": 17796
+    },
+    {
+      "epoch": 35.594,
+      "grad_norm": 1.3459787368774414,
+      "learning_rate": 2e-05,
+      "loss": 0.04722689,
+      "step": 17797
+    },
+    {
+      "epoch": 35.596,
+      "grad_norm": 1.6134040355682373,
+      "learning_rate": 2e-05,
+      "loss": 0.05787089,
+      "step": 17798
+    },
+    {
+      "epoch": 35.598,
+      "grad_norm": 0.8614208698272705,
+      "learning_rate": 2e-05,
+      "loss": 0.02445607,
+      "step": 17799
+    },
+    {
+      "epoch": 35.6,
+      "grad_norm": 1.3574453592300415,
+      "learning_rate": 2e-05,
+      "loss": 0.05048421,
+      "step": 17800
+    },
+    {
+      "epoch": 35.602,
+      "grad_norm": 1.5967652797698975,
+      "learning_rate": 2e-05,
+      "loss": 0.05926722,
+      "step": 17801
+    },
+    {
+      "epoch": 35.604,
+      "grad_norm": 1.3235586881637573,
+      "learning_rate": 2e-05,
+      "loss": 0.04076,
+      "step": 17802
+    },
+    {
+      "epoch": 35.606,
+      "grad_norm": 1.4626275300979614,
+      "learning_rate": 2e-05,
+      "loss": 0.03682541,
+      "step": 17803
+    },
+    {
+      "epoch": 35.608,
+      "grad_norm": 1.6847271919250488,
+      "learning_rate": 2e-05,
+      "loss": 0.0393777,
+      "step": 17804
+    },
+    {
+      "epoch": 35.61,
+      "grad_norm": 2.216646194458008,
+      "learning_rate": 2e-05,
+      "loss": 0.03593418,
+      "step": 17805
+    },
+    {
+      "epoch": 35.612,
+      "grad_norm": 1.109135627746582,
+      "learning_rate": 2e-05,
+      "loss": 0.03311303,
+      "step": 17806
+    },
+    {
+      "epoch": 35.614,
+      "grad_norm": 1.2525417804718018,
+      "learning_rate": 2e-05,
+      "loss": 0.05682575,
+      "step": 17807
+    },
+    {
+      "epoch": 35.616,
+      "grad_norm": 1.1396535634994507,
+      "learning_rate": 2e-05,
+      "loss": 0.0337383,
+      "step": 17808
+    },
+    {
+      "epoch": 35.618,
+      "grad_norm": 0.8801586031913757,
+      "learning_rate": 2e-05,
+      "loss": 0.02758865,
+      "step": 17809
+    },
+    {
+      "epoch": 35.62,
+      "grad_norm": 1.4734177589416504,
+      "learning_rate": 2e-05,
+      "loss": 0.04279312,
+      "step": 17810
+    },
+    {
+      "epoch": 35.622,
+      "grad_norm": 1.2697399854660034,
+      "learning_rate": 2e-05,
+      "loss": 0.03729413,
+      "step": 17811
+    },
+    {
+      "epoch": 35.624,
+      "grad_norm": 1.2622113227844238,
+      "learning_rate": 2e-05,
+      "loss": 0.05834858,
+      "step": 17812
+    },
+    {
+      "epoch": 35.626,
+      "grad_norm": 1.3223472833633423,
+      "learning_rate": 2e-05,
+      "loss": 0.04107845,
+      "step": 17813
+    },
+    {
+      "epoch": 35.628,
+      "grad_norm": 1.3146884441375732,
+      "learning_rate": 2e-05,
+      "loss": 0.05463982,
+      "step": 17814
+    },
+    {
+      "epoch": 35.63,
+      "grad_norm": 0.9225746393203735,
+      "learning_rate": 2e-05,
+      "loss": 0.02599794,
+      "step": 17815
+    },
+    {
+      "epoch": 35.632,
+      "grad_norm": 1.4230635166168213,
+      "learning_rate": 2e-05,
+      "loss": 0.04019064,
+      "step": 17816
+    },
+    {
+      "epoch": 35.634,
+      "grad_norm": 1.6114380359649658,
+      "learning_rate": 2e-05,
+      "loss": 0.05018803,
+      "step": 17817
+    },
+    {
+      "epoch": 35.636,
+      "grad_norm": 1.1191251277923584,
+      "learning_rate": 2e-05,
+      "loss": 0.03190347,
+      "step": 17818
+    },
+    {
+      "epoch": 35.638,
+      "grad_norm": 2.0187177658081055,
+      "learning_rate": 2e-05,
+      "loss": 0.04886641,
+      "step": 17819
+    },
+    {
+      "epoch": 35.64,
+      "grad_norm": 1.0409353971481323,
+      "learning_rate": 2e-05,
+      "loss": 0.03332942,
+      "step": 17820
+    },
+    {
+      "epoch": 35.642,
+      "grad_norm": 1.3523070812225342,
+      "learning_rate": 2e-05,
+      "loss": 0.05481984,
+      "step": 17821
+    },
+    {
+      "epoch": 35.644,
+      "grad_norm": 0.8652260899543762,
+      "learning_rate": 2e-05,
+      "loss": 0.02689623,
+      "step": 17822
+    },
+    {
+      "epoch": 35.646,
+      "grad_norm": 1.0217618942260742,
+      "learning_rate": 2e-05,
+      "loss": 0.03837987,
+      "step": 17823
+    },
+    {
+      "epoch": 35.648,
+      "grad_norm": 2.860708713531494,
+      "learning_rate": 2e-05,
+      "loss": 0.0455092,
+      "step": 17824
+    },
+    {
+      "epoch": 35.65,
+      "grad_norm": 1.2912145853042603,
+      "learning_rate": 2e-05,
+      "loss": 0.04535623,
+      "step": 17825
+    },
+    {
+      "epoch": 35.652,
+      "grad_norm": 1.1695078611373901,
+      "learning_rate": 2e-05,
+      "loss": 0.047709,
+      "step": 17826
+    },
+    {
+      "epoch": 35.654,
+      "grad_norm": 1.5502129793167114,
+      "learning_rate": 2e-05,
+      "loss": 0.05648194,
+      "step": 17827
+    },
+    {
+      "epoch": 35.656,
+      "grad_norm": 0.9700777530670166,
+      "learning_rate": 2e-05,
+      "loss": 0.02404735,
+      "step": 17828
+    },
+    {
+      "epoch": 35.658,
+      "grad_norm": 1.0643705129623413,
+      "learning_rate": 2e-05,
+      "loss": 0.02897286,
+      "step": 17829
+    },
+    {
+      "epoch": 35.66,
+      "grad_norm": 1.0671511888504028,
+      "learning_rate": 2e-05,
+      "loss": 0.03070662,
+      "step": 17830
+    },
+    {
+      "epoch": 35.662,
+      "grad_norm": 1.3201302289962769,
+      "learning_rate": 2e-05,
+      "loss": 0.04533841,
+      "step": 17831
+    },
+    {
+      "epoch": 35.664,
+      "grad_norm": 1.1541919708251953,
+      "learning_rate": 2e-05,
+      "loss": 0.044394,
+      "step": 17832
+    },
+    {
+      "epoch": 35.666,
+      "grad_norm": 1.3019088506698608,
+      "learning_rate": 2e-05,
+      "loss": 0.06686329,
+      "step": 17833
+    },
+    {
+      "epoch": 35.668,
+      "grad_norm": 1.0324286222457886,
+      "learning_rate": 2e-05,
+      "loss": 0.04079415,
+      "step": 17834
+    },
+    {
+      "epoch": 35.67,
+      "grad_norm": 1.0423561334609985,
+      "learning_rate": 2e-05,
+      "loss": 0.03136035,
+      "step": 17835
+    },
+    {
+      "epoch": 35.672,
+      "grad_norm": 1.6549469232559204,
+      "learning_rate": 2e-05,
+      "loss": 0.0495429,
+      "step": 17836
+    },
+    {
+      "epoch": 35.674,
+      "grad_norm": 3.071321725845337,
+      "learning_rate": 2e-05,
+      "loss": 0.04680189,
+      "step": 17837
+    },
+    {
+      "epoch": 35.676,
+      "grad_norm": 1.4656696319580078,
+      "learning_rate": 2e-05,
+      "loss": 0.04357903,
+      "step": 17838
+    },
+    {
+      "epoch": 35.678,
+      "grad_norm": 1.0006139278411865,
+      "learning_rate": 2e-05,
+      "loss": 0.02649287,
+      "step": 17839
+    },
+    {
+      "epoch": 35.68,
+      "grad_norm": 0.8368288278579712,
+      "learning_rate": 2e-05,
+      "loss": 0.02872172,
+      "step": 17840
+    },
+    {
+      "epoch": 35.682,
+      "grad_norm": 1.1067603826522827,
+      "learning_rate": 2e-05,
+      "loss": 0.04076967,
+      "step": 17841
+    },
+    {
+      "epoch": 35.684,
+      "grad_norm": 1.4088644981384277,
+      "learning_rate": 2e-05,
+      "loss": 0.04598125,
+      "step": 17842
+    },
+    {
+      "epoch": 35.686,
+      "grad_norm": 1.1588618755340576,
+      "learning_rate": 2e-05,
+      "loss": 0.03723881,
+      "step": 17843
+    },
+    {
+      "epoch": 35.688,
+      "grad_norm": 1.1325336694717407,
+      "learning_rate": 2e-05,
+      "loss": 0.04327301,
+      "step": 17844
+    },
+    {
+      "epoch": 35.69,
+      "grad_norm": 2.3693675994873047,
+      "learning_rate": 2e-05,
+      "loss": 0.05009006,
+      "step": 17845
+    },
+    {
+      "epoch": 35.692,
+      "grad_norm": 1.2514508962631226,
+      "learning_rate": 2e-05,
+      "loss": 0.03267618,
+      "step": 17846
+    },
+    {
+      "epoch": 35.694,
+      "grad_norm": 1.3177813291549683,
+      "learning_rate": 2e-05,
+      "loss": 0.02680693,
+      "step": 17847
+    },
+    {
+      "epoch": 35.696,
+      "grad_norm": 1.2522759437561035,
+      "learning_rate": 2e-05,
+      "loss": 0.04521048,
+      "step": 17848
+    },
+    {
+      "epoch": 35.698,
+      "grad_norm": 1.2396931648254395,
+      "learning_rate": 2e-05,
+      "loss": 0.04618759,
+      "step": 17849
+    },
+    {
+      "epoch": 35.7,
+      "grad_norm": 1.2118010520935059,
+      "learning_rate": 2e-05,
+      "loss": 0.03588796,
+      "step": 17850
+    },
+    {
+      "epoch": 35.702,
+      "grad_norm": 1.1289559602737427,
+      "learning_rate": 2e-05,
+      "loss": 0.04039676,
+      "step": 17851
+    },
+    {
+      "epoch": 35.704,
+      "grad_norm": 2.2850546836853027,
+      "learning_rate": 2e-05,
+      "loss": 0.04581477,
+      "step": 17852
+    },
+    {
+      "epoch": 35.706,
+      "grad_norm": 1.0945602655410767,
+      "learning_rate": 2e-05,
+      "loss": 0.02884125,
+      "step": 17853
+    },
+    {
+      "epoch": 35.708,
+      "grad_norm": 1.218193769454956,
+      "learning_rate": 2e-05,
+      "loss": 0.03631239,
+      "step": 17854
+    },
+    {
+      "epoch": 35.71,
+      "grad_norm": 1.4844825267791748,
+      "learning_rate": 2e-05,
+      "loss": 0.02633794,
+      "step": 17855
+    },
+    {
+      "epoch": 35.712,
+      "grad_norm": 2.8186700344085693,
+      "learning_rate": 2e-05,
+      "loss": 0.05136289,
+      "step": 17856
+    },
+    {
+      "epoch": 35.714,
+      "grad_norm": 1.0413309335708618,
+      "learning_rate": 2e-05,
+      "loss": 0.02841013,
+      "step": 17857
+    },
+    {
+      "epoch": 35.716,
+      "grad_norm": 1.4248929023742676,
+      "learning_rate": 2e-05,
+      "loss": 0.04046667,
+      "step": 17858
+    },
+    {
+      "epoch": 35.718,
+      "grad_norm": 1.086219310760498,
+      "learning_rate": 2e-05,
+      "loss": 0.03643753,
+      "step": 17859
+    },
+    {
+      "epoch": 35.72,
+      "grad_norm": 1.0413047075271606,
+      "learning_rate": 2e-05,
+      "loss": 0.03506773,
+      "step": 17860
+    },
+    {
+      "epoch": 35.722,
+      "grad_norm": 1.138818383216858,
+      "learning_rate": 2e-05,
+      "loss": 0.04383482,
+      "step": 17861
+    },
+    {
+      "epoch": 35.724,
+      "grad_norm": 1.0784032344818115,
+      "learning_rate": 2e-05,
+      "loss": 0.04035743,
+      "step": 17862
+    },
+    {
+      "epoch": 35.726,
+      "grad_norm": 1.1928517818450928,
+      "learning_rate": 2e-05,
+      "loss": 0.04039065,
+      "step": 17863
+    },
+    {
+      "epoch": 35.728,
+      "grad_norm": 1.3450567722320557,
+      "learning_rate": 2e-05,
+      "loss": 0.051686,
+      "step": 17864
+    },
+    {
+      "epoch": 35.73,
+      "grad_norm": 1.3145865201950073,
+      "learning_rate": 2e-05,
+      "loss": 0.05856508,
+      "step": 17865
+    },
+    {
+      "epoch": 35.732,
+      "grad_norm": 1.7338427305221558,
+      "learning_rate": 2e-05,
+      "loss": 0.05495846,
+      "step": 17866
+    },
+    {
+      "epoch": 35.734,
+      "grad_norm": 1.2222492694854736,
+      "learning_rate": 2e-05,
+      "loss": 0.04681749,
+      "step": 17867
+    },
+    {
+      "epoch": 35.736,
+      "grad_norm": 2.382176399230957,
+      "learning_rate": 2e-05,
+      "loss": 0.0592317,
+      "step": 17868
+    },
+    {
+      "epoch": 35.738,
+      "grad_norm": 1.3851975202560425,
+      "learning_rate": 2e-05,
+      "loss": 0.05830447,
+      "step": 17869
+    },
+    {
+      "epoch": 35.74,
+      "grad_norm": 1.8277791738510132,
+      "learning_rate": 2e-05,
+      "loss": 0.04027114,
+      "step": 17870
+    },
+    {
+      "epoch": 35.742,
+      "grad_norm": 1.3364810943603516,
+      "learning_rate": 2e-05,
+      "loss": 0.04788211,
+      "step": 17871
+    },
+    {
+      "epoch": 35.744,
+      "grad_norm": 1.4539287090301514,
+      "learning_rate": 2e-05,
+      "loss": 0.05281156,
+      "step": 17872
+    },
+    {
+      "epoch": 35.746,
+      "grad_norm": 2.4535436630249023,
+      "learning_rate": 2e-05,
+      "loss": 0.05732207,
+      "step": 17873
+    },
+    {
+      "epoch": 35.748,
+      "grad_norm": 1.3156609535217285,
+      "learning_rate": 2e-05,
+      "loss": 0.04180748,
+      "step": 17874
+    },
+    {
+      "epoch": 35.75,
+      "grad_norm": 1.083983063697815,
+      "learning_rate": 2e-05,
+      "loss": 0.02724944,
+      "step": 17875
+    },
+    {
+      "epoch": 35.752,
+      "grad_norm": 1.2590404748916626,
+      "learning_rate": 2e-05,
+      "loss": 0.04488721,
+      "step": 17876
+    },
+    {
+      "epoch": 35.754,
+      "grad_norm": 1.1767133474349976,
+      "learning_rate": 2e-05,
+      "loss": 0.03508144,
+      "step": 17877
+    },
+    {
+      "epoch": 35.756,
+      "grad_norm": 0.9333146810531616,
+      "learning_rate": 2e-05,
+      "loss": 0.03762346,
+      "step": 17878
+    },
+    {
+      "epoch": 35.758,
+      "grad_norm": 3.9581522941589355,
+      "learning_rate": 2e-05,
+      "loss": 0.05363543,
+      "step": 17879
+    },
+    {
+      "epoch": 35.76,
+      "grad_norm": 1.9999139308929443,
+      "learning_rate": 2e-05,
+      "loss": 0.05396073,
+      "step": 17880
+    },
+    {
+      "epoch": 35.762,
+      "grad_norm": 1.12847900390625,
+      "learning_rate": 2e-05,
+      "loss": 0.04188031,
+      "step": 17881
+    },
+    {
+      "epoch": 35.764,
+      "grad_norm": 1.221652626991272,
+      "learning_rate": 2e-05,
+      "loss": 0.04514348,
+      "step": 17882
+    },
+    {
+      "epoch": 35.766,
+      "grad_norm": 1.2314302921295166,
+      "learning_rate": 2e-05,
+      "loss": 0.04036722,
+      "step": 17883
+    },
+    {
+      "epoch": 35.768,
+      "grad_norm": 1.4413890838623047,
+      "learning_rate": 2e-05,
+      "loss": 0.05193814,
+      "step": 17884
+    },
+    {
+      "epoch": 35.77,
+      "grad_norm": 1.3951367139816284,
+      "learning_rate": 2e-05,
+      "loss": 0.03265073,
+      "step": 17885
+    },
+    {
+      "epoch": 35.772,
+      "grad_norm": 1.0324435234069824,
+      "learning_rate": 2e-05,
+      "loss": 0.03620376,
+      "step": 17886
+    },
+    {
+      "epoch": 35.774,
+      "grad_norm": 1.1713546514511108,
+      "learning_rate": 2e-05,
+      "loss": 0.04896335,
+      "step": 17887
+    },
+    {
+      "epoch": 35.776,
+      "grad_norm": 0.991287350654602,
+      "learning_rate": 2e-05,
+      "loss": 0.03548204,
+      "step": 17888
+    },
+    {
+      "epoch": 35.778,
+      "grad_norm": 2.7516236305236816,
+      "learning_rate": 2e-05,
+      "loss": 0.06211228,
+      "step": 17889
+    },
+    {
+      "epoch": 35.78,
+      "grad_norm": 2.7214620113372803,
+      "learning_rate": 2e-05,
+      "loss": 0.05013406,
+      "step": 17890
+    },
+    {
+      "epoch": 35.782,
+      "grad_norm": 1.233435034751892,
+      "learning_rate": 2e-05,
+      "loss": 0.04874311,
+      "step": 17891
+    },
+    {
+      "epoch": 35.784,
+      "grad_norm": 1.7725192308425903,
+      "learning_rate": 2e-05,
+      "loss": 0.04005573,
+      "step": 17892
+    },
+    {
+      "epoch": 35.786,
+      "grad_norm": 1.3703975677490234,
+      "learning_rate": 2e-05,
+      "loss": 0.04984409,
+      "step": 17893
+    },
+    {
+      "epoch": 35.788,
+      "grad_norm": 1.3456453084945679,
+      "learning_rate": 2e-05,
+      "loss": 0.04753321,
+      "step": 17894
+    },
+    {
+      "epoch": 35.79,
+      "grad_norm": 1.2621350288391113,
+      "learning_rate": 2e-05,
+      "loss": 0.05199986,
+      "step": 17895
+    },
+    {
+      "epoch": 35.792,
+      "grad_norm": 1.3152614831924438,
+      "learning_rate": 2e-05,
+      "loss": 0.03647928,
+      "step": 17896
+    },
+    {
+      "epoch": 35.794,
+      "grad_norm": 0.9398272633552551,
+      "learning_rate": 2e-05,
+      "loss": 0.02891801,
+      "step": 17897
+    },
+    {
+      "epoch": 35.796,
+      "grad_norm": 1.0658025741577148,
+      "learning_rate": 2e-05,
+      "loss": 0.03435162,
+      "step": 17898
+    },
+    {
+      "epoch": 35.798,
+      "grad_norm": 1.157807469367981,
+      "learning_rate": 2e-05,
+      "loss": 0.03711173,
+      "step": 17899
+    },
+    {
+      "epoch": 35.8,
+      "grad_norm": 2.451808214187622,
+      "learning_rate": 2e-05,
+      "loss": 0.03681424,
+      "step": 17900
+    },
+    {
+      "epoch": 35.802,
+      "grad_norm": 1.0727334022521973,
+      "learning_rate": 2e-05,
+      "loss": 0.04209089,
+      "step": 17901
+    },
+    {
+      "epoch": 35.804,
+      "grad_norm": 1.314285159111023,
+      "learning_rate": 2e-05,
+      "loss": 0.0380992,
+      "step": 17902
+    },
+    {
+      "epoch": 35.806,
+      "grad_norm": 1.1336785554885864,
+      "learning_rate": 2e-05,
+      "loss": 0.03314412,
+      "step": 17903
+    },
+    {
+      "epoch": 35.808,
+      "grad_norm": 1.7368667125701904,
+      "learning_rate": 2e-05,
+      "loss": 0.0345618,
+      "step": 17904
+    },
+    {
+      "epoch": 35.81,
+      "grad_norm": 2.1174581050872803,
+      "learning_rate": 2e-05,
+      "loss": 0.06510151,
+      "step": 17905
+    },
+    {
+      "epoch": 35.812,
+      "grad_norm": 1.6650149822235107,
+      "learning_rate": 2e-05,
+      "loss": 0.04526421,
+      "step": 17906
+    },
+    {
+      "epoch": 35.814,
+      "grad_norm": 1.2398277521133423,
+      "learning_rate": 2e-05,
+      "loss": 0.037236,
+      "step": 17907
+    },
+    {
+      "epoch": 35.816,
+      "grad_norm": 1.1168713569641113,
+      "learning_rate": 2e-05,
+      "loss": 0.03919902,
+      "step": 17908
+    },
+    {
+      "epoch": 35.818,
+      "grad_norm": 1.6926127672195435,
+      "learning_rate": 2e-05,
+      "loss": 0.05567209,
+      "step": 17909
+    },
+    {
+      "epoch": 35.82,
+      "grad_norm": 0.9734522104263306,
+      "learning_rate": 2e-05,
+      "loss": 0.0383275,
+      "step": 17910
+    },
+    {
+      "epoch": 35.822,
+      "grad_norm": 1.0680774450302124,
+      "learning_rate": 2e-05,
+      "loss": 0.04609864,
+      "step": 17911
+    },
+    {
+      "epoch": 35.824,
+      "grad_norm": 1.3214465379714966,
+      "learning_rate": 2e-05,
+      "loss": 0.04425301,
+      "step": 17912
+    },
+    {
+      "epoch": 35.826,
+      "grad_norm": 1.187393307685852,
+      "learning_rate": 2e-05,
+      "loss": 0.04773477,
+      "step": 17913
+    },
+    {
+      "epoch": 35.828,
+      "grad_norm": 2.846890926361084,
+      "learning_rate": 2e-05,
+      "loss": 0.04056527,
+      "step": 17914
+    },
+    {
+      "epoch": 35.83,
+      "grad_norm": 1.2541725635528564,
+      "learning_rate": 2e-05,
+      "loss": 0.04198845,
+      "step": 17915
+    },
+    {
+      "epoch": 35.832,
+      "grad_norm": 1.2594901323318481,
+      "learning_rate": 2e-05,
+      "loss": 0.04509019,
+      "step": 17916
+    },
+    {
+      "epoch": 35.834,
+      "grad_norm": 1.6290315389633179,
+      "learning_rate": 2e-05,
+      "loss": 0.04840211,
+      "step": 17917
+    },
+    {
+      "epoch": 35.836,
+      "grad_norm": 1.9645651578903198,
+      "learning_rate": 2e-05,
+      "loss": 0.04485439,
+      "step": 17918
+    },
+    {
+      "epoch": 35.838,
+      "grad_norm": 1.1047340631484985,
+      "learning_rate": 2e-05,
+      "loss": 0.04456278,
+      "step": 17919
+    },
+    {
+      "epoch": 35.84,
+      "grad_norm": 1.4236819744110107,
+      "learning_rate": 2e-05,
+      "loss": 0.05413448,
+      "step": 17920
+    },
+    {
+      "epoch": 35.842,
+      "grad_norm": 1.3188161849975586,
+      "learning_rate": 2e-05,
+      "loss": 0.05367437,
+      "step": 17921
+    },
+    {
+      "epoch": 35.844,
+      "grad_norm": 1.1377904415130615,
+      "learning_rate": 2e-05,
+      "loss": 0.04965197,
+      "step": 17922
+    },
+    {
+      "epoch": 35.846,
+      "grad_norm": 2.1205976009368896,
+      "learning_rate": 2e-05,
+      "loss": 0.05599382,
+      "step": 17923
+    },
+    {
+      "epoch": 35.848,
+      "grad_norm": 1.2316977977752686,
+      "learning_rate": 2e-05,
+      "loss": 0.04126142,
+      "step": 17924
+    },
+    {
+      "epoch": 35.85,
+      "grad_norm": 1.2489413022994995,
+      "learning_rate": 2e-05,
+      "loss": 0.04298563,
+      "step": 17925
+    },
+    {
+      "epoch": 35.852,
+      "grad_norm": 1.4723291397094727,
+      "learning_rate": 2e-05,
+      "loss": 0.04364316,
+      "step": 17926
+    },
+    {
+      "epoch": 35.854,
+      "grad_norm": 0.9845232963562012,
+      "learning_rate": 2e-05,
+      "loss": 0.03328026,
+      "step": 17927
+    },
+    {
+      "epoch": 35.856,
+      "grad_norm": 1.11769700050354,
+      "learning_rate": 2e-05,
+      "loss": 0.03830117,
+      "step": 17928
+    },
+    {
+      "epoch": 35.858,
+      "grad_norm": 1.0789967775344849,
+      "learning_rate": 2e-05,
+      "loss": 0.03304796,
+      "step": 17929
+    },
+    {
+      "epoch": 35.86,
+      "grad_norm": 1.4628854990005493,
+      "learning_rate": 2e-05,
+      "loss": 0.04162867,
+      "step": 17930
+    },
+    {
+      "epoch": 35.862,
+      "grad_norm": 1.1505697965621948,
+      "learning_rate": 2e-05,
+      "loss": 0.04003832,
+      "step": 17931
+    },
+    {
+      "epoch": 35.864,
+      "grad_norm": 0.9786707758903503,
+      "learning_rate": 2e-05,
+      "loss": 0.04291322,
+      "step": 17932
+    },
+    {
+      "epoch": 35.866,
+      "grad_norm": 1.514482021331787,
+      "learning_rate": 2e-05,
+      "loss": 0.04682028,
+      "step": 17933
+    },
+    {
+      "epoch": 35.868,
+      "grad_norm": 1.150888442993164,
+      "learning_rate": 2e-05,
+      "loss": 0.04786021,
+      "step": 17934
+    },
+    {
+      "epoch": 35.87,
+      "grad_norm": 1.2002613544464111,
+      "learning_rate": 2e-05,
+      "loss": 0.0451806,
+      "step": 17935
+    },
+    {
+      "epoch": 35.872,
+      "grad_norm": 1.3272571563720703,
+      "learning_rate": 2e-05,
+      "loss": 0.05705176,
+      "step": 17936
+    },
+    {
+      "epoch": 35.874,
+      "grad_norm": 1.1266093254089355,
+      "learning_rate": 2e-05,
+      "loss": 0.03312952,
+      "step": 17937
+    },
+    {
+      "epoch": 35.876,
+      "grad_norm": 1.2619445323944092,
+      "learning_rate": 2e-05,
+      "loss": 0.04542191,
+      "step": 17938
+    },
+    {
+      "epoch": 35.878,
+      "grad_norm": 1.2090903520584106,
+      "learning_rate": 2e-05,
+      "loss": 0.04657109,
+      "step": 17939
+    },
+    {
+      "epoch": 35.88,
+      "grad_norm": 1.0083225965499878,
+      "learning_rate": 2e-05,
+      "loss": 0.03867437,
+      "step": 17940
+    },
+    {
+      "epoch": 35.882,
+      "grad_norm": 1.0939233303070068,
+      "learning_rate": 2e-05,
+      "loss": 0.02842722,
+      "step": 17941
+    },
+    {
+      "epoch": 35.884,
+      "grad_norm": 1.0067133903503418,
+      "learning_rate": 2e-05,
+      "loss": 0.03263469,
+      "step": 17942
+    },
+    {
+      "epoch": 35.886,
+      "grad_norm": 1.1299545764923096,
+      "learning_rate": 2e-05,
+      "loss": 0.04446633,
+      "step": 17943
+    },
+    {
+      "epoch": 35.888,
+      "grad_norm": 1.0273183584213257,
+      "learning_rate": 2e-05,
+      "loss": 0.05045258,
+      "step": 17944
+    },
+    {
+      "epoch": 35.89,
+      "grad_norm": 1.154341697692871,
+      "learning_rate": 2e-05,
+      "loss": 0.03366087,
+      "step": 17945
+    },
+    {
+      "epoch": 35.892,
+      "grad_norm": 1.1345313787460327,
+      "learning_rate": 2e-05,
+      "loss": 0.03205415,
+      "step": 17946
+    },
+    {
+      "epoch": 35.894,
+      "grad_norm": 1.1819157600402832,
+      "learning_rate": 2e-05,
+      "loss": 0.04350836,
+      "step": 17947
+    },
+    {
+      "epoch": 35.896,
+      "grad_norm": 1.7022780179977417,
+      "learning_rate": 2e-05,
+      "loss": 0.048735,
+      "step": 17948
+    },
+    {
+      "epoch": 35.898,
+      "grad_norm": 0.8520524501800537,
+      "learning_rate": 2e-05,
+      "loss": 0.02894574,
+      "step": 17949
+    },
+    {
+      "epoch": 35.9,
+      "grad_norm": 1.0947223901748657,
+      "learning_rate": 2e-05,
+      "loss": 0.03767444,
+      "step": 17950
+    },
+    {
+      "epoch": 35.902,
+      "grad_norm": 2.9594931602478027,
+      "learning_rate": 2e-05,
+      "loss": 0.0309197,
+      "step": 17951
+    },
+    {
+      "epoch": 35.904,
+      "grad_norm": 1.3062098026275635,
+      "learning_rate": 2e-05,
+      "loss": 0.04968993,
+      "step": 17952
+    },
+    {
+      "epoch": 35.906,
+      "grad_norm": 1.0888330936431885,
+      "learning_rate": 2e-05,
+      "loss": 0.04075143,
+      "step": 17953
+    },
+    {
+      "epoch": 35.908,
+      "grad_norm": 1.0041322708129883,
+      "learning_rate": 2e-05,
+      "loss": 0.03644698,
+      "step": 17954
+    },
+    {
+      "epoch": 35.91,
+      "grad_norm": 1.0307637453079224,
+      "learning_rate": 2e-05,
+      "loss": 0.04232953,
+      "step": 17955
+    },
+    {
+      "epoch": 35.912,
+      "grad_norm": 1.2270115613937378,
+      "learning_rate": 2e-05,
+      "loss": 0.04430905,
+      "step": 17956
+    },
+    {
+      "epoch": 35.914,
+      "grad_norm": 1.2956207990646362,
+      "learning_rate": 2e-05,
+      "loss": 0.06014199,
+      "step": 17957
+    },
+    {
+      "epoch": 35.916,
+      "grad_norm": 1.5076323747634888,
+      "learning_rate": 2e-05,
+      "loss": 0.0593874,
+      "step": 17958
+    },
+    {
+      "epoch": 35.918,
+      "grad_norm": 1.1515308618545532,
+      "learning_rate": 2e-05,
+      "loss": 0.04382184,
+      "step": 17959
+    },
+    {
+      "epoch": 35.92,
+      "grad_norm": 1.1946406364440918,
+      "learning_rate": 2e-05,
+      "loss": 0.05360639,
+      "step": 17960
+    },
+    {
+      "epoch": 35.922,
+      "grad_norm": 1.332234501838684,
+      "learning_rate": 2e-05,
+      "loss": 0.03495939,
+      "step": 17961
+    },
+    {
+      "epoch": 35.924,
+      "grad_norm": 1.5882817506790161,
+      "learning_rate": 2e-05,
+      "loss": 0.02272969,
+      "step": 17962
+    },
+    {
+      "epoch": 35.926,
+      "grad_norm": 1.1679266691207886,
+      "learning_rate": 2e-05,
+      "loss": 0.05053775,
+      "step": 17963
+    },
+    {
+      "epoch": 35.928,
+      "grad_norm": 1.0632550716400146,
+      "learning_rate": 2e-05,
+      "loss": 0.04534806,
+      "step": 17964
+    },
+    {
+      "epoch": 35.93,
+      "grad_norm": 1.6096125841140747,
+      "learning_rate": 2e-05,
+      "loss": 0.0572957,
+      "step": 17965
+    },
+    {
+      "epoch": 35.932,
+      "grad_norm": 1.2726202011108398,
+      "learning_rate": 2e-05,
+      "loss": 0.04721672,
+      "step": 17966
+    },
+    {
+      "epoch": 35.934,
+      "grad_norm": 1.637152075767517,
+      "learning_rate": 2e-05,
+      "loss": 0.03014551,
+      "step": 17967
+    },
+    {
+      "epoch": 35.936,
+      "grad_norm": 1.2941720485687256,
+      "learning_rate": 2e-05,
+      "loss": 0.04506875,
+      "step": 17968
+    },
+    {
+      "epoch": 35.938,
+      "grad_norm": 1.165555477142334,
+      "learning_rate": 2e-05,
+      "loss": 0.04141459,
+      "step": 17969
+    },
+    {
+      "epoch": 35.94,
+      "grad_norm": 1.2198277711868286,
+      "learning_rate": 2e-05,
+      "loss": 0.04705646,
+      "step": 17970
+    },
+    {
+      "epoch": 35.942,
+      "grad_norm": 1.2274084091186523,
+      "learning_rate": 2e-05,
+      "loss": 0.04948511,
+      "step": 17971
+    },
+    {
+      "epoch": 35.944,
+      "grad_norm": 1.474108338356018,
+      "learning_rate": 2e-05,
+      "loss": 0.04234532,
+      "step": 17972
+    },
+    {
+      "epoch": 35.946,
+      "grad_norm": 1.1247254610061646,
+      "learning_rate": 2e-05,
+      "loss": 0.04175204,
+      "step": 17973
+    },
+    {
+      "epoch": 35.948,
+      "grad_norm": 1.2691006660461426,
+      "learning_rate": 2e-05,
+      "loss": 0.05016943,
+      "step": 17974
+    },
+    {
+      "epoch": 35.95,
+      "grad_norm": 1.319199800491333,
+      "learning_rate": 2e-05,
+      "loss": 0.05450393,
+      "step": 17975
+    },
+    {
+      "epoch": 35.952,
+      "grad_norm": 1.6631337404251099,
+      "learning_rate": 2e-05,
+      "loss": 0.05106764,
+      "step": 17976
+    },
+    {
+      "epoch": 35.954,
+      "grad_norm": 1.2125028371810913,
+      "learning_rate": 2e-05,
+      "loss": 0.05636699,
+      "step": 17977
+    },
+    {
+      "epoch": 35.956,
+      "grad_norm": 1.0986932516098022,
+      "learning_rate": 2e-05,
+      "loss": 0.04804093,
+      "step": 17978
+    },
+    {
+      "epoch": 35.958,
+      "grad_norm": 2.018481731414795,
+      "learning_rate": 2e-05,
+      "loss": 0.04954034,
+      "step": 17979
+    },
+    {
+      "epoch": 35.96,
+      "grad_norm": 1.1213892698287964,
+      "learning_rate": 2e-05,
+      "loss": 0.03687691,
+      "step": 17980
+    },
+    {
+      "epoch": 35.962,
+      "grad_norm": 1.2770687341690063,
+      "learning_rate": 2e-05,
+      "loss": 0.04670438,
+      "step": 17981
+    },
+    {
+      "epoch": 35.964,
+      "grad_norm": 1.8598698377609253,
+      "learning_rate": 2e-05,
+      "loss": 0.04108175,
+      "step": 17982
+    },
+    {
+      "epoch": 35.966,
+      "grad_norm": 0.9518316388130188,
+      "learning_rate": 2e-05,
+      "loss": 0.03023814,
+      "step": 17983
+    },
+    {
+      "epoch": 35.968,
+      "grad_norm": 1.3764249086380005,
+      "learning_rate": 2e-05,
+      "loss": 0.03840397,
+      "step": 17984
+    },
+    {
+      "epoch": 35.97,
+      "grad_norm": 0.9351934194564819,
+      "learning_rate": 2e-05,
+      "loss": 0.03416009,
+      "step": 17985
+    },
+    {
+      "epoch": 35.972,
+      "grad_norm": 1.1933887004852295,
+      "learning_rate": 2e-05,
+      "loss": 0.03491713,
+      "step": 17986
+    },
+    {
+      "epoch": 35.974,
+      "grad_norm": 1.3929616212844849,
+      "learning_rate": 2e-05,
+      "loss": 0.05122756,
+      "step": 17987
+    },
+    {
+      "epoch": 35.976,
+      "grad_norm": 1.3909257650375366,
+      "learning_rate": 2e-05,
+      "loss": 0.04380251,
+      "step": 17988
+    },
+    {
+      "epoch": 35.978,
+      "grad_norm": 1.3311553001403809,
+      "learning_rate": 2e-05,
+      "loss": 0.04476759,
+      "step": 17989
+    },
+    {
+      "epoch": 35.98,
+      "grad_norm": 1.3036085367202759,
+      "learning_rate": 2e-05,
+      "loss": 0.05047145,
+      "step": 17990
+    },
+    {
+      "epoch": 35.982,
+      "grad_norm": 1.5605629682540894,
+      "learning_rate": 2e-05,
+      "loss": 0.04389017,
+      "step": 17991
+    },
+    {
+      "epoch": 35.984,
+      "grad_norm": 1.2548414468765259,
+      "learning_rate": 2e-05,
+      "loss": 0.03170571,
+      "step": 17992
+    },
+    {
+      "epoch": 35.986,
+      "grad_norm": 1.2417798042297363,
+      "learning_rate": 2e-05,
+      "loss": 0.04844061,
+      "step": 17993
+    },
+    {
+      "epoch": 35.988,
+      "grad_norm": 1.1286331415176392,
+      "learning_rate": 2e-05,
+      "loss": 0.04077397,
+      "step": 17994
+    },
+    {
+      "epoch": 35.99,
+      "grad_norm": 1.581292748451233,
+      "learning_rate": 2e-05,
+      "loss": 0.03007511,
+      "step": 17995
+    },
+    {
+      "epoch": 35.992,
+      "grad_norm": 1.491782784461975,
+      "learning_rate": 2e-05,
+      "loss": 0.03442695,
+      "step": 17996
+    },
+    {
+      "epoch": 35.994,
+      "grad_norm": 1.134804368019104,
+      "learning_rate": 2e-05,
+      "loss": 0.03684225,
+      "step": 17997
+    },
+    {
+      "epoch": 35.996,
+      "grad_norm": 2.2671315670013428,
+      "learning_rate": 2e-05,
+      "loss": 0.04276287,
+      "step": 17998
+    },
+    {
+      "epoch": 35.998,
+      "grad_norm": 1.1745227575302124,
+      "learning_rate": 2e-05,
+      "loss": 0.047242,
+      "step": 17999
+    },
+    {
+      "epoch": 36.0,
+      "grad_norm": 1.1241718530654907,
+      "learning_rate": 2e-05,
+      "loss": 0.0284595,
+      "step": 18000
+    },
+    {
+      "epoch": 36.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9820359281437125,
+        "Equal_1": 0.998,
+        "Equal_2": 0.9800399201596807,
+        "Equal_3": 0.9900199600798403,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.998003992015968,
+        "Parallel_1": 0.9959919839679359,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 0.994,
+        "Perpendicular_2": 0.994,
+        "Perpendicular_3": 0.8837675350701403,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 1.0,
+        "PointLiesOnCircle_3": 0.996,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9880239520958084
+      },
+      "eval_runtime": 319.9643,
+      "eval_samples_per_second": 32.816,
+      "eval_steps_per_second": 0.656,
+      "step": 18000
+    },
+    {
+      "epoch": 36.002,
+      "grad_norm": 1.679438829421997,
+      "learning_rate": 2e-05,
+      "loss": 0.06195679,
+      "step": 18001
+    },
+    {
+      "epoch": 36.004,
+      "grad_norm": 1.7376255989074707,
+      "learning_rate": 2e-05,
+      "loss": 0.06136629,
+      "step": 18002
+    },
+    {
+      "epoch": 36.006,
+      "grad_norm": 1.0917325019836426,
+      "learning_rate": 2e-05,
+      "loss": 0.04106084,
+      "step": 18003
+    },
+    {
+      "epoch": 36.008,
+      "grad_norm": 1.2398790121078491,
+      "learning_rate": 2e-05,
+      "loss": 0.05541268,
+      "step": 18004
+    },
+    {
+      "epoch": 36.01,
+      "grad_norm": 1.5430097579956055,
+      "learning_rate": 2e-05,
+      "loss": 0.04105628,
+      "step": 18005
+    },
+    {
+      "epoch": 36.012,
+      "grad_norm": 1.0548675060272217,
+      "learning_rate": 2e-05,
+      "loss": 0.03720319,
+      "step": 18006
+    },
+    {
+      "epoch": 36.014,
+      "grad_norm": 1.1223161220550537,
+      "learning_rate": 2e-05,
+      "loss": 0.03772051,
+      "step": 18007
+    },
+    {
+      "epoch": 36.016,
+      "grad_norm": 1.4329164028167725,
+      "learning_rate": 2e-05,
+      "loss": 0.04604128,
+      "step": 18008
+    },
+    {
+      "epoch": 36.018,
+      "grad_norm": 2.9451615810394287,
+      "learning_rate": 2e-05,
+      "loss": 0.0591033,
+      "step": 18009
+    },
+    {
+      "epoch": 36.02,
+      "grad_norm": 1.7627770900726318,
+      "learning_rate": 2e-05,
+      "loss": 0.05250013,
+      "step": 18010
+    },
+    {
+      "epoch": 36.022,
+      "grad_norm": 0.9771580696105957,
+      "learning_rate": 2e-05,
+      "loss": 0.04982607,
+      "step": 18011
+    },
+    {
+      "epoch": 36.024,
+      "grad_norm": 1.4142186641693115,
+      "learning_rate": 2e-05,
+      "loss": 0.05801858,
+      "step": 18012
+    },
+    {
+      "epoch": 36.026,
+      "grad_norm": 0.9408489465713501,
+      "learning_rate": 2e-05,
+      "loss": 0.0294435,
+      "step": 18013
+    },
+    {
+      "epoch": 36.028,
+      "grad_norm": 1.266186237335205,
+      "learning_rate": 2e-05,
+      "loss": 0.04173048,
+      "step": 18014
+    },
+    {
+      "epoch": 36.03,
+      "grad_norm": 1.2469063997268677,
+      "learning_rate": 2e-05,
+      "loss": 0.05035538,
+      "step": 18015
+    },
+    {
+      "epoch": 36.032,
+      "grad_norm": 0.9710808396339417,
+      "learning_rate": 2e-05,
+      "loss": 0.03061551,
+      "step": 18016
+    },
+    {
+      "epoch": 36.034,
+      "grad_norm": 1.1993094682693481,
+      "learning_rate": 2e-05,
+      "loss": 0.0386519,
+      "step": 18017
+    },
+    {
+      "epoch": 36.036,
+      "grad_norm": 1.0367350578308105,
+      "learning_rate": 2e-05,
+      "loss": 0.04080002,
+      "step": 18018
+    },
+    {
+      "epoch": 36.038,
+      "grad_norm": 1.0484521389007568,
+      "learning_rate": 2e-05,
+      "loss": 0.03632751,
+      "step": 18019
+    },
+    {
+      "epoch": 36.04,
+      "grad_norm": 1.096776008605957,
+      "learning_rate": 2e-05,
+      "loss": 0.03241234,
+      "step": 18020
+    },
+    {
+      "epoch": 36.042,
+      "grad_norm": 1.2367771863937378,
+      "learning_rate": 2e-05,
+      "loss": 0.04970008,
+      "step": 18021
+    },
+    {
+      "epoch": 36.044,
+      "grad_norm": 1.095942735671997,
+      "learning_rate": 2e-05,
+      "loss": 0.03803166,
+      "step": 18022
+    },
+    {
+      "epoch": 36.046,
+      "grad_norm": 1.2419805526733398,
+      "learning_rate": 2e-05,
+      "loss": 0.03578827,
+      "step": 18023
+    },
+    {
+      "epoch": 36.048,
+      "grad_norm": 0.9755739569664001,
+      "learning_rate": 2e-05,
+      "loss": 0.03225515,
+      "step": 18024
+    },
+    {
+      "epoch": 36.05,
+      "grad_norm": 0.9210935235023499,
+      "learning_rate": 2e-05,
+      "loss": 0.03545808,
+      "step": 18025
+    },
+    {
+      "epoch": 36.052,
+      "grad_norm": 1.514643669128418,
+      "learning_rate": 2e-05,
+      "loss": 0.04578691,
+      "step": 18026
+    },
+    {
+      "epoch": 36.054,
+      "grad_norm": 1.3199200630187988,
+      "learning_rate": 2e-05,
+      "loss": 0.04699472,
+      "step": 18027
+    },
+    {
+      "epoch": 36.056,
+      "grad_norm": 1.3400589227676392,
+      "learning_rate": 2e-05,
+      "loss": 0.04195932,
+      "step": 18028
+    },
+    {
+      "epoch": 36.058,
+      "grad_norm": 1.3302844762802124,
+      "learning_rate": 2e-05,
+      "loss": 0.05917486,
+      "step": 18029
+    },
+    {
+      "epoch": 36.06,
+      "grad_norm": 0.9098374247550964,
+      "learning_rate": 2e-05,
+      "loss": 0.02756627,
+      "step": 18030
+    },
+    {
+      "epoch": 36.062,
+      "grad_norm": 0.9550777673721313,
+      "learning_rate": 2e-05,
+      "loss": 0.02522951,
+      "step": 18031
+    },
+    {
+      "epoch": 36.064,
+      "grad_norm": 1.3498256206512451,
+      "learning_rate": 2e-05,
+      "loss": 0.03331952,
+      "step": 18032
+    },
+    {
+      "epoch": 36.066,
+      "grad_norm": 1.149773359298706,
+      "learning_rate": 2e-05,
+      "loss": 0.04008397,
+      "step": 18033
+    },
+    {
+      "epoch": 36.068,
+      "grad_norm": 1.6260327100753784,
+      "learning_rate": 2e-05,
+      "loss": 0.02002457,
+      "step": 18034
+    },
+    {
+      "epoch": 36.07,
+      "grad_norm": 1.0886062383651733,
+      "learning_rate": 2e-05,
+      "loss": 0.03689205,
+      "step": 18035
+    },
+    {
+      "epoch": 36.072,
+      "grad_norm": 2.1114742755889893,
+      "learning_rate": 2e-05,
+      "loss": 0.04384237,
+      "step": 18036
+    },
+    {
+      "epoch": 36.074,
+      "grad_norm": 1.199912428855896,
+      "learning_rate": 2e-05,
+      "loss": 0.03832989,
+      "step": 18037
+    },
+    {
+      "epoch": 36.076,
+      "grad_norm": 1.1416548490524292,
+      "learning_rate": 2e-05,
+      "loss": 0.04295663,
+      "step": 18038
+    },
+    {
+      "epoch": 36.078,
+      "grad_norm": 5.6245245933532715,
+      "learning_rate": 2e-05,
+      "loss": 0.04268225,
+      "step": 18039
+    },
+    {
+      "epoch": 36.08,
+      "grad_norm": 1.2393498420715332,
+      "learning_rate": 2e-05,
+      "loss": 0.03811008,
+      "step": 18040
+    },
+    {
+      "epoch": 36.082,
+      "grad_norm": 1.0027004480361938,
+      "learning_rate": 2e-05,
+      "loss": 0.03410269,
+      "step": 18041
+    },
+    {
+      "epoch": 36.084,
+      "grad_norm": 1.7057132720947266,
+      "learning_rate": 2e-05,
+      "loss": 0.03965345,
+      "step": 18042
+    },
+    {
+      "epoch": 36.086,
+      "grad_norm": 0.9601574540138245,
+      "learning_rate": 2e-05,
+      "loss": 0.02651468,
+      "step": 18043
+    },
+    {
+      "epoch": 36.088,
+      "grad_norm": 1.2537764310836792,
+      "learning_rate": 2e-05,
+      "loss": 0.03860223,
+      "step": 18044
+    },
+    {
+      "epoch": 36.09,
+      "grad_norm": 1.3177660703659058,
+      "learning_rate": 2e-05,
+      "loss": 0.05462038,
+      "step": 18045
+    },
+    {
+      "epoch": 36.092,
+      "grad_norm": 1.2594581842422485,
+      "learning_rate": 2e-05,
+      "loss": 0.04445657,
+      "step": 18046
+    },
+    {
+      "epoch": 36.094,
+      "grad_norm": 1.132057547569275,
+      "learning_rate": 2e-05,
+      "loss": 0.05026545,
+      "step": 18047
+    },
+    {
+      "epoch": 36.096,
+      "grad_norm": 1.5587725639343262,
+      "learning_rate": 2e-05,
+      "loss": 0.049737,
+      "step": 18048
+    },
+    {
+      "epoch": 36.098,
+      "grad_norm": 1.235729694366455,
+      "learning_rate": 2e-05,
+      "loss": 0.04665924,
+      "step": 18049
+    },
+    {
+      "epoch": 36.1,
+      "grad_norm": 1.4784846305847168,
+      "learning_rate": 2e-05,
+      "loss": 0.04161521,
+      "step": 18050
+    },
+    {
+      "epoch": 36.102,
+      "grad_norm": 1.898224115371704,
+      "learning_rate": 2e-05,
+      "loss": 0.05342315,
+      "step": 18051
+    },
+    {
+      "epoch": 36.104,
+      "grad_norm": 1.1921885013580322,
+      "learning_rate": 2e-05,
+      "loss": 0.03857575,
+      "step": 18052
+    },
+    {
+      "epoch": 36.106,
+      "grad_norm": 1.1267322301864624,
+      "learning_rate": 2e-05,
+      "loss": 0.04110962,
+      "step": 18053
+    },
+    {
+      "epoch": 36.108,
+      "grad_norm": 1.1211832761764526,
+      "learning_rate": 2e-05,
+      "loss": 0.04066541,
+      "step": 18054
+    },
+    {
+      "epoch": 36.11,
+      "grad_norm": 1.1396983861923218,
+      "learning_rate": 2e-05,
+      "loss": 0.03554956,
+      "step": 18055
+    },
+    {
+      "epoch": 36.112,
+      "grad_norm": 4.26861572265625,
+      "learning_rate": 2e-05,
+      "loss": 0.04480337,
+      "step": 18056
+    },
+    {
+      "epoch": 36.114,
+      "grad_norm": 1.3064028024673462,
+      "learning_rate": 2e-05,
+      "loss": 0.05451906,
+      "step": 18057
+    },
+    {
+      "epoch": 36.116,
+      "grad_norm": 1.1184580326080322,
+      "learning_rate": 2e-05,
+      "loss": 0.04466115,
+      "step": 18058
+    },
+    {
+      "epoch": 36.118,
+      "grad_norm": 2.734436273574829,
+      "learning_rate": 2e-05,
+      "loss": 0.04473677,
+      "step": 18059
+    },
+    {
+      "epoch": 36.12,
+      "grad_norm": 1.1926535367965698,
+      "learning_rate": 2e-05,
+      "loss": 0.04510562,
+      "step": 18060
+    },
+    {
+      "epoch": 36.122,
+      "grad_norm": 2.227550506591797,
+      "learning_rate": 2e-05,
+      "loss": 0.03720244,
+      "step": 18061
+    },
+    {
+      "epoch": 36.124,
+      "grad_norm": 1.2409284114837646,
+      "learning_rate": 2e-05,
+      "loss": 0.04574725,
+      "step": 18062
+    },
+    {
+      "epoch": 36.126,
+      "grad_norm": 1.0186352729797363,
+      "learning_rate": 2e-05,
+      "loss": 0.03966548,
+      "step": 18063
+    },
+    {
+      "epoch": 36.128,
+      "grad_norm": 1.098799467086792,
+      "learning_rate": 2e-05,
+      "loss": 0.04066562,
+      "step": 18064
+    },
+    {
+      "epoch": 36.13,
+      "grad_norm": 0.9615515470504761,
+      "learning_rate": 2e-05,
+      "loss": 0.02897996,
+      "step": 18065
+    },
+    {
+      "epoch": 36.132,
+      "grad_norm": 2.371126413345337,
+      "learning_rate": 2e-05,
+      "loss": 0.06829099,
+      "step": 18066
+    },
+    {
+      "epoch": 36.134,
+      "grad_norm": 1.4561325311660767,
+      "learning_rate": 2e-05,
+      "loss": 0.04208488,
+      "step": 18067
+    },
+    {
+      "epoch": 36.136,
+      "grad_norm": 1.1017025709152222,
+      "learning_rate": 2e-05,
+      "loss": 0.04764595,
+      "step": 18068
+    },
+    {
+      "epoch": 36.138,
+      "grad_norm": 1.6677707433700562,
+      "learning_rate": 2e-05,
+      "loss": 0.03898593,
+      "step": 18069
+    },
+    {
+      "epoch": 36.14,
+      "grad_norm": 1.2435901165008545,
+      "learning_rate": 2e-05,
+      "loss": 0.03861672,
+      "step": 18070
+    },
+    {
+      "epoch": 36.142,
+      "grad_norm": 1.3686686754226685,
+      "learning_rate": 2e-05,
+      "loss": 0.05133841,
+      "step": 18071
+    },
+    {
+      "epoch": 36.144,
+      "grad_norm": 1.1071763038635254,
+      "learning_rate": 2e-05,
+      "loss": 0.04213175,
+      "step": 18072
+    },
+    {
+      "epoch": 36.146,
+      "grad_norm": 1.1424134969711304,
+      "learning_rate": 2e-05,
+      "loss": 0.04421389,
+      "step": 18073
+    },
+    {
+      "epoch": 36.148,
+      "grad_norm": 1.2619765996932983,
+      "learning_rate": 2e-05,
+      "loss": 0.05113876,
+      "step": 18074
+    },
+    {
+      "epoch": 36.15,
+      "grad_norm": 2.3322865962982178,
+      "learning_rate": 2e-05,
+      "loss": 0.05380969,
+      "step": 18075
+    },
+    {
+      "epoch": 36.152,
+      "grad_norm": 1.5360093116760254,
+      "learning_rate": 2e-05,
+      "loss": 0.03704157,
+      "step": 18076
+    },
+    {
+      "epoch": 36.154,
+      "grad_norm": 2.7045676708221436,
+      "learning_rate": 2e-05,
+      "loss": 0.03724046,
+      "step": 18077
+    },
+    {
+      "epoch": 36.156,
+      "grad_norm": 1.778171181678772,
+      "learning_rate": 2e-05,
+      "loss": 0.06204282,
+      "step": 18078
+    },
+    {
+      "epoch": 36.158,
+      "grad_norm": 2.811089277267456,
+      "learning_rate": 2e-05,
+      "loss": 0.04121366,
+      "step": 18079
+    },
+    {
+      "epoch": 36.16,
+      "grad_norm": 1.1129862070083618,
+      "learning_rate": 2e-05,
+      "loss": 0.03516244,
+      "step": 18080
+    },
+    {
+      "epoch": 36.162,
+      "grad_norm": 1.1569979190826416,
+      "learning_rate": 2e-05,
+      "loss": 0.04278459,
+      "step": 18081
+    },
+    {
+      "epoch": 36.164,
+      "grad_norm": 1.064875841140747,
+      "learning_rate": 2e-05,
+      "loss": 0.03698279,
+      "step": 18082
+    },
+    {
+      "epoch": 36.166,
+      "grad_norm": 1.0886690616607666,
+      "learning_rate": 2e-05,
+      "loss": 0.04283421,
+      "step": 18083
+    },
+    {
+      "epoch": 36.168,
+      "grad_norm": 3.398261785507202,
+      "learning_rate": 2e-05,
+      "loss": 0.03867529,
+      "step": 18084
+    },
+    {
+      "epoch": 36.17,
+      "grad_norm": 1.657079815864563,
+      "learning_rate": 2e-05,
+      "loss": 0.04585808,
+      "step": 18085
+    },
+    {
+      "epoch": 36.172,
+      "grad_norm": 1.3320543766021729,
+      "learning_rate": 2e-05,
+      "loss": 0.04661385,
+      "step": 18086
+    },
+    {
+      "epoch": 36.174,
+      "grad_norm": 0.8988568782806396,
+      "learning_rate": 2e-05,
+      "loss": 0.02427765,
+      "step": 18087
+    },
+    {
+      "epoch": 36.176,
+      "grad_norm": 1.284536361694336,
+      "learning_rate": 2e-05,
+      "loss": 0.0532205,
+      "step": 18088
+    },
+    {
+      "epoch": 36.178,
+      "grad_norm": 1.1643673181533813,
+      "learning_rate": 2e-05,
+      "loss": 0.03632015,
+      "step": 18089
+    },
+    {
+      "epoch": 36.18,
+      "grad_norm": 1.3021336793899536,
+      "learning_rate": 2e-05,
+      "loss": 0.05177925,
+      "step": 18090
+    },
+    {
+      "epoch": 36.182,
+      "grad_norm": 2.0210723876953125,
+      "learning_rate": 2e-05,
+      "loss": 0.04610447,
+      "step": 18091
+    },
+    {
+      "epoch": 36.184,
+      "grad_norm": 0.9699888229370117,
+      "learning_rate": 2e-05,
+      "loss": 0.0398217,
+      "step": 18092
+    },
+    {
+      "epoch": 36.186,
+      "grad_norm": 1.0085978507995605,
+      "learning_rate": 2e-05,
+      "loss": 0.03745851,
+      "step": 18093
+    },
+    {
+      "epoch": 36.188,
+      "grad_norm": 1.3533179759979248,
+      "learning_rate": 2e-05,
+      "loss": 0.05687679,
+      "step": 18094
+    },
+    {
+      "epoch": 36.19,
+      "grad_norm": 1.0797785520553589,
+      "learning_rate": 2e-05,
+      "loss": 0.04193022,
+      "step": 18095
+    },
+    {
+      "epoch": 36.192,
+      "grad_norm": 1.4561572074890137,
+      "learning_rate": 2e-05,
+      "loss": 0.03306325,
+      "step": 18096
+    },
+    {
+      "epoch": 36.194,
+      "grad_norm": 1.2475993633270264,
+      "learning_rate": 2e-05,
+      "loss": 0.03667925,
+      "step": 18097
+    },
+    {
+      "epoch": 36.196,
+      "grad_norm": 0.9133176803588867,
+      "learning_rate": 2e-05,
+      "loss": 0.02343579,
+      "step": 18098
+    },
+    {
+      "epoch": 36.198,
+      "grad_norm": 1.6272944211959839,
+      "learning_rate": 2e-05,
+      "loss": 0.04554627,
+      "step": 18099
+    },
+    {
+      "epoch": 36.2,
+      "grad_norm": 1.5745600461959839,
+      "learning_rate": 2e-05,
+      "loss": 0.04788917,
+      "step": 18100
+    },
+    {
+      "epoch": 36.202,
+      "grad_norm": 1.2428606748580933,
+      "learning_rate": 2e-05,
+      "loss": 0.04026973,
+      "step": 18101
+    },
+    {
+      "epoch": 36.204,
+      "grad_norm": 1.6128360033035278,
+      "learning_rate": 2e-05,
+      "loss": 0.05024995,
+      "step": 18102
+    },
+    {
+      "epoch": 36.206,
+      "grad_norm": 1.0230610370635986,
+      "learning_rate": 2e-05,
+      "loss": 0.03382932,
+      "step": 18103
+    },
+    {
+      "epoch": 36.208,
+      "grad_norm": 1.075156331062317,
+      "learning_rate": 2e-05,
+      "loss": 0.04463224,
+      "step": 18104
+    },
+    {
+      "epoch": 36.21,
+      "grad_norm": 3.846419095993042,
+      "learning_rate": 2e-05,
+      "loss": 0.05278007,
+      "step": 18105
+    },
+    {
+      "epoch": 36.212,
+      "grad_norm": 0.9034450054168701,
+      "learning_rate": 2e-05,
+      "loss": 0.03557969,
+      "step": 18106
+    },
+    {
+      "epoch": 36.214,
+      "grad_norm": 0.923078179359436,
+      "learning_rate": 2e-05,
+      "loss": 0.03522691,
+      "step": 18107
+    },
+    {
+      "epoch": 36.216,
+      "grad_norm": 1.3479729890823364,
+      "learning_rate": 2e-05,
+      "loss": 0.0439695,
+      "step": 18108
+    },
+    {
+      "epoch": 36.218,
+      "grad_norm": 2.1420326232910156,
+      "learning_rate": 2e-05,
+      "loss": 0.06162531,
+      "step": 18109
+    },
+    {
+      "epoch": 36.22,
+      "grad_norm": 1.184019684791565,
+      "learning_rate": 2e-05,
+      "loss": 0.04064212,
+      "step": 18110
+    },
+    {
+      "epoch": 36.222,
+      "grad_norm": 1.0127112865447998,
+      "learning_rate": 2e-05,
+      "loss": 0.03657559,
+      "step": 18111
+    },
+    {
+      "epoch": 36.224,
+      "grad_norm": 0.9742773175239563,
+      "learning_rate": 2e-05,
+      "loss": 0.03809122,
+      "step": 18112
+    },
+    {
+      "epoch": 36.226,
+      "grad_norm": 1.1490448713302612,
+      "learning_rate": 2e-05,
+      "loss": 0.04694427,
+      "step": 18113
+    },
+    {
+      "epoch": 36.228,
+      "grad_norm": 0.9744361042976379,
+      "learning_rate": 2e-05,
+      "loss": 0.02423439,
+      "step": 18114
+    },
+    {
+      "epoch": 36.23,
+      "grad_norm": 1.1082568168640137,
+      "learning_rate": 2e-05,
+      "loss": 0.04402753,
+      "step": 18115
+    },
+    {
+      "epoch": 36.232,
+      "grad_norm": 1.1237958669662476,
+      "learning_rate": 2e-05,
+      "loss": 0.03403135,
+      "step": 18116
+    },
+    {
+      "epoch": 36.234,
+      "grad_norm": 1.0820505619049072,
+      "learning_rate": 2e-05,
+      "loss": 0.03654093,
+      "step": 18117
+    },
+    {
+      "epoch": 36.236,
+      "grad_norm": 1.0331096649169922,
+      "learning_rate": 2e-05,
+      "loss": 0.03929486,
+      "step": 18118
+    },
+    {
+      "epoch": 36.238,
+      "grad_norm": 1.2376075983047485,
+      "learning_rate": 2e-05,
+      "loss": 0.03313967,
+      "step": 18119
+    },
+    {
+      "epoch": 36.24,
+      "grad_norm": 1.596076250076294,
+      "learning_rate": 2e-05,
+      "loss": 0.04521656,
+      "step": 18120
+    },
+    {
+      "epoch": 36.242,
+      "grad_norm": 1.2083992958068848,
+      "learning_rate": 2e-05,
+      "loss": 0.04801769,
+      "step": 18121
+    },
+    {
+      "epoch": 36.244,
+      "grad_norm": 1.3693628311157227,
+      "learning_rate": 2e-05,
+      "loss": 0.05652841,
+      "step": 18122
+    },
+    {
+      "epoch": 36.246,
+      "grad_norm": 1.8372113704681396,
+      "learning_rate": 2e-05,
+      "loss": 0.03900075,
+      "step": 18123
+    },
+    {
+      "epoch": 36.248,
+      "grad_norm": 1.6199109554290771,
+      "learning_rate": 2e-05,
+      "loss": 0.05532604,
+      "step": 18124
+    },
+    {
+      "epoch": 36.25,
+      "grad_norm": 1.1799906492233276,
+      "learning_rate": 2e-05,
+      "loss": 0.03721724,
+      "step": 18125
+    },
+    {
+      "epoch": 36.252,
+      "grad_norm": 1.0089119672775269,
+      "learning_rate": 2e-05,
+      "loss": 0.03412523,
+      "step": 18126
+    },
+    {
+      "epoch": 36.254,
+      "grad_norm": 1.3146682977676392,
+      "learning_rate": 2e-05,
+      "loss": 0.03678686,
+      "step": 18127
+    },
+    {
+      "epoch": 36.256,
+      "grad_norm": 1.4293696880340576,
+      "learning_rate": 2e-05,
+      "loss": 0.05230099,
+      "step": 18128
+    },
+    {
+      "epoch": 36.258,
+      "grad_norm": 1.1753870248794556,
+      "learning_rate": 2e-05,
+      "loss": 0.04117392,
+      "step": 18129
+    },
+    {
+      "epoch": 36.26,
+      "grad_norm": 1.2926392555236816,
+      "learning_rate": 2e-05,
+      "loss": 0.04781886,
+      "step": 18130
+    },
+    {
+      "epoch": 36.262,
+      "grad_norm": 1.7987608909606934,
+      "learning_rate": 2e-05,
+      "loss": 0.05109328,
+      "step": 18131
+    },
+    {
+      "epoch": 36.264,
+      "grad_norm": 1.0095752477645874,
+      "learning_rate": 2e-05,
+      "loss": 0.04271962,
+      "step": 18132
+    },
+    {
+      "epoch": 36.266,
+      "grad_norm": 0.901914656162262,
+      "learning_rate": 2e-05,
+      "loss": 0.02623034,
+      "step": 18133
+    },
+    {
+      "epoch": 36.268,
+      "grad_norm": 1.1711057424545288,
+      "learning_rate": 2e-05,
+      "loss": 0.03598663,
+      "step": 18134
+    },
+    {
+      "epoch": 36.27,
+      "grad_norm": 0.9837378859519958,
+      "learning_rate": 2e-05,
+      "loss": 0.03688063,
+      "step": 18135
+    },
+    {
+      "epoch": 36.272,
+      "grad_norm": 1.2390130758285522,
+      "learning_rate": 2e-05,
+      "loss": 0.05092961,
+      "step": 18136
+    },
+    {
+      "epoch": 36.274,
+      "grad_norm": 1.270090103149414,
+      "learning_rate": 2e-05,
+      "loss": 0.0404551,
+      "step": 18137
+    },
+    {
+      "epoch": 36.276,
+      "grad_norm": 1.5756562948226929,
+      "learning_rate": 2e-05,
+      "loss": 0.04972642,
+      "step": 18138
+    },
+    {
+      "epoch": 36.278,
+      "grad_norm": 0.9940088987350464,
+      "learning_rate": 2e-05,
+      "loss": 0.03693637,
+      "step": 18139
+    },
+    {
+      "epoch": 36.28,
+      "grad_norm": 0.9043747782707214,
+      "learning_rate": 2e-05,
+      "loss": 0.02635769,
+      "step": 18140
+    },
+    {
+      "epoch": 36.282,
+      "grad_norm": 1.2040432691574097,
+      "learning_rate": 2e-05,
+      "loss": 0.04434269,
+      "step": 18141
+    },
+    {
+      "epoch": 36.284,
+      "grad_norm": 1.0949373245239258,
+      "learning_rate": 2e-05,
+      "loss": 0.03474148,
+      "step": 18142
+    },
+    {
+      "epoch": 36.286,
+      "grad_norm": 1.09050452709198,
+      "learning_rate": 2e-05,
+      "loss": 0.04845259,
+      "step": 18143
+    },
+    {
+      "epoch": 36.288,
+      "grad_norm": 1.3407831192016602,
+      "learning_rate": 2e-05,
+      "loss": 0.04304389,
+      "step": 18144
+    },
+    {
+      "epoch": 36.29,
+      "grad_norm": 1.1936886310577393,
+      "learning_rate": 2e-05,
+      "loss": 0.04481979,
+      "step": 18145
+    },
+    {
+      "epoch": 36.292,
+      "grad_norm": 2.4183106422424316,
+      "learning_rate": 2e-05,
+      "loss": 0.05331695,
+      "step": 18146
+    },
+    {
+      "epoch": 36.294,
+      "grad_norm": 1.1295486688613892,
+      "learning_rate": 2e-05,
+      "loss": 0.03321164,
+      "step": 18147
+    },
+    {
+      "epoch": 36.296,
+      "grad_norm": 1.360809087753296,
+      "learning_rate": 2e-05,
+      "loss": 0.03916159,
+      "step": 18148
+    },
+    {
+      "epoch": 36.298,
+      "grad_norm": 1.247570276260376,
+      "learning_rate": 2e-05,
+      "loss": 0.0428711,
+      "step": 18149
+    },
+    {
+      "epoch": 36.3,
+      "grad_norm": 1.036298394203186,
+      "learning_rate": 2e-05,
+      "loss": 0.04656066,
+      "step": 18150
+    },
+    {
+      "epoch": 36.302,
+      "grad_norm": 0.9646769165992737,
+      "learning_rate": 2e-05,
+      "loss": 0.03399408,
+      "step": 18151
+    },
+    {
+      "epoch": 36.304,
+      "grad_norm": 1.164666771888733,
+      "learning_rate": 2e-05,
+      "loss": 0.04837671,
+      "step": 18152
+    },
+    {
+      "epoch": 36.306,
+      "grad_norm": 1.2248945236206055,
+      "learning_rate": 2e-05,
+      "loss": 0.0307962,
+      "step": 18153
+    },
+    {
+      "epoch": 36.308,
+      "grad_norm": 1.1758040189743042,
+      "learning_rate": 2e-05,
+      "loss": 0.0404579,
+      "step": 18154
+    },
+    {
+      "epoch": 36.31,
+      "grad_norm": 1.0307793617248535,
+      "learning_rate": 2e-05,
+      "loss": 0.0290431,
+      "step": 18155
+    },
+    {
+      "epoch": 36.312,
+      "grad_norm": 0.920478880405426,
+      "learning_rate": 2e-05,
+      "loss": 0.03528172,
+      "step": 18156
+    },
+    {
+      "epoch": 36.314,
+      "grad_norm": 1.0334807634353638,
+      "learning_rate": 2e-05,
+      "loss": 0.05046558,
+      "step": 18157
+    },
+    {
+      "epoch": 36.316,
+      "grad_norm": 0.9907777309417725,
+      "learning_rate": 2e-05,
+      "loss": 0.03023825,
+      "step": 18158
+    },
+    {
+      "epoch": 36.318,
+      "grad_norm": 1.227904200553894,
+      "learning_rate": 2e-05,
+      "loss": 0.04913838,
+      "step": 18159
+    },
+    {
+      "epoch": 36.32,
+      "grad_norm": 1.2009419202804565,
+      "learning_rate": 2e-05,
+      "loss": 0.04932972,
+      "step": 18160
+    },
+    {
+      "epoch": 36.322,
+      "grad_norm": 0.9665030241012573,
+      "learning_rate": 2e-05,
+      "loss": 0.03171739,
+      "step": 18161
+    },
+    {
+      "epoch": 36.324,
+      "grad_norm": 1.2279425859451294,
+      "learning_rate": 2e-05,
+      "loss": 0.0608729,
+      "step": 18162
+    },
+    {
+      "epoch": 36.326,
+      "grad_norm": 0.9468924403190613,
+      "learning_rate": 2e-05,
+      "loss": 0.03156189,
+      "step": 18163
+    },
+    {
+      "epoch": 36.328,
+      "grad_norm": 1.0261448621749878,
+      "learning_rate": 2e-05,
+      "loss": 0.03610597,
+      "step": 18164
+    },
+    {
+      "epoch": 36.33,
+      "grad_norm": 1.276781678199768,
+      "learning_rate": 2e-05,
+      "loss": 0.03549629,
+      "step": 18165
+    },
+    {
+      "epoch": 36.332,
+      "grad_norm": 0.9759575724601746,
+      "learning_rate": 2e-05,
+      "loss": 0.03994817,
+      "step": 18166
+    },
+    {
+      "epoch": 36.334,
+      "grad_norm": 1.7276750802993774,
+      "learning_rate": 2e-05,
+      "loss": 0.04654774,
+      "step": 18167
+    },
+    {
+      "epoch": 36.336,
+      "grad_norm": 0.9664547443389893,
+      "learning_rate": 2e-05,
+      "loss": 0.03177688,
+      "step": 18168
+    },
+    {
+      "epoch": 36.338,
+      "grad_norm": 1.9670281410217285,
+      "learning_rate": 2e-05,
+      "loss": 0.04651101,
+      "step": 18169
+    },
+    {
+      "epoch": 36.34,
+      "grad_norm": 1.802821159362793,
+      "learning_rate": 2e-05,
+      "loss": 0.04297075,
+      "step": 18170
+    },
+    {
+      "epoch": 36.342,
+      "grad_norm": 1.661774754524231,
+      "learning_rate": 2e-05,
+      "loss": 0.06117053,
+      "step": 18171
+    },
+    {
+      "epoch": 36.344,
+      "grad_norm": 2.3734986782073975,
+      "learning_rate": 2e-05,
+      "loss": 0.05952629,
+      "step": 18172
+    },
+    {
+      "epoch": 36.346,
+      "grad_norm": 1.2025349140167236,
+      "learning_rate": 2e-05,
+      "loss": 0.04102109,
+      "step": 18173
+    },
+    {
+      "epoch": 36.348,
+      "grad_norm": 1.0982835292816162,
+      "learning_rate": 2e-05,
+      "loss": 0.04104698,
+      "step": 18174
+    },
+    {
+      "epoch": 36.35,
+      "grad_norm": 1.157949686050415,
+      "learning_rate": 2e-05,
+      "loss": 0.03813836,
+      "step": 18175
+    },
+    {
+      "epoch": 36.352,
+      "grad_norm": 1.055647611618042,
+      "learning_rate": 2e-05,
+      "loss": 0.03598933,
+      "step": 18176
+    },
+    {
+      "epoch": 36.354,
+      "grad_norm": 1.3801038265228271,
+      "learning_rate": 2e-05,
+      "loss": 0.04944832,
+      "step": 18177
+    },
+    {
+      "epoch": 36.356,
+      "grad_norm": 1.322588324546814,
+      "learning_rate": 2e-05,
+      "loss": 0.03539138,
+      "step": 18178
+    },
+    {
+      "epoch": 36.358,
+      "grad_norm": 1.4064815044403076,
+      "learning_rate": 2e-05,
+      "loss": 0.05578845,
+      "step": 18179
+    },
+    {
+      "epoch": 36.36,
+      "grad_norm": 3.2461321353912354,
+      "learning_rate": 2e-05,
+      "loss": 0.04246819,
+      "step": 18180
+    },
+    {
+      "epoch": 36.362,
+      "grad_norm": 1.1064374446868896,
+      "learning_rate": 2e-05,
+      "loss": 0.03681324,
+      "step": 18181
+    },
+    {
+      "epoch": 36.364,
+      "grad_norm": 1.1665987968444824,
+      "learning_rate": 2e-05,
+      "loss": 0.04058225,
+      "step": 18182
+    },
+    {
+      "epoch": 36.366,
+      "grad_norm": 1.029982089996338,
+      "learning_rate": 2e-05,
+      "loss": 0.03287161,
+      "step": 18183
+    },
+    {
+      "epoch": 36.368,
+      "grad_norm": 1.1927478313446045,
+      "learning_rate": 2e-05,
+      "loss": 0.05367187,
+      "step": 18184
+    },
+    {
+      "epoch": 36.37,
+      "grad_norm": 1.2972198724746704,
+      "learning_rate": 2e-05,
+      "loss": 0.04106817,
+      "step": 18185
+    },
+    {
+      "epoch": 36.372,
+      "grad_norm": 1.027411699295044,
+      "learning_rate": 2e-05,
+      "loss": 0.04088916,
+      "step": 18186
+    },
+    {
+      "epoch": 36.374,
+      "grad_norm": 1.171830415725708,
+      "learning_rate": 2e-05,
+      "loss": 0.05644122,
+      "step": 18187
+    },
+    {
+      "epoch": 36.376,
+      "grad_norm": 1.1561020612716675,
+      "learning_rate": 2e-05,
+      "loss": 0.04188239,
+      "step": 18188
+    },
+    {
+      "epoch": 36.378,
+      "grad_norm": 1.0463422536849976,
+      "learning_rate": 2e-05,
+      "loss": 0.04296407,
+      "step": 18189
+    },
+    {
+      "epoch": 36.38,
+      "grad_norm": 1.2229862213134766,
+      "learning_rate": 2e-05,
+      "loss": 0.03844842,
+      "step": 18190
+    },
+    {
+      "epoch": 36.382,
+      "grad_norm": 0.8768106698989868,
+      "learning_rate": 2e-05,
+      "loss": 0.03120964,
+      "step": 18191
+    },
+    {
+      "epoch": 36.384,
+      "grad_norm": 1.1767603158950806,
+      "learning_rate": 2e-05,
+      "loss": 0.03380742,
+      "step": 18192
+    },
+    {
+      "epoch": 36.386,
+      "grad_norm": 0.9635298848152161,
+      "learning_rate": 2e-05,
+      "loss": 0.03339863,
+      "step": 18193
+    },
+    {
+      "epoch": 36.388,
+      "grad_norm": 0.9768321514129639,
+      "learning_rate": 2e-05,
+      "loss": 0.03579611,
+      "step": 18194
+    },
+    {
+      "epoch": 36.39,
+      "grad_norm": 0.8538011312484741,
+      "learning_rate": 2e-05,
+      "loss": 0.02970689,
+      "step": 18195
+    },
+    {
+      "epoch": 36.392,
+      "grad_norm": 1.0878775119781494,
+      "learning_rate": 2e-05,
+      "loss": 0.04001588,
+      "step": 18196
+    },
+    {
+      "epoch": 36.394,
+      "grad_norm": 1.142221212387085,
+      "learning_rate": 2e-05,
+      "loss": 0.04551936,
+      "step": 18197
+    },
+    {
+      "epoch": 36.396,
+      "grad_norm": 1.5530418157577515,
+      "learning_rate": 2e-05,
+      "loss": 0.03043629,
+      "step": 18198
+    },
+    {
+      "epoch": 36.398,
+      "grad_norm": 1.637148141860962,
+      "learning_rate": 2e-05,
+      "loss": 0.03448223,
+      "step": 18199
+    },
+    {
+      "epoch": 36.4,
+      "grad_norm": 1.4483563899993896,
+      "learning_rate": 2e-05,
+      "loss": 0.05348013,
+      "step": 18200
+    },
+    {
+      "epoch": 36.402,
+      "grad_norm": 1.0635143518447876,
+      "learning_rate": 2e-05,
+      "loss": 0.03588112,
+      "step": 18201
+    },
+    {
+      "epoch": 36.404,
+      "grad_norm": 2.611689329147339,
+      "learning_rate": 2e-05,
+      "loss": 0.06078885,
+      "step": 18202
+    },
+    {
+      "epoch": 36.406,
+      "grad_norm": 1.502625584602356,
+      "learning_rate": 2e-05,
+      "loss": 0.03423886,
+      "step": 18203
+    },
+    {
+      "epoch": 36.408,
+      "grad_norm": 1.3298054933547974,
+      "learning_rate": 2e-05,
+      "loss": 0.04151591,
+      "step": 18204
+    },
+    {
+      "epoch": 36.41,
+      "grad_norm": 1.27322518825531,
+      "learning_rate": 2e-05,
+      "loss": 0.05073592,
+      "step": 18205
+    },
+    {
+      "epoch": 36.412,
+      "grad_norm": 1.035677433013916,
+      "learning_rate": 2e-05,
+      "loss": 0.04113561,
+      "step": 18206
+    },
+    {
+      "epoch": 36.414,
+      "grad_norm": 1.7269580364227295,
+      "learning_rate": 2e-05,
+      "loss": 0.04068903,
+      "step": 18207
+    },
+    {
+      "epoch": 36.416,
+      "grad_norm": 1.314218521118164,
+      "learning_rate": 2e-05,
+      "loss": 0.04115914,
+      "step": 18208
+    },
+    {
+      "epoch": 36.418,
+      "grad_norm": 1.3631287813186646,
+      "learning_rate": 2e-05,
+      "loss": 0.05018599,
+      "step": 18209
+    },
+    {
+      "epoch": 36.42,
+      "grad_norm": 1.4252984523773193,
+      "learning_rate": 2e-05,
+      "loss": 0.05225445,
+      "step": 18210
+    },
+    {
+      "epoch": 36.422,
+      "grad_norm": 1.0618059635162354,
+      "learning_rate": 2e-05,
+      "loss": 0.03559672,
+      "step": 18211
+    },
+    {
+      "epoch": 36.424,
+      "grad_norm": 1.217882752418518,
+      "learning_rate": 2e-05,
+      "loss": 0.0485566,
+      "step": 18212
+    },
+    {
+      "epoch": 36.426,
+      "grad_norm": 1.2114055156707764,
+      "learning_rate": 2e-05,
+      "loss": 0.04762184,
+      "step": 18213
+    },
+    {
+      "epoch": 36.428,
+      "grad_norm": 2.0187506675720215,
+      "learning_rate": 2e-05,
+      "loss": 0.05630653,
+      "step": 18214
+    },
+    {
+      "epoch": 36.43,
+      "grad_norm": 1.9597657918930054,
+      "learning_rate": 2e-05,
+      "loss": 0.06257641,
+      "step": 18215
+    },
+    {
+      "epoch": 36.432,
+      "grad_norm": 0.9971281886100769,
+      "learning_rate": 2e-05,
+      "loss": 0.03872079,
+      "step": 18216
+    },
+    {
+      "epoch": 36.434,
+      "grad_norm": 1.2572100162506104,
+      "learning_rate": 2e-05,
+      "loss": 0.03614823,
+      "step": 18217
+    },
+    {
+      "epoch": 36.436,
+      "grad_norm": 1.1719560623168945,
+      "learning_rate": 2e-05,
+      "loss": 0.03753065,
+      "step": 18218
+    },
+    {
+      "epoch": 36.438,
+      "grad_norm": 1.0278385877609253,
+      "learning_rate": 2e-05,
+      "loss": 0.03851878,
+      "step": 18219
+    },
+    {
+      "epoch": 36.44,
+      "grad_norm": 1.2253730297088623,
+      "learning_rate": 2e-05,
+      "loss": 0.0503806,
+      "step": 18220
+    },
+    {
+      "epoch": 36.442,
+      "grad_norm": 0.88956218957901,
+      "learning_rate": 2e-05,
+      "loss": 0.02707586,
+      "step": 18221
+    },
+    {
+      "epoch": 36.444,
+      "grad_norm": 0.8710851669311523,
+      "learning_rate": 2e-05,
+      "loss": 0.0222133,
+      "step": 18222
+    },
+    {
+      "epoch": 36.446,
+      "grad_norm": 1.2004845142364502,
+      "learning_rate": 2e-05,
+      "loss": 0.05466457,
+      "step": 18223
+    },
+    {
+      "epoch": 36.448,
+      "grad_norm": 1.3014144897460938,
+      "learning_rate": 2e-05,
+      "loss": 0.05169172,
+      "step": 18224
+    },
+    {
+      "epoch": 36.45,
+      "grad_norm": 1.0197980403900146,
+      "learning_rate": 2e-05,
+      "loss": 0.03822663,
+      "step": 18225
+    },
+    {
+      "epoch": 36.452,
+      "grad_norm": 1.3792431354522705,
+      "learning_rate": 2e-05,
+      "loss": 0.04293468,
+      "step": 18226
+    },
+    {
+      "epoch": 36.454,
+      "grad_norm": 1.1744886636734009,
+      "learning_rate": 2e-05,
+      "loss": 0.05211157,
+      "step": 18227
+    },
+    {
+      "epoch": 36.456,
+      "grad_norm": 1.598494052886963,
+      "learning_rate": 2e-05,
+      "loss": 0.04538838,
+      "step": 18228
+    },
+    {
+      "epoch": 36.458,
+      "grad_norm": 0.8673466444015503,
+      "learning_rate": 2e-05,
+      "loss": 0.0319301,
+      "step": 18229
+    },
+    {
+      "epoch": 36.46,
+      "grad_norm": 1.918244481086731,
+      "learning_rate": 2e-05,
+      "loss": 0.05689289,
+      "step": 18230
+    },
+    {
+      "epoch": 36.462,
+      "grad_norm": 1.1343375444412231,
+      "learning_rate": 2e-05,
+      "loss": 0.04852476,
+      "step": 18231
+    },
+    {
+      "epoch": 36.464,
+      "grad_norm": 1.910357117652893,
+      "learning_rate": 2e-05,
+      "loss": 0.04690136,
+      "step": 18232
+    },
+    {
+      "epoch": 36.466,
+      "grad_norm": 1.1695517301559448,
+      "learning_rate": 2e-05,
+      "loss": 0.04719057,
+      "step": 18233
+    },
+    {
+      "epoch": 36.468,
+      "grad_norm": 1.595226526260376,
+      "learning_rate": 2e-05,
+      "loss": 0.04680973,
+      "step": 18234
+    },
+    {
+      "epoch": 36.47,
+      "grad_norm": 1.0426281690597534,
+      "learning_rate": 2e-05,
+      "loss": 0.03990892,
+      "step": 18235
+    },
+    {
+      "epoch": 36.472,
+      "grad_norm": 1.362654209136963,
+      "learning_rate": 2e-05,
+      "loss": 0.03500029,
+      "step": 18236
+    },
+    {
+      "epoch": 36.474,
+      "grad_norm": 1.1419274806976318,
+      "learning_rate": 2e-05,
+      "loss": 0.04769664,
+      "step": 18237
+    },
+    {
+      "epoch": 36.476,
+      "grad_norm": 1.1063685417175293,
+      "learning_rate": 2e-05,
+      "loss": 0.03118099,
+      "step": 18238
+    },
+    {
+      "epoch": 36.478,
+      "grad_norm": 1.0712615251541138,
+      "learning_rate": 2e-05,
+      "loss": 0.0473005,
+      "step": 18239
+    },
+    {
+      "epoch": 36.48,
+      "grad_norm": 0.9738886952400208,
+      "learning_rate": 2e-05,
+      "loss": 0.03309727,
+      "step": 18240
+    },
+    {
+      "epoch": 36.482,
+      "grad_norm": 1.6552538871765137,
+      "learning_rate": 2e-05,
+      "loss": 0.05177336,
+      "step": 18241
+    },
+    {
+      "epoch": 36.484,
+      "grad_norm": 0.9547532200813293,
+      "learning_rate": 2e-05,
+      "loss": 0.03158896,
+      "step": 18242
+    },
+    {
+      "epoch": 36.486,
+      "grad_norm": 1.1574902534484863,
+      "learning_rate": 2e-05,
+      "loss": 0.03545589,
+      "step": 18243
+    },
+    {
+      "epoch": 36.488,
+      "grad_norm": 1.1547956466674805,
+      "learning_rate": 2e-05,
+      "loss": 0.05037494,
+      "step": 18244
+    },
+    {
+      "epoch": 36.49,
+      "grad_norm": 1.5758951902389526,
+      "learning_rate": 2e-05,
+      "loss": 0.03462842,
+      "step": 18245
+    },
+    {
+      "epoch": 36.492,
+      "grad_norm": 1.3058934211730957,
+      "learning_rate": 2e-05,
+      "loss": 0.05186077,
+      "step": 18246
+    },
+    {
+      "epoch": 36.494,
+      "grad_norm": 2.153010129928589,
+      "learning_rate": 2e-05,
+      "loss": 0.05587346,
+      "step": 18247
+    },
+    {
+      "epoch": 36.496,
+      "grad_norm": 1.0306788682937622,
+      "learning_rate": 2e-05,
+      "loss": 0.02917742,
+      "step": 18248
+    },
+    {
+      "epoch": 36.498,
+      "grad_norm": 1.1398119926452637,
+      "learning_rate": 2e-05,
+      "loss": 0.0482577,
+      "step": 18249
+    },
+    {
+      "epoch": 36.5,
+      "grad_norm": 1.2190545797348022,
+      "learning_rate": 2e-05,
+      "loss": 0.04970558,
+      "step": 18250
+    },
+    {
+      "epoch": 36.502,
+      "grad_norm": 1.0682768821716309,
+      "learning_rate": 2e-05,
+      "loss": 0.03955474,
+      "step": 18251
+    },
+    {
+      "epoch": 36.504,
+      "grad_norm": 1.061971664428711,
+      "learning_rate": 2e-05,
+      "loss": 0.03687964,
+      "step": 18252
+    },
+    {
+      "epoch": 36.506,
+      "grad_norm": 1.0158944129943848,
+      "learning_rate": 2e-05,
+      "loss": 0.03078072,
+      "step": 18253
+    },
+    {
+      "epoch": 36.508,
+      "grad_norm": 1.3401365280151367,
+      "learning_rate": 2e-05,
+      "loss": 0.0381853,
+      "step": 18254
+    },
+    {
+      "epoch": 36.51,
+      "grad_norm": 1.350685715675354,
+      "learning_rate": 2e-05,
+      "loss": 0.04724355,
+      "step": 18255
+    },
+    {
+      "epoch": 36.512,
+      "grad_norm": 2.3168015480041504,
+      "learning_rate": 2e-05,
+      "loss": 0.04308641,
+      "step": 18256
+    },
+    {
+      "epoch": 36.514,
+      "grad_norm": 1.6673002243041992,
+      "learning_rate": 2e-05,
+      "loss": 0.04929385,
+      "step": 18257
+    },
+    {
+      "epoch": 36.516,
+      "grad_norm": 0.9662206768989563,
+      "learning_rate": 2e-05,
+      "loss": 0.03089986,
+      "step": 18258
+    },
+    {
+      "epoch": 36.518,
+      "grad_norm": 1.062555193901062,
+      "learning_rate": 2e-05,
+      "loss": 0.0395766,
+      "step": 18259
+    },
+    {
+      "epoch": 36.52,
+      "grad_norm": 1.0343207120895386,
+      "learning_rate": 2e-05,
+      "loss": 0.02805191,
+      "step": 18260
+    },
+    {
+      "epoch": 36.522,
+      "grad_norm": 0.9246268272399902,
+      "learning_rate": 2e-05,
+      "loss": 0.02393855,
+      "step": 18261
+    },
+    {
+      "epoch": 36.524,
+      "grad_norm": 1.8312151432037354,
+      "learning_rate": 2e-05,
+      "loss": 0.04711398,
+      "step": 18262
+    },
+    {
+      "epoch": 36.526,
+      "grad_norm": 1.2133851051330566,
+      "learning_rate": 2e-05,
+      "loss": 0.04653526,
+      "step": 18263
+    },
+    {
+      "epoch": 36.528,
+      "grad_norm": 0.9863196015357971,
+      "learning_rate": 2e-05,
+      "loss": 0.02355134,
+      "step": 18264
+    },
+    {
+      "epoch": 36.53,
+      "grad_norm": 1.0515233278274536,
+      "learning_rate": 2e-05,
+      "loss": 0.03952796,
+      "step": 18265
+    },
+    {
+      "epoch": 36.532,
+      "grad_norm": 0.971024751663208,
+      "learning_rate": 2e-05,
+      "loss": 0.03000312,
+      "step": 18266
+    },
+    {
+      "epoch": 36.534,
+      "grad_norm": 1.2390265464782715,
+      "learning_rate": 2e-05,
+      "loss": 0.03372075,
+      "step": 18267
+    },
+    {
+      "epoch": 36.536,
+      "grad_norm": 0.8119271993637085,
+      "learning_rate": 2e-05,
+      "loss": 0.0300536,
+      "step": 18268
+    },
+    {
+      "epoch": 36.538,
+      "grad_norm": 0.9378778338432312,
+      "learning_rate": 2e-05,
+      "loss": 0.02700995,
+      "step": 18269
+    },
+    {
+      "epoch": 36.54,
+      "grad_norm": 2.0789153575897217,
+      "learning_rate": 2e-05,
+      "loss": 0.03948624,
+      "step": 18270
+    },
+    {
+      "epoch": 36.542,
+      "grad_norm": 1.76309335231781,
+      "learning_rate": 2e-05,
+      "loss": 0.04286005,
+      "step": 18271
+    },
+    {
+      "epoch": 36.544,
+      "grad_norm": 4.2802581787109375,
+      "learning_rate": 2e-05,
+      "loss": 0.03132661,
+      "step": 18272
+    },
+    {
+      "epoch": 36.546,
+      "grad_norm": 2.573993444442749,
+      "learning_rate": 2e-05,
+      "loss": 0.04764672,
+      "step": 18273
+    },
+    {
+      "epoch": 36.548,
+      "grad_norm": 0.87221759557724,
+      "learning_rate": 2e-05,
+      "loss": 0.02382921,
+      "step": 18274
+    },
+    {
+      "epoch": 36.55,
+      "grad_norm": 1.0562106370925903,
+      "learning_rate": 2e-05,
+      "loss": 0.04094611,
+      "step": 18275
+    },
+    {
+      "epoch": 36.552,
+      "grad_norm": 1.2717176675796509,
+      "learning_rate": 2e-05,
+      "loss": 0.0410012,
+      "step": 18276
+    },
+    {
+      "epoch": 36.554,
+      "grad_norm": 1.2382874488830566,
+      "learning_rate": 2e-05,
+      "loss": 0.03663464,
+      "step": 18277
+    },
+    {
+      "epoch": 36.556,
+      "grad_norm": 1.964185118675232,
+      "learning_rate": 2e-05,
+      "loss": 0.04018112,
+      "step": 18278
+    },
+    {
+      "epoch": 36.558,
+      "grad_norm": 1.2324621677398682,
+      "learning_rate": 2e-05,
+      "loss": 0.03608364,
+      "step": 18279
+    },
+    {
+      "epoch": 36.56,
+      "grad_norm": 1.2070492506027222,
+      "learning_rate": 2e-05,
+      "loss": 0.05842055,
+      "step": 18280
+    },
+    {
+      "epoch": 36.562,
+      "grad_norm": 1.1493645906448364,
+      "learning_rate": 2e-05,
+      "loss": 0.05052029,
+      "step": 18281
+    },
+    {
+      "epoch": 36.564,
+      "grad_norm": 2.735915422439575,
+      "learning_rate": 2e-05,
+      "loss": 0.04691009,
+      "step": 18282
+    },
+    {
+      "epoch": 36.566,
+      "grad_norm": 1.2203706502914429,
+      "learning_rate": 2e-05,
+      "loss": 0.04751462,
+      "step": 18283
+    },
+    {
+      "epoch": 36.568,
+      "grad_norm": 1.1800801753997803,
+      "learning_rate": 2e-05,
+      "loss": 0.05025848,
+      "step": 18284
+    },
+    {
+      "epoch": 36.57,
+      "grad_norm": 1.0068162679672241,
+      "learning_rate": 2e-05,
+      "loss": 0.03809405,
+      "step": 18285
+    },
+    {
+      "epoch": 36.572,
+      "grad_norm": 1.2168195247650146,
+      "learning_rate": 2e-05,
+      "loss": 0.04326154,
+      "step": 18286
+    },
+    {
+      "epoch": 36.574,
+      "grad_norm": 1.434417724609375,
+      "learning_rate": 2e-05,
+      "loss": 0.0574247,
+      "step": 18287
+    },
+    {
+      "epoch": 36.576,
+      "grad_norm": 1.1110270023345947,
+      "learning_rate": 2e-05,
+      "loss": 0.0267412,
+      "step": 18288
+    },
+    {
+      "epoch": 36.578,
+      "grad_norm": 1.2842190265655518,
+      "learning_rate": 2e-05,
+      "loss": 0.04922368,
+      "step": 18289
+    },
+    {
+      "epoch": 36.58,
+      "grad_norm": 1.4136754274368286,
+      "learning_rate": 2e-05,
+      "loss": 0.03303508,
+      "step": 18290
+    },
+    {
+      "epoch": 36.582,
+      "grad_norm": 1.025667667388916,
+      "learning_rate": 2e-05,
+      "loss": 0.03115687,
+      "step": 18291
+    },
+    {
+      "epoch": 36.584,
+      "grad_norm": 1.3057501316070557,
+      "learning_rate": 2e-05,
+      "loss": 0.05253988,
+      "step": 18292
+    },
+    {
+      "epoch": 36.586,
+      "grad_norm": 1.0783129930496216,
+      "learning_rate": 2e-05,
+      "loss": 0.03761948,
+      "step": 18293
+    },
+    {
+      "epoch": 36.588,
+      "grad_norm": 1.1081454753875732,
+      "learning_rate": 2e-05,
+      "loss": 0.03277981,
+      "step": 18294
+    },
+    {
+      "epoch": 36.59,
+      "grad_norm": 1.0502467155456543,
+      "learning_rate": 2e-05,
+      "loss": 0.03914522,
+      "step": 18295
+    },
+    {
+      "epoch": 36.592,
+      "grad_norm": 1.080041766166687,
+      "learning_rate": 2e-05,
+      "loss": 0.03669895,
+      "step": 18296
+    },
+    {
+      "epoch": 36.594,
+      "grad_norm": 2.066941976547241,
+      "learning_rate": 2e-05,
+      "loss": 0.05867566,
+      "step": 18297
+    },
+    {
+      "epoch": 36.596,
+      "grad_norm": 1.1148244142532349,
+      "learning_rate": 2e-05,
+      "loss": 0.04393861,
+      "step": 18298
+    },
+    {
+      "epoch": 36.598,
+      "grad_norm": 1.0712250471115112,
+      "learning_rate": 2e-05,
+      "loss": 0.03914363,
+      "step": 18299
+    },
+    {
+      "epoch": 36.6,
+      "grad_norm": 1.2081435918807983,
+      "learning_rate": 2e-05,
+      "loss": 0.03771123,
+      "step": 18300
+    },
+    {
+      "epoch": 36.602,
+      "grad_norm": 1.143244743347168,
+      "learning_rate": 2e-05,
+      "loss": 0.04224782,
+      "step": 18301
+    },
+    {
+      "epoch": 36.604,
+      "grad_norm": 1.2254302501678467,
+      "learning_rate": 2e-05,
+      "loss": 0.04013958,
+      "step": 18302
+    },
+    {
+      "epoch": 36.606,
+      "grad_norm": 1.3752824068069458,
+      "learning_rate": 2e-05,
+      "loss": 0.050212,
+      "step": 18303
+    },
+    {
+      "epoch": 36.608,
+      "grad_norm": 0.94117671251297,
+      "learning_rate": 2e-05,
+      "loss": 0.03055703,
+      "step": 18304
+    },
+    {
+      "epoch": 36.61,
+      "grad_norm": 1.1006180047988892,
+      "learning_rate": 2e-05,
+      "loss": 0.03390218,
+      "step": 18305
+    },
+    {
+      "epoch": 36.612,
+      "grad_norm": 1.3862262964248657,
+      "learning_rate": 2e-05,
+      "loss": 0.05241469,
+      "step": 18306
+    },
+    {
+      "epoch": 36.614,
+      "grad_norm": 1.8648090362548828,
+      "learning_rate": 2e-05,
+      "loss": 0.04927859,
+      "step": 18307
+    },
+    {
+      "epoch": 36.616,
+      "grad_norm": 1.4699233770370483,
+      "learning_rate": 2e-05,
+      "loss": 0.04433728,
+      "step": 18308
+    },
+    {
+      "epoch": 36.618,
+      "grad_norm": 1.4989550113677979,
+      "learning_rate": 2e-05,
+      "loss": 0.05942933,
+      "step": 18309
+    },
+    {
+      "epoch": 36.62,
+      "grad_norm": 1.1221908330917358,
+      "learning_rate": 2e-05,
+      "loss": 0.047569,
+      "step": 18310
+    },
+    {
+      "epoch": 36.622,
+      "grad_norm": 1.540700078010559,
+      "learning_rate": 2e-05,
+      "loss": 0.05225103,
+      "step": 18311
+    },
+    {
+      "epoch": 36.624,
+      "grad_norm": 1.2131165266036987,
+      "learning_rate": 2e-05,
+      "loss": 0.04262118,
+      "step": 18312
+    },
+    {
+      "epoch": 36.626,
+      "grad_norm": 1.1024413108825684,
+      "learning_rate": 2e-05,
+      "loss": 0.03573479,
+      "step": 18313
+    },
+    {
+      "epoch": 36.628,
+      "grad_norm": 1.9609568119049072,
+      "learning_rate": 2e-05,
+      "loss": 0.04972448,
+      "step": 18314
+    },
+    {
+      "epoch": 36.63,
+      "grad_norm": 1.236175775527954,
+      "learning_rate": 2e-05,
+      "loss": 0.03609653,
+      "step": 18315
+    },
+    {
+      "epoch": 36.632,
+      "grad_norm": 1.202167272567749,
+      "learning_rate": 2e-05,
+      "loss": 0.04706704,
+      "step": 18316
+    },
+    {
+      "epoch": 36.634,
+      "grad_norm": 1.1365604400634766,
+      "learning_rate": 2e-05,
+      "loss": 0.02697523,
+      "step": 18317
+    },
+    {
+      "epoch": 36.636,
+      "grad_norm": 2.451118230819702,
+      "learning_rate": 2e-05,
+      "loss": 0.04429175,
+      "step": 18318
+    },
+    {
+      "epoch": 36.638,
+      "grad_norm": 1.0182795524597168,
+      "learning_rate": 2e-05,
+      "loss": 0.03436865,
+      "step": 18319
+    },
+    {
+      "epoch": 36.64,
+      "grad_norm": 1.2875375747680664,
+      "learning_rate": 2e-05,
+      "loss": 0.0550098,
+      "step": 18320
+    },
+    {
+      "epoch": 36.642,
+      "grad_norm": 1.559504508972168,
+      "learning_rate": 2e-05,
+      "loss": 0.05564836,
+      "step": 18321
+    },
+    {
+      "epoch": 36.644,
+      "grad_norm": 1.1889086961746216,
+      "learning_rate": 2e-05,
+      "loss": 0.03784901,
+      "step": 18322
+    },
+    {
+      "epoch": 36.646,
+      "grad_norm": 0.8898120522499084,
+      "learning_rate": 2e-05,
+      "loss": 0.0278656,
+      "step": 18323
+    },
+    {
+      "epoch": 36.648,
+      "grad_norm": 1.1782766580581665,
+      "learning_rate": 2e-05,
+      "loss": 0.03811631,
+      "step": 18324
+    },
+    {
+      "epoch": 36.65,
+      "grad_norm": 1.2543221712112427,
+      "learning_rate": 2e-05,
+      "loss": 0.03640786,
+      "step": 18325
+    },
+    {
+      "epoch": 36.652,
+      "grad_norm": 1.0462356805801392,
+      "learning_rate": 2e-05,
+      "loss": 0.03829445,
+      "step": 18326
+    },
+    {
+      "epoch": 36.654,
+      "grad_norm": 1.3017123937606812,
+      "learning_rate": 2e-05,
+      "loss": 0.03575253,
+      "step": 18327
+    },
+    {
+      "epoch": 36.656,
+      "grad_norm": 1.342948317527771,
+      "learning_rate": 2e-05,
+      "loss": 0.03757192,
+      "step": 18328
+    },
+    {
+      "epoch": 36.658,
+      "grad_norm": 1.066953420639038,
+      "learning_rate": 2e-05,
+      "loss": 0.04221145,
+      "step": 18329
+    },
+    {
+      "epoch": 36.66,
+      "grad_norm": 1.3225663900375366,
+      "learning_rate": 2e-05,
+      "loss": 0.05248147,
+      "step": 18330
+    },
+    {
+      "epoch": 36.662,
+      "grad_norm": 2.983029842376709,
+      "learning_rate": 2e-05,
+      "loss": 0.03616619,
+      "step": 18331
+    },
+    {
+      "epoch": 36.664,
+      "grad_norm": 2.2221550941467285,
+      "learning_rate": 2e-05,
+      "loss": 0.04058126,
+      "step": 18332
+    },
+    {
+      "epoch": 36.666,
+      "grad_norm": 1.315987467765808,
+      "learning_rate": 2e-05,
+      "loss": 0.03588897,
+      "step": 18333
+    },
+    {
+      "epoch": 36.668,
+      "grad_norm": 0.9215741157531738,
+      "learning_rate": 2e-05,
+      "loss": 0.03925972,
+      "step": 18334
+    },
+    {
+      "epoch": 36.67,
+      "grad_norm": 1.2016059160232544,
+      "learning_rate": 2e-05,
+      "loss": 0.05430229,
+      "step": 18335
+    },
+    {
+      "epoch": 36.672,
+      "grad_norm": 0.921889066696167,
+      "learning_rate": 2e-05,
+      "loss": 0.03469507,
+      "step": 18336
+    },
+    {
+      "epoch": 36.674,
+      "grad_norm": 0.9315686821937561,
+      "learning_rate": 2e-05,
+      "loss": 0.02187958,
+      "step": 18337
+    },
+    {
+      "epoch": 36.676,
+      "grad_norm": 0.9967896938323975,
+      "learning_rate": 2e-05,
+      "loss": 0.0331551,
+      "step": 18338
+    },
+    {
+      "epoch": 36.678,
+      "grad_norm": 1.1743967533111572,
+      "learning_rate": 2e-05,
+      "loss": 0.04423882,
+      "step": 18339
+    },
+    {
+      "epoch": 36.68,
+      "grad_norm": 1.2126519680023193,
+      "learning_rate": 2e-05,
+      "loss": 0.03514692,
+      "step": 18340
+    },
+    {
+      "epoch": 36.682,
+      "grad_norm": 1.3199232816696167,
+      "learning_rate": 2e-05,
+      "loss": 0.03419685,
+      "step": 18341
+    },
+    {
+      "epoch": 36.684,
+      "grad_norm": 1.7518872022628784,
+      "learning_rate": 2e-05,
+      "loss": 0.05697691,
+      "step": 18342
+    },
+    {
+      "epoch": 36.686,
+      "grad_norm": 1.0076147317886353,
+      "learning_rate": 2e-05,
+      "loss": 0.0281866,
+      "step": 18343
+    },
+    {
+      "epoch": 36.688,
+      "grad_norm": 1.5172227621078491,
+      "learning_rate": 2e-05,
+      "loss": 0.05067704,
+      "step": 18344
+    },
+    {
+      "epoch": 36.69,
+      "grad_norm": 1.6824908256530762,
+      "learning_rate": 2e-05,
+      "loss": 0.04503788,
+      "step": 18345
+    },
+    {
+      "epoch": 36.692,
+      "grad_norm": 1.3745967149734497,
+      "learning_rate": 2e-05,
+      "loss": 0.0483752,
+      "step": 18346
+    },
+    {
+      "epoch": 36.694,
+      "grad_norm": 1.2463769912719727,
+      "learning_rate": 2e-05,
+      "loss": 0.05347317,
+      "step": 18347
+    },
+    {
+      "epoch": 36.696,
+      "grad_norm": 1.0087859630584717,
+      "learning_rate": 2e-05,
+      "loss": 0.03465716,
+      "step": 18348
+    },
+    {
+      "epoch": 36.698,
+      "grad_norm": 1.2997995615005493,
+      "learning_rate": 2e-05,
+      "loss": 0.03587429,
+      "step": 18349
+    },
+    {
+      "epoch": 36.7,
+      "grad_norm": 1.1634178161621094,
+      "learning_rate": 2e-05,
+      "loss": 0.02643599,
+      "step": 18350
+    },
+    {
+      "epoch": 36.702,
+      "grad_norm": 1.2005338668823242,
+      "learning_rate": 2e-05,
+      "loss": 0.04984972,
+      "step": 18351
+    },
+    {
+      "epoch": 36.704,
+      "grad_norm": 0.9318526387214661,
+      "learning_rate": 2e-05,
+      "loss": 0.02411564,
+      "step": 18352
+    },
+    {
+      "epoch": 36.706,
+      "grad_norm": 1.1339325904846191,
+      "learning_rate": 2e-05,
+      "loss": 0.03448115,
+      "step": 18353
+    },
+    {
+      "epoch": 36.708,
+      "grad_norm": 1.1679779291152954,
+      "learning_rate": 2e-05,
+      "loss": 0.03647584,
+      "step": 18354
+    },
+    {
+      "epoch": 36.71,
+      "grad_norm": 1.2104750871658325,
+      "learning_rate": 2e-05,
+      "loss": 0.0515867,
+      "step": 18355
+    },
+    {
+      "epoch": 36.712,
+      "grad_norm": 2.100127935409546,
+      "learning_rate": 2e-05,
+      "loss": 0.05232673,
+      "step": 18356
+    },
+    {
+      "epoch": 36.714,
+      "grad_norm": 1.754447340965271,
+      "learning_rate": 2e-05,
+      "loss": 0.04861075,
+      "step": 18357
+    },
+    {
+      "epoch": 36.716,
+      "grad_norm": 1.008870005607605,
+      "learning_rate": 2e-05,
+      "loss": 0.03333037,
+      "step": 18358
+    },
+    {
+      "epoch": 36.718,
+      "grad_norm": 1.7049437761306763,
+      "learning_rate": 2e-05,
+      "loss": 0.03618935,
+      "step": 18359
+    },
+    {
+      "epoch": 36.72,
+      "grad_norm": 1.8496776819229126,
+      "learning_rate": 2e-05,
+      "loss": 0.05417331,
+      "step": 18360
+    },
+    {
+      "epoch": 36.722,
+      "grad_norm": 1.0589567422866821,
+      "learning_rate": 2e-05,
+      "loss": 0.02868169,
+      "step": 18361
+    },
+    {
+      "epoch": 36.724,
+      "grad_norm": 0.9607617259025574,
+      "learning_rate": 2e-05,
+      "loss": 0.03263365,
+      "step": 18362
+    },
+    {
+      "epoch": 36.726,
+      "grad_norm": 2.0316452980041504,
+      "learning_rate": 2e-05,
+      "loss": 0.04388903,
+      "step": 18363
+    },
+    {
+      "epoch": 36.728,
+      "grad_norm": 1.2382222414016724,
+      "learning_rate": 2e-05,
+      "loss": 0.04471454,
+      "step": 18364
+    },
+    {
+      "epoch": 36.73,
+      "grad_norm": 1.287028193473816,
+      "learning_rate": 2e-05,
+      "loss": 0.04440287,
+      "step": 18365
+    },
+    {
+      "epoch": 36.732,
+      "grad_norm": 1.2623907327651978,
+      "learning_rate": 2e-05,
+      "loss": 0.03233304,
+      "step": 18366
+    },
+    {
+      "epoch": 36.734,
+      "grad_norm": 1.9164531230926514,
+      "learning_rate": 2e-05,
+      "loss": 0.04776657,
+      "step": 18367
+    },
+    {
+      "epoch": 36.736,
+      "grad_norm": 1.3518916368484497,
+      "learning_rate": 2e-05,
+      "loss": 0.06090958,
+      "step": 18368
+    },
+    {
+      "epoch": 36.738,
+      "grad_norm": 1.8850350379943848,
+      "learning_rate": 2e-05,
+      "loss": 0.03893983,
+      "step": 18369
+    },
+    {
+      "epoch": 36.74,
+      "grad_norm": 1.0571264028549194,
+      "learning_rate": 2e-05,
+      "loss": 0.03137238,
+      "step": 18370
+    },
+    {
+      "epoch": 36.742,
+      "grad_norm": 1.1914355754852295,
+      "learning_rate": 2e-05,
+      "loss": 0.03433579,
+      "step": 18371
+    },
+    {
+      "epoch": 36.744,
+      "grad_norm": 0.8919147849082947,
+      "learning_rate": 2e-05,
+      "loss": 0.0218019,
+      "step": 18372
+    },
+    {
+      "epoch": 36.746,
+      "grad_norm": 1.1925781965255737,
+      "learning_rate": 2e-05,
+      "loss": 0.05873271,
+      "step": 18373
+    },
+    {
+      "epoch": 36.748,
+      "grad_norm": 1.1038137674331665,
+      "learning_rate": 2e-05,
+      "loss": 0.04641431,
+      "step": 18374
+    },
+    {
+      "epoch": 36.75,
+      "grad_norm": 0.965733528137207,
+      "learning_rate": 2e-05,
+      "loss": 0.02856233,
+      "step": 18375
+    },
+    {
+      "epoch": 36.752,
+      "grad_norm": 1.5396867990493774,
+      "learning_rate": 2e-05,
+      "loss": 0.04703198,
+      "step": 18376
+    },
+    {
+      "epoch": 36.754,
+      "grad_norm": 2.1341171264648438,
+      "learning_rate": 2e-05,
+      "loss": 0.04486314,
+      "step": 18377
+    },
+    {
+      "epoch": 36.756,
+      "grad_norm": 1.7395333051681519,
+      "learning_rate": 2e-05,
+      "loss": 0.04411499,
+      "step": 18378
+    },
+    {
+      "epoch": 36.758,
+      "grad_norm": 1.706061840057373,
+      "learning_rate": 2e-05,
+      "loss": 0.03239989,
+      "step": 18379
+    },
+    {
+      "epoch": 36.76,
+      "grad_norm": 1.6504026651382446,
+      "learning_rate": 2e-05,
+      "loss": 0.03501267,
+      "step": 18380
+    },
+    {
+      "epoch": 36.762,
+      "grad_norm": 1.3423389196395874,
+      "learning_rate": 2e-05,
+      "loss": 0.05261109,
+      "step": 18381
+    },
+    {
+      "epoch": 36.764,
+      "grad_norm": 1.3317161798477173,
+      "learning_rate": 2e-05,
+      "loss": 0.04133341,
+      "step": 18382
+    },
+    {
+      "epoch": 36.766,
+      "grad_norm": 1.6479923725128174,
+      "learning_rate": 2e-05,
+      "loss": 0.04605314,
+      "step": 18383
+    },
+    {
+      "epoch": 36.768,
+      "grad_norm": 1.1072478294372559,
+      "learning_rate": 2e-05,
+      "loss": 0.03600177,
+      "step": 18384
+    },
+    {
+      "epoch": 36.77,
+      "grad_norm": 1.1794970035552979,
+      "learning_rate": 2e-05,
+      "loss": 0.05206569,
+      "step": 18385
+    },
+    {
+      "epoch": 36.772,
+      "grad_norm": 1.075430989265442,
+      "learning_rate": 2e-05,
+      "loss": 0.03614403,
+      "step": 18386
+    },
+    {
+      "epoch": 36.774,
+      "grad_norm": 1.2344156503677368,
+      "learning_rate": 2e-05,
+      "loss": 0.06035028,
+      "step": 18387
+    },
+    {
+      "epoch": 36.776,
+      "grad_norm": 1.0581718683242798,
+      "learning_rate": 2e-05,
+      "loss": 0.0294181,
+      "step": 18388
+    },
+    {
+      "epoch": 36.778,
+      "grad_norm": 1.7531042098999023,
+      "learning_rate": 2e-05,
+      "loss": 0.04625558,
+      "step": 18389
+    },
+    {
+      "epoch": 36.78,
+      "grad_norm": 1.1771091222763062,
+      "learning_rate": 2e-05,
+      "loss": 0.04027297,
+      "step": 18390
+    },
+    {
+      "epoch": 36.782,
+      "grad_norm": 0.9724439978599548,
+      "learning_rate": 2e-05,
+      "loss": 0.03231468,
+      "step": 18391
+    },
+    {
+      "epoch": 36.784,
+      "grad_norm": 1.1481809616088867,
+      "learning_rate": 2e-05,
+      "loss": 0.03959997,
+      "step": 18392
+    },
+    {
+      "epoch": 36.786,
+      "grad_norm": 1.2926446199417114,
+      "learning_rate": 2e-05,
+      "loss": 0.05425737,
+      "step": 18393
+    },
+    {
+      "epoch": 36.788,
+      "grad_norm": 1.4414113759994507,
+      "learning_rate": 2e-05,
+      "loss": 0.04928648,
+      "step": 18394
+    },
+    {
+      "epoch": 36.79,
+      "grad_norm": 1.5177888870239258,
+      "learning_rate": 2e-05,
+      "loss": 0.03690218,
+      "step": 18395
+    },
+    {
+      "epoch": 36.792,
+      "grad_norm": 1.5356523990631104,
+      "learning_rate": 2e-05,
+      "loss": 0.04718888,
+      "step": 18396
+    },
+    {
+      "epoch": 36.794,
+      "grad_norm": 2.2311973571777344,
+      "learning_rate": 2e-05,
+      "loss": 0.03536499,
+      "step": 18397
+    },
+    {
+      "epoch": 36.796,
+      "grad_norm": 2.834324836730957,
+      "learning_rate": 2e-05,
+      "loss": 0.04099808,
+      "step": 18398
+    },
+    {
+      "epoch": 36.798,
+      "grad_norm": 1.1920080184936523,
+      "learning_rate": 2e-05,
+      "loss": 0.04433641,
+      "step": 18399
+    },
+    {
+      "epoch": 36.8,
+      "grad_norm": 1.1526068449020386,
+      "learning_rate": 2e-05,
+      "loss": 0.04516728,
+      "step": 18400
+    },
+    {
+      "epoch": 36.802,
+      "grad_norm": 1.2687426805496216,
+      "learning_rate": 2e-05,
+      "loss": 0.03586322,
+      "step": 18401
+    },
+    {
+      "epoch": 36.804,
+      "grad_norm": 0.9818292856216431,
+      "learning_rate": 2e-05,
+      "loss": 0.02395935,
+      "step": 18402
+    },
+    {
+      "epoch": 36.806,
+      "grad_norm": 2.2147302627563477,
+      "learning_rate": 2e-05,
+      "loss": 0.04670376,
+      "step": 18403
+    },
+    {
+      "epoch": 36.808,
+      "grad_norm": 1.6681506633758545,
+      "learning_rate": 2e-05,
+      "loss": 0.05808145,
+      "step": 18404
+    },
+    {
+      "epoch": 36.81,
+      "grad_norm": 1.1924699544906616,
+      "learning_rate": 2e-05,
+      "loss": 0.04742082,
+      "step": 18405
+    },
+    {
+      "epoch": 36.812,
+      "grad_norm": 2.0133984088897705,
+      "learning_rate": 2e-05,
+      "loss": 0.03901449,
+      "step": 18406
+    },
+    {
+      "epoch": 36.814,
+      "grad_norm": 1.1701860427856445,
+      "learning_rate": 2e-05,
+      "loss": 0.02848022,
+      "step": 18407
+    },
+    {
+      "epoch": 36.816,
+      "grad_norm": 1.089707374572754,
+      "learning_rate": 2e-05,
+      "loss": 0.04020487,
+      "step": 18408
+    },
+    {
+      "epoch": 36.818,
+      "grad_norm": 1.0396581888198853,
+      "learning_rate": 2e-05,
+      "loss": 0.03704511,
+      "step": 18409
+    },
+    {
+      "epoch": 36.82,
+      "grad_norm": 1.6695880889892578,
+      "learning_rate": 2e-05,
+      "loss": 0.0519683,
+      "step": 18410
+    },
+    {
+      "epoch": 36.822,
+      "grad_norm": 1.1158127784729004,
+      "learning_rate": 2e-05,
+      "loss": 0.04286835,
+      "step": 18411
+    },
+    {
+      "epoch": 36.824,
+      "grad_norm": 1.1745147705078125,
+      "learning_rate": 2e-05,
+      "loss": 0.03942792,
+      "step": 18412
+    },
+    {
+      "epoch": 36.826,
+      "grad_norm": 0.8886756896972656,
+      "learning_rate": 2e-05,
+      "loss": 0.02771359,
+      "step": 18413
+    },
+    {
+      "epoch": 36.828,
+      "grad_norm": 1.2267519235610962,
+      "learning_rate": 2e-05,
+      "loss": 0.0511631,
+      "step": 18414
+    },
+    {
+      "epoch": 36.83,
+      "grad_norm": 2.179556369781494,
+      "learning_rate": 2e-05,
+      "loss": 0.06088002,
+      "step": 18415
+    },
+    {
+      "epoch": 36.832,
+      "grad_norm": 1.2211542129516602,
+      "learning_rate": 2e-05,
+      "loss": 0.04303684,
+      "step": 18416
+    },
+    {
+      "epoch": 36.834,
+      "grad_norm": 2.2510719299316406,
+      "learning_rate": 2e-05,
+      "loss": 0.05186936,
+      "step": 18417
+    },
+    {
+      "epoch": 36.836,
+      "grad_norm": 1.2228981256484985,
+      "learning_rate": 2e-05,
+      "loss": 0.04879232,
+      "step": 18418
+    },
+    {
+      "epoch": 36.838,
+      "grad_norm": 1.098894715309143,
+      "learning_rate": 2e-05,
+      "loss": 0.03979594,
+      "step": 18419
+    },
+    {
+      "epoch": 36.84,
+      "grad_norm": 1.2254807949066162,
+      "learning_rate": 2e-05,
+      "loss": 0.03765118,
+      "step": 18420
+    },
+    {
+      "epoch": 36.842,
+      "grad_norm": 1.175650954246521,
+      "learning_rate": 2e-05,
+      "loss": 0.04586347,
+      "step": 18421
+    },
+    {
+      "epoch": 36.844,
+      "grad_norm": 1.22563636302948,
+      "learning_rate": 2e-05,
+      "loss": 0.0531856,
+      "step": 18422
+    },
+    {
+      "epoch": 36.846,
+      "grad_norm": 1.2885513305664062,
+      "learning_rate": 2e-05,
+      "loss": 0.04552082,
+      "step": 18423
+    },
+    {
+      "epoch": 36.848,
+      "grad_norm": 1.1295452117919922,
+      "learning_rate": 2e-05,
+      "loss": 0.0402256,
+      "step": 18424
+    },
+    {
+      "epoch": 36.85,
+      "grad_norm": 1.0516687631607056,
+      "learning_rate": 2e-05,
+      "loss": 0.04456215,
+      "step": 18425
+    },
+    {
+      "epoch": 36.852,
+      "grad_norm": 1.0733994245529175,
+      "learning_rate": 2e-05,
+      "loss": 0.03550437,
+      "step": 18426
+    },
+    {
+      "epoch": 36.854,
+      "grad_norm": 1.1007059812545776,
+      "learning_rate": 2e-05,
+      "loss": 0.04091225,
+      "step": 18427
+    },
+    {
+      "epoch": 36.856,
+      "grad_norm": 1.5547980070114136,
+      "learning_rate": 2e-05,
+      "loss": 0.04072155,
+      "step": 18428
+    },
+    {
+      "epoch": 36.858,
+      "grad_norm": 1.1082590818405151,
+      "learning_rate": 2e-05,
+      "loss": 0.03597355,
+      "step": 18429
+    },
+    {
+      "epoch": 36.86,
+      "grad_norm": 1.299734354019165,
+      "learning_rate": 2e-05,
+      "loss": 0.03989294,
+      "step": 18430
+    },
+    {
+      "epoch": 36.862,
+      "grad_norm": 0.9722706079483032,
+      "learning_rate": 2e-05,
+      "loss": 0.03567673,
+      "step": 18431
+    },
+    {
+      "epoch": 36.864,
+      "grad_norm": 1.2522664070129395,
+      "learning_rate": 2e-05,
+      "loss": 0.04869508,
+      "step": 18432
+    },
+    {
+      "epoch": 36.866,
+      "grad_norm": 1.638421654701233,
+      "learning_rate": 2e-05,
+      "loss": 0.03735339,
+      "step": 18433
+    },
+    {
+      "epoch": 36.868,
+      "grad_norm": 1.0237358808517456,
+      "learning_rate": 2e-05,
+      "loss": 0.04364359,
+      "step": 18434
+    },
+    {
+      "epoch": 36.87,
+      "grad_norm": 1.2411482334136963,
+      "learning_rate": 2e-05,
+      "loss": 0.03947395,
+      "step": 18435
+    },
+    {
+      "epoch": 36.872,
+      "grad_norm": 1.159613847732544,
+      "learning_rate": 2e-05,
+      "loss": 0.04042497,
+      "step": 18436
+    },
+    {
+      "epoch": 36.874,
+      "grad_norm": 1.2544336318969727,
+      "learning_rate": 2e-05,
+      "loss": 0.0312546,
+      "step": 18437
+    },
+    {
+      "epoch": 36.876,
+      "grad_norm": 1.9354642629623413,
+      "learning_rate": 2e-05,
+      "loss": 0.0527093,
+      "step": 18438
+    },
+    {
+      "epoch": 36.878,
+      "grad_norm": 3.310471773147583,
+      "learning_rate": 2e-05,
+      "loss": 0.04944807,
+      "step": 18439
+    },
+    {
+      "epoch": 36.88,
+      "grad_norm": 1.1568819284439087,
+      "learning_rate": 2e-05,
+      "loss": 0.0369446,
+      "step": 18440
+    },
+    {
+      "epoch": 36.882,
+      "grad_norm": 1.2008459568023682,
+      "learning_rate": 2e-05,
+      "loss": 0.04868155,
+      "step": 18441
+    },
+    {
+      "epoch": 36.884,
+      "grad_norm": 1.064301609992981,
+      "learning_rate": 2e-05,
+      "loss": 0.03515521,
+      "step": 18442
+    },
+    {
+      "epoch": 36.886,
+      "grad_norm": 1.3205647468566895,
+      "learning_rate": 2e-05,
+      "loss": 0.03227288,
+      "step": 18443
+    },
+    {
+      "epoch": 36.888,
+      "grad_norm": 1.5607926845550537,
+      "learning_rate": 2e-05,
+      "loss": 0.04700986,
+      "step": 18444
+    },
+    {
+      "epoch": 36.89,
+      "grad_norm": 1.1326090097427368,
+      "learning_rate": 2e-05,
+      "loss": 0.03524204,
+      "step": 18445
+    },
+    {
+      "epoch": 36.892,
+      "grad_norm": 0.9644931554794312,
+      "learning_rate": 2e-05,
+      "loss": 0.0342176,
+      "step": 18446
+    },
+    {
+      "epoch": 36.894,
+      "grad_norm": 1.2272518873214722,
+      "learning_rate": 2e-05,
+      "loss": 0.03322006,
+      "step": 18447
+    },
+    {
+      "epoch": 36.896,
+      "grad_norm": 1.061239242553711,
+      "learning_rate": 2e-05,
+      "loss": 0.03333382,
+      "step": 18448
+    },
+    {
+      "epoch": 36.898,
+      "grad_norm": 1.1636090278625488,
+      "learning_rate": 2e-05,
+      "loss": 0.04022244,
+      "step": 18449
+    },
+    {
+      "epoch": 36.9,
+      "grad_norm": 1.0282191038131714,
+      "learning_rate": 2e-05,
+      "loss": 0.03084,
+      "step": 18450
+    },
+    {
+      "epoch": 36.902,
+      "grad_norm": 2.087050199508667,
+      "learning_rate": 2e-05,
+      "loss": 0.0529213,
+      "step": 18451
+    },
+    {
+      "epoch": 36.904,
+      "grad_norm": 1.0408107042312622,
+      "learning_rate": 2e-05,
+      "loss": 0.04363842,
+      "step": 18452
+    },
+    {
+      "epoch": 36.906,
+      "grad_norm": 1.284910798072815,
+      "learning_rate": 2e-05,
+      "loss": 0.0489226,
+      "step": 18453
+    },
+    {
+      "epoch": 36.908,
+      "grad_norm": 1.6543158292770386,
+      "learning_rate": 2e-05,
+      "loss": 0.06356652,
+      "step": 18454
+    },
+    {
+      "epoch": 36.91,
+      "grad_norm": 1.2143512964248657,
+      "learning_rate": 2e-05,
+      "loss": 0.05602705,
+      "step": 18455
+    },
+    {
+      "epoch": 36.912,
+      "grad_norm": 1.095168113708496,
+      "learning_rate": 2e-05,
+      "loss": 0.04102691,
+      "step": 18456
+    },
+    {
+      "epoch": 36.914,
+      "grad_norm": 1.4219430685043335,
+      "learning_rate": 2e-05,
+      "loss": 0.04681015,
+      "step": 18457
+    },
+    {
+      "epoch": 36.916,
+      "grad_norm": 1.1736583709716797,
+      "learning_rate": 2e-05,
+      "loss": 0.04344864,
+      "step": 18458
+    },
+    {
+      "epoch": 36.918,
+      "grad_norm": 0.9385769963264465,
+      "learning_rate": 2e-05,
+      "loss": 0.02884827,
+      "step": 18459
+    },
+    {
+      "epoch": 36.92,
+      "grad_norm": 1.4176242351531982,
+      "learning_rate": 2e-05,
+      "loss": 0.06416596,
+      "step": 18460
+    },
+    {
+      "epoch": 36.922,
+      "grad_norm": 1.288359522819519,
+      "learning_rate": 2e-05,
+      "loss": 0.04654737,
+      "step": 18461
+    },
+    {
+      "epoch": 36.924,
+      "grad_norm": 1.1389371156692505,
+      "learning_rate": 2e-05,
+      "loss": 0.04628037,
+      "step": 18462
+    },
+    {
+      "epoch": 36.926,
+      "grad_norm": 1.098805546760559,
+      "learning_rate": 2e-05,
+      "loss": 0.03729876,
+      "step": 18463
+    },
+    {
+      "epoch": 36.928,
+      "grad_norm": 1.4324229955673218,
+      "learning_rate": 2e-05,
+      "loss": 0.0529637,
+      "step": 18464
+    },
+    {
+      "epoch": 36.93,
+      "grad_norm": 1.4143197536468506,
+      "learning_rate": 2e-05,
+      "loss": 0.03819642,
+      "step": 18465
+    },
+    {
+      "epoch": 36.932,
+      "grad_norm": 1.4157121181488037,
+      "learning_rate": 2e-05,
+      "loss": 0.04541536,
+      "step": 18466
+    },
+    {
+      "epoch": 36.934,
+      "grad_norm": 1.393878698348999,
+      "learning_rate": 2e-05,
+      "loss": 0.04265588,
+      "step": 18467
+    },
+    {
+      "epoch": 36.936,
+      "grad_norm": 1.0789780616760254,
+      "learning_rate": 2e-05,
+      "loss": 0.04042588,
+      "step": 18468
+    },
+    {
+      "epoch": 36.938,
+      "grad_norm": 1.6265769004821777,
+      "learning_rate": 2e-05,
+      "loss": 0.05098584,
+      "step": 18469
+    },
+    {
+      "epoch": 36.94,
+      "grad_norm": 1.1690245866775513,
+      "learning_rate": 2e-05,
+      "loss": 0.04554411,
+      "step": 18470
+    },
+    {
+      "epoch": 36.942,
+      "grad_norm": 1.3794010877609253,
+      "learning_rate": 2e-05,
+      "loss": 0.044034,
+      "step": 18471
+    },
+    {
+      "epoch": 36.944,
+      "grad_norm": 0.8413645625114441,
+      "learning_rate": 2e-05,
+      "loss": 0.02523428,
+      "step": 18472
+    },
+    {
+      "epoch": 36.946,
+      "grad_norm": 0.9403594732284546,
+      "learning_rate": 2e-05,
+      "loss": 0.03138512,
+      "step": 18473
+    },
+    {
+      "epoch": 36.948,
+      "grad_norm": 1.1274261474609375,
+      "learning_rate": 2e-05,
+      "loss": 0.0406789,
+      "step": 18474
+    },
+    {
+      "epoch": 36.95,
+      "grad_norm": 1.4938697814941406,
+      "learning_rate": 2e-05,
+      "loss": 0.05272559,
+      "step": 18475
+    },
+    {
+      "epoch": 36.952,
+      "grad_norm": 1.0316252708435059,
+      "learning_rate": 2e-05,
+      "loss": 0.03562522,
+      "step": 18476
+    },
+    {
+      "epoch": 36.954,
+      "grad_norm": 1.3863626718521118,
+      "learning_rate": 2e-05,
+      "loss": 0.03377339,
+      "step": 18477
+    },
+    {
+      "epoch": 36.956,
+      "grad_norm": 1.7842824459075928,
+      "learning_rate": 2e-05,
+      "loss": 0.05903577,
+      "step": 18478
+    },
+    {
+      "epoch": 36.958,
+      "grad_norm": 1.35276198387146,
+      "learning_rate": 2e-05,
+      "loss": 0.03431714,
+      "step": 18479
+    },
+    {
+      "epoch": 36.96,
+      "grad_norm": 1.2699631452560425,
+      "learning_rate": 2e-05,
+      "loss": 0.0535304,
+      "step": 18480
+    },
+    {
+      "epoch": 36.962,
+      "grad_norm": 1.2348072528839111,
+      "learning_rate": 2e-05,
+      "loss": 0.04691994,
+      "step": 18481
+    },
+    {
+      "epoch": 36.964,
+      "grad_norm": 0.9137774705886841,
+      "learning_rate": 2e-05,
+      "loss": 0.0308315,
+      "step": 18482
+    },
+    {
+      "epoch": 36.966,
+      "grad_norm": 1.0427780151367188,
+      "learning_rate": 2e-05,
+      "loss": 0.0346647,
+      "step": 18483
+    },
+    {
+      "epoch": 36.968,
+      "grad_norm": 2.189744472503662,
+      "learning_rate": 2e-05,
+      "loss": 0.03301712,
+      "step": 18484
+    },
+    {
+      "epoch": 36.97,
+      "grad_norm": 1.0989826917648315,
+      "learning_rate": 2e-05,
+      "loss": 0.02908636,
+      "step": 18485
+    },
+    {
+      "epoch": 36.972,
+      "grad_norm": 1.6174060106277466,
+      "learning_rate": 2e-05,
+      "loss": 0.04544824,
+      "step": 18486
+    },
+    {
+      "epoch": 36.974,
+      "grad_norm": 1.2668603658676147,
+      "learning_rate": 2e-05,
+      "loss": 0.05053727,
+      "step": 18487
+    },
+    {
+      "epoch": 36.976,
+      "grad_norm": 1.0576096773147583,
+      "learning_rate": 2e-05,
+      "loss": 0.03743074,
+      "step": 18488
+    },
+    {
+      "epoch": 36.978,
+      "grad_norm": 1.1801321506500244,
+      "learning_rate": 2e-05,
+      "loss": 0.03683414,
+      "step": 18489
+    },
+    {
+      "epoch": 36.98,
+      "grad_norm": 1.1503815650939941,
+      "learning_rate": 2e-05,
+      "loss": 0.03822848,
+      "step": 18490
+    },
+    {
+      "epoch": 36.982,
+      "grad_norm": 1.3549894094467163,
+      "learning_rate": 2e-05,
+      "loss": 0.05866414,
+      "step": 18491
+    },
+    {
+      "epoch": 36.984,
+      "grad_norm": 1.0250877141952515,
+      "learning_rate": 2e-05,
+      "loss": 0.03789775,
+      "step": 18492
+    },
+    {
+      "epoch": 36.986,
+      "grad_norm": 1.0828768014907837,
+      "learning_rate": 2e-05,
+      "loss": 0.04611751,
+      "step": 18493
+    },
+    {
+      "epoch": 36.988,
+      "grad_norm": 1.385727882385254,
+      "learning_rate": 2e-05,
+      "loss": 0.05194426,
+      "step": 18494
+    },
+    {
+      "epoch": 36.99,
+      "grad_norm": 1.17097008228302,
+      "learning_rate": 2e-05,
+      "loss": 0.05635056,
+      "step": 18495
+    },
+    {
+      "epoch": 36.992,
+      "grad_norm": 1.1106641292572021,
+      "learning_rate": 2e-05,
+      "loss": 0.04129127,
+      "step": 18496
+    },
+    {
+      "epoch": 36.994,
+      "grad_norm": 1.0685235261917114,
+      "learning_rate": 2e-05,
+      "loss": 0.04287387,
+      "step": 18497
+    },
+    {
+      "epoch": 36.996,
+      "grad_norm": 1.188862919807434,
+      "learning_rate": 2e-05,
+      "loss": 0.04415998,
+      "step": 18498
+    },
+    {
+      "epoch": 36.998,
+      "grad_norm": 0.956224799156189,
+      "learning_rate": 2e-05,
+      "loss": 0.03282576,
+      "step": 18499
+    },
+    {
+      "epoch": 37.0,
+      "grad_norm": 1.2222694158554077,
+      "learning_rate": 2e-05,
+      "loss": 0.0414125,
+      "step": 18500
+    },
+    {
+      "epoch": 37.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.998,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9880239520958084,
+        "Equal_1": 0.994,
+        "Equal_2": 0.9880239520958084,
+        "Equal_3": 0.9900199600798403,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9940119760479041,
+        "Parallel_1": 0.9899799599198397,
+        "Parallel_2": 0.9959919839679359,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 0.996,
+        "Perpendicular_2": 0.988,
+        "Perpendicular_3": 0.9018036072144289,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 0.996,
+        "PointLiesOnCircle_3": 0.996,
+        "PointLiesOnLine_1": 1.0,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9880239520958084
+      },
+      "eval_runtime": 323.3803,
+      "eval_samples_per_second": 32.47,
+      "eval_steps_per_second": 0.649,
+      "step": 18500
+    },
+    {
+      "epoch": 37.002,
+      "grad_norm": 1.2086259126663208,
+      "learning_rate": 2e-05,
+      "loss": 0.05474807,
+      "step": 18501
+    },
+    {
+      "epoch": 37.004,
+      "grad_norm": 1.1692421436309814,
+      "learning_rate": 2e-05,
+      "loss": 0.04277514,
+      "step": 18502
+    },
+    {
+      "epoch": 37.006,
+      "grad_norm": 1.1070613861083984,
+      "learning_rate": 2e-05,
+      "loss": 0.03776816,
+      "step": 18503
+    },
+    {
+      "epoch": 37.008,
+      "grad_norm": 1.263321876525879,
+      "learning_rate": 2e-05,
+      "loss": 0.04286732,
+      "step": 18504
+    },
+    {
+      "epoch": 37.01,
+      "grad_norm": 1.0955123901367188,
+      "learning_rate": 2e-05,
+      "loss": 0.02977836,
+      "step": 18505
+    },
+    {
+      "epoch": 37.012,
+      "grad_norm": 1.2032184600830078,
+      "learning_rate": 2e-05,
+      "loss": 0.03929453,
+      "step": 18506
+    },
+    {
+      "epoch": 37.014,
+      "grad_norm": 1.0266116857528687,
+      "learning_rate": 2e-05,
+      "loss": 0.02832175,
+      "step": 18507
+    },
+    {
+      "epoch": 37.016,
+      "grad_norm": 1.862282633781433,
+      "learning_rate": 2e-05,
+      "loss": 0.04781446,
+      "step": 18508
+    },
+    {
+      "epoch": 37.018,
+      "grad_norm": 1.4959344863891602,
+      "learning_rate": 2e-05,
+      "loss": 0.05474871,
+      "step": 18509
+    },
+    {
+      "epoch": 37.02,
+      "grad_norm": 2.0433266162872314,
+      "learning_rate": 2e-05,
+      "loss": 0.05582491,
+      "step": 18510
+    },
+    {
+      "epoch": 37.022,
+      "grad_norm": 1.0982203483581543,
+      "learning_rate": 2e-05,
+      "loss": 0.03969976,
+      "step": 18511
+    },
+    {
+      "epoch": 37.024,
+      "grad_norm": 1.8421516418457031,
+      "learning_rate": 2e-05,
+      "loss": 0.05257302,
+      "step": 18512
+    },
+    {
+      "epoch": 37.026,
+      "grad_norm": 1.0704379081726074,
+      "learning_rate": 2e-05,
+      "loss": 0.03272648,
+      "step": 18513
+    },
+    {
+      "epoch": 37.028,
+      "grad_norm": 1.063231110572815,
+      "learning_rate": 2e-05,
+      "loss": 0.03291452,
+      "step": 18514
+    },
+    {
+      "epoch": 37.03,
+      "grad_norm": 1.1486153602600098,
+      "learning_rate": 2e-05,
+      "loss": 0.04854581,
+      "step": 18515
+    },
+    {
+      "epoch": 37.032,
+      "grad_norm": 1.0761146545410156,
+      "learning_rate": 2e-05,
+      "loss": 0.04210125,
+      "step": 18516
+    },
+    {
+      "epoch": 37.034,
+      "grad_norm": 1.477710247039795,
+      "learning_rate": 2e-05,
+      "loss": 0.05219606,
+      "step": 18517
+    },
+    {
+      "epoch": 37.036,
+      "grad_norm": 1.0120596885681152,
+      "learning_rate": 2e-05,
+      "loss": 0.02972842,
+      "step": 18518
+    },
+    {
+      "epoch": 37.038,
+      "grad_norm": 1.341340184211731,
+      "learning_rate": 2e-05,
+      "loss": 0.03905867,
+      "step": 18519
+    },
+    {
+      "epoch": 37.04,
+      "grad_norm": 1.8006370067596436,
+      "learning_rate": 2e-05,
+      "loss": 0.04028159,
+      "step": 18520
+    },
+    {
+      "epoch": 37.042,
+      "grad_norm": 1.06367027759552,
+      "learning_rate": 2e-05,
+      "loss": 0.0451144,
+      "step": 18521
+    },
+    {
+      "epoch": 37.044,
+      "grad_norm": 1.3736926317214966,
+      "learning_rate": 2e-05,
+      "loss": 0.05268521,
+      "step": 18522
+    },
+    {
+      "epoch": 37.046,
+      "grad_norm": 3.3430514335632324,
+      "learning_rate": 2e-05,
+      "loss": 0.05855541,
+      "step": 18523
+    },
+    {
+      "epoch": 37.048,
+      "grad_norm": 1.0005395412445068,
+      "learning_rate": 2e-05,
+      "loss": 0.03178389,
+      "step": 18524
+    },
+    {
+      "epoch": 37.05,
+      "grad_norm": 1.4999123811721802,
+      "learning_rate": 2e-05,
+      "loss": 0.03547132,
+      "step": 18525
+    },
+    {
+      "epoch": 37.052,
+      "grad_norm": 1.2250981330871582,
+      "learning_rate": 2e-05,
+      "loss": 0.04226706,
+      "step": 18526
+    },
+    {
+      "epoch": 37.054,
+      "grad_norm": 1.1662254333496094,
+      "learning_rate": 2e-05,
+      "loss": 0.03948696,
+      "step": 18527
+    },
+    {
+      "epoch": 37.056,
+      "grad_norm": 1.1504510641098022,
+      "learning_rate": 2e-05,
+      "loss": 0.04601368,
+      "step": 18528
+    },
+    {
+      "epoch": 37.058,
+      "grad_norm": 1.0720382928848267,
+      "learning_rate": 2e-05,
+      "loss": 0.03848777,
+      "step": 18529
+    },
+    {
+      "epoch": 37.06,
+      "grad_norm": 1.3089898824691772,
+      "learning_rate": 2e-05,
+      "loss": 0.03624338,
+      "step": 18530
+    },
+    {
+      "epoch": 37.062,
+      "grad_norm": 1.2138144969940186,
+      "learning_rate": 2e-05,
+      "loss": 0.0507899,
+      "step": 18531
+    },
+    {
+      "epoch": 37.064,
+      "grad_norm": 1.297494888305664,
+      "learning_rate": 2e-05,
+      "loss": 0.04721798,
+      "step": 18532
+    },
+    {
+      "epoch": 37.066,
+      "grad_norm": 1.2399834394454956,
+      "learning_rate": 2e-05,
+      "loss": 0.04290456,
+      "step": 18533
+    },
+    {
+      "epoch": 37.068,
+      "grad_norm": 1.258151888847351,
+      "learning_rate": 2e-05,
+      "loss": 0.03323229,
+      "step": 18534
+    },
+    {
+      "epoch": 37.07,
+      "grad_norm": 2.544591188430786,
+      "learning_rate": 2e-05,
+      "loss": 0.04624221,
+      "step": 18535
+    },
+    {
+      "epoch": 37.072,
+      "grad_norm": 1.0994566679000854,
+      "learning_rate": 2e-05,
+      "loss": 0.03203875,
+      "step": 18536
+    },
+    {
+      "epoch": 37.074,
+      "grad_norm": 1.5202358961105347,
+      "learning_rate": 2e-05,
+      "loss": 0.04309528,
+      "step": 18537
+    },
+    {
+      "epoch": 37.076,
+      "grad_norm": 1.162284255027771,
+      "learning_rate": 2e-05,
+      "loss": 0.04106382,
+      "step": 18538
+    },
+    {
+      "epoch": 37.078,
+      "grad_norm": 1.5119906663894653,
+      "learning_rate": 2e-05,
+      "loss": 0.05221222,
+      "step": 18539
+    },
+    {
+      "epoch": 37.08,
+      "grad_norm": 1.0148025751113892,
+      "learning_rate": 2e-05,
+      "loss": 0.04121029,
+      "step": 18540
+    },
+    {
+      "epoch": 37.082,
+      "grad_norm": 1.2499613761901855,
+      "learning_rate": 2e-05,
+      "loss": 0.05167337,
+      "step": 18541
+    },
+    {
+      "epoch": 37.084,
+      "grad_norm": 1.3583709001541138,
+      "learning_rate": 2e-05,
+      "loss": 0.04471046,
+      "step": 18542
+    },
+    {
+      "epoch": 37.086,
+      "grad_norm": 2.9629805088043213,
+      "learning_rate": 2e-05,
+      "loss": 0.04842234,
+      "step": 18543
+    },
+    {
+      "epoch": 37.088,
+      "grad_norm": 1.327346920967102,
+      "learning_rate": 2e-05,
+      "loss": 0.05041461,
+      "step": 18544
+    },
+    {
+      "epoch": 37.09,
+      "grad_norm": 1.7716262340545654,
+      "learning_rate": 2e-05,
+      "loss": 0.03213685,
+      "step": 18545
+    },
+    {
+      "epoch": 37.092,
+      "grad_norm": 1.3329129219055176,
+      "learning_rate": 2e-05,
+      "loss": 0.05127243,
+      "step": 18546
+    },
+    {
+      "epoch": 37.094,
+      "grad_norm": 1.2340625524520874,
+      "learning_rate": 2e-05,
+      "loss": 0.04882501,
+      "step": 18547
+    },
+    {
+      "epoch": 37.096,
+      "grad_norm": 1.5020076036453247,
+      "learning_rate": 2e-05,
+      "loss": 0.05312416,
+      "step": 18548
+    },
+    {
+      "epoch": 37.098,
+      "grad_norm": 1.0506970882415771,
+      "learning_rate": 2e-05,
+      "loss": 0.04135638,
+      "step": 18549
+    },
+    {
+      "epoch": 37.1,
+      "grad_norm": 0.8881466388702393,
+      "learning_rate": 2e-05,
+      "loss": 0.02312918,
+      "step": 18550
+    },
+    {
+      "epoch": 37.102,
+      "grad_norm": 2.499708652496338,
+      "learning_rate": 2e-05,
+      "loss": 0.05453001,
+      "step": 18551
+    },
+    {
+      "epoch": 37.104,
+      "grad_norm": 0.9243218302726746,
+      "learning_rate": 2e-05,
+      "loss": 0.02658354,
+      "step": 18552
+    },
+    {
+      "epoch": 37.106,
+      "grad_norm": 0.8885984420776367,
+      "learning_rate": 2e-05,
+      "loss": 0.03515668,
+      "step": 18553
+    },
+    {
+      "epoch": 37.108,
+      "grad_norm": 1.3641797304153442,
+      "learning_rate": 2e-05,
+      "loss": 0.05175686,
+      "step": 18554
+    },
+    {
+      "epoch": 37.11,
+      "grad_norm": 1.5672239065170288,
+      "learning_rate": 2e-05,
+      "loss": 0.05572821,
+      "step": 18555
+    },
+    {
+      "epoch": 37.112,
+      "grad_norm": 1.1886248588562012,
+      "learning_rate": 2e-05,
+      "loss": 0.03903022,
+      "step": 18556
+    },
+    {
+      "epoch": 37.114,
+      "grad_norm": 2.1297879219055176,
+      "learning_rate": 2e-05,
+      "loss": 0.04060325,
+      "step": 18557
+    },
+    {
+      "epoch": 37.116,
+      "grad_norm": 1.0565165281295776,
+      "learning_rate": 2e-05,
+      "loss": 0.03858557,
+      "step": 18558
+    },
+    {
+      "epoch": 37.118,
+      "grad_norm": 0.9887270927429199,
+      "learning_rate": 2e-05,
+      "loss": 0.0331764,
+      "step": 18559
+    },
+    {
+      "epoch": 37.12,
+      "grad_norm": 1.3902815580368042,
+      "learning_rate": 2e-05,
+      "loss": 0.04316443,
+      "step": 18560
+    },
+    {
+      "epoch": 37.122,
+      "grad_norm": 1.7422220706939697,
+      "learning_rate": 2e-05,
+      "loss": 0.05364904,
+      "step": 18561
+    },
+    {
+      "epoch": 37.124,
+      "grad_norm": 1.6973824501037598,
+      "learning_rate": 2e-05,
+      "loss": 0.06084717,
+      "step": 18562
+    },
+    {
+      "epoch": 37.126,
+      "grad_norm": 1.2839064598083496,
+      "learning_rate": 2e-05,
+      "loss": 0.04210863,
+      "step": 18563
+    },
+    {
+      "epoch": 37.128,
+      "grad_norm": 1.687252402305603,
+      "learning_rate": 2e-05,
+      "loss": 0.03391319,
+      "step": 18564
+    },
+    {
+      "epoch": 37.13,
+      "grad_norm": 1.2945353984832764,
+      "learning_rate": 2e-05,
+      "loss": 0.04598271,
+      "step": 18565
+    },
+    {
+      "epoch": 37.132,
+      "grad_norm": 1.3326228857040405,
+      "learning_rate": 2e-05,
+      "loss": 0.04093219,
+      "step": 18566
+    },
+    {
+      "epoch": 37.134,
+      "grad_norm": 2.1192924976348877,
+      "learning_rate": 2e-05,
+      "loss": 0.03747287,
+      "step": 18567
+    },
+    {
+      "epoch": 37.136,
+      "grad_norm": 0.9054780006408691,
+      "learning_rate": 2e-05,
+      "loss": 0.02849467,
+      "step": 18568
+    },
+    {
+      "epoch": 37.138,
+      "grad_norm": 1.0603820085525513,
+      "learning_rate": 2e-05,
+      "loss": 0.0402334,
+      "step": 18569
+    },
+    {
+      "epoch": 37.14,
+      "grad_norm": 1.367003083229065,
+      "learning_rate": 2e-05,
+      "loss": 0.05233489,
+      "step": 18570
+    },
+    {
+      "epoch": 37.142,
+      "grad_norm": 0.9761447906494141,
+      "learning_rate": 2e-05,
+      "loss": 0.03226305,
+      "step": 18571
+    },
+    {
+      "epoch": 37.144,
+      "grad_norm": 1.0491434335708618,
+      "learning_rate": 2e-05,
+      "loss": 0.03535546,
+      "step": 18572
+    },
+    {
+      "epoch": 37.146,
+      "grad_norm": 1.1767827272415161,
+      "learning_rate": 2e-05,
+      "loss": 0.0434572,
+      "step": 18573
+    },
+    {
+      "epoch": 37.148,
+      "grad_norm": 1.0024957656860352,
+      "learning_rate": 2e-05,
+      "loss": 0.03991302,
+      "step": 18574
+    },
+    {
+      "epoch": 37.15,
+      "grad_norm": 1.4435670375823975,
+      "learning_rate": 2e-05,
+      "loss": 0.04887334,
+      "step": 18575
+    },
+    {
+      "epoch": 37.152,
+      "grad_norm": 1.0476168394088745,
+      "learning_rate": 2e-05,
+      "loss": 0.03045881,
+      "step": 18576
+    },
+    {
+      "epoch": 37.154,
+      "grad_norm": 3.0771420001983643,
+      "learning_rate": 2e-05,
+      "loss": 0.05882487,
+      "step": 18577
+    },
+    {
+      "epoch": 37.156,
+      "grad_norm": 1.1227463483810425,
+      "learning_rate": 2e-05,
+      "loss": 0.04296419,
+      "step": 18578
+    },
+    {
+      "epoch": 37.158,
+      "grad_norm": 0.8635389804840088,
+      "learning_rate": 2e-05,
+      "loss": 0.02800985,
+      "step": 18579
+    },
+    {
+      "epoch": 37.16,
+      "grad_norm": 1.0923333168029785,
+      "learning_rate": 2e-05,
+      "loss": 0.03294976,
+      "step": 18580
+    },
+    {
+      "epoch": 37.162,
+      "grad_norm": 1.2803630828857422,
+      "learning_rate": 2e-05,
+      "loss": 0.05049956,
+      "step": 18581
+    },
+    {
+      "epoch": 37.164,
+      "grad_norm": 2.036414384841919,
+      "learning_rate": 2e-05,
+      "loss": 0.03803521,
+      "step": 18582
+    },
+    {
+      "epoch": 37.166,
+      "grad_norm": 1.534833312034607,
+      "learning_rate": 2e-05,
+      "loss": 0.03264375,
+      "step": 18583
+    },
+    {
+      "epoch": 37.168,
+      "grad_norm": 1.0317829847335815,
+      "learning_rate": 2e-05,
+      "loss": 0.02741832,
+      "step": 18584
+    },
+    {
+      "epoch": 37.17,
+      "grad_norm": 1.0632944107055664,
+      "learning_rate": 2e-05,
+      "loss": 0.03481572,
+      "step": 18585
+    },
+    {
+      "epoch": 37.172,
+      "grad_norm": 1.263419270515442,
+      "learning_rate": 2e-05,
+      "loss": 0.04441663,
+      "step": 18586
+    },
+    {
+      "epoch": 37.174,
+      "grad_norm": 1.0531634092330933,
+      "learning_rate": 2e-05,
+      "loss": 0.03552748,
+      "step": 18587
+    },
+    {
+      "epoch": 37.176,
+      "grad_norm": 1.569259524345398,
+      "learning_rate": 2e-05,
+      "loss": 0.05546115,
+      "step": 18588
+    },
+    {
+      "epoch": 37.178,
+      "grad_norm": 1.3170031309127808,
+      "learning_rate": 2e-05,
+      "loss": 0.05026477,
+      "step": 18589
+    },
+    {
+      "epoch": 37.18,
+      "grad_norm": 1.1640214920043945,
+      "learning_rate": 2e-05,
+      "loss": 0.04955488,
+      "step": 18590
+    },
+    {
+      "epoch": 37.182,
+      "grad_norm": 1.6744143962860107,
+      "learning_rate": 2e-05,
+      "loss": 0.03792101,
+      "step": 18591
+    },
+    {
+      "epoch": 37.184,
+      "grad_norm": 0.9118578433990479,
+      "learning_rate": 2e-05,
+      "loss": 0.02556996,
+      "step": 18592
+    },
+    {
+      "epoch": 37.186,
+      "grad_norm": 1.644417643547058,
+      "learning_rate": 2e-05,
+      "loss": 0.04530199,
+      "step": 18593
+    },
+    {
+      "epoch": 37.188,
+      "grad_norm": 1.7604995965957642,
+      "learning_rate": 2e-05,
+      "loss": 0.06252389,
+      "step": 18594
+    },
+    {
+      "epoch": 37.19,
+      "grad_norm": 1.0273133516311646,
+      "learning_rate": 2e-05,
+      "loss": 0.04975293,
+      "step": 18595
+    },
+    {
+      "epoch": 37.192,
+      "grad_norm": 1.05352783203125,
+      "learning_rate": 2e-05,
+      "loss": 0.04171479,
+      "step": 18596
+    },
+    {
+      "epoch": 37.194,
+      "grad_norm": 0.9587264060974121,
+      "learning_rate": 2e-05,
+      "loss": 0.03576777,
+      "step": 18597
+    },
+    {
+      "epoch": 37.196,
+      "grad_norm": 1.094084620475769,
+      "learning_rate": 2e-05,
+      "loss": 0.02543401,
+      "step": 18598
+    },
+    {
+      "epoch": 37.198,
+      "grad_norm": 1.139223575592041,
+      "learning_rate": 2e-05,
+      "loss": 0.03940034,
+      "step": 18599
+    },
+    {
+      "epoch": 37.2,
+      "grad_norm": 1.050518274307251,
+      "learning_rate": 2e-05,
+      "loss": 0.0513064,
+      "step": 18600
+    },
+    {
+      "epoch": 37.202,
+      "grad_norm": 1.5931733846664429,
+      "learning_rate": 2e-05,
+      "loss": 0.03497151,
+      "step": 18601
+    },
+    {
+      "epoch": 37.204,
+      "grad_norm": 1.2428843975067139,
+      "learning_rate": 2e-05,
+      "loss": 0.04900038,
+      "step": 18602
+    },
+    {
+      "epoch": 37.206,
+      "grad_norm": 5.434542179107666,
+      "learning_rate": 2e-05,
+      "loss": 0.05245988,
+      "step": 18603
+    },
+    {
+      "epoch": 37.208,
+      "grad_norm": 1.1642683744430542,
+      "learning_rate": 2e-05,
+      "loss": 0.05161486,
+      "step": 18604
+    },
+    {
+      "epoch": 37.21,
+      "grad_norm": 1.0116956233978271,
+      "learning_rate": 2e-05,
+      "loss": 0.03130523,
+      "step": 18605
+    },
+    {
+      "epoch": 37.212,
+      "grad_norm": 1.7632371187210083,
+      "learning_rate": 2e-05,
+      "loss": 0.04773512,
+      "step": 18606
+    },
+    {
+      "epoch": 37.214,
+      "grad_norm": 0.9744578003883362,
+      "learning_rate": 2e-05,
+      "loss": 0.03005046,
+      "step": 18607
+    },
+    {
+      "epoch": 37.216,
+      "grad_norm": 1.5116866827011108,
+      "learning_rate": 2e-05,
+      "loss": 0.04617791,
+      "step": 18608
+    },
+    {
+      "epoch": 37.218,
+      "grad_norm": 2.665987730026245,
+      "learning_rate": 2e-05,
+      "loss": 0.03156227,
+      "step": 18609
+    },
+    {
+      "epoch": 37.22,
+      "grad_norm": 1.0697249174118042,
+      "learning_rate": 2e-05,
+      "loss": 0.03917909,
+      "step": 18610
+    },
+    {
+      "epoch": 37.222,
+      "grad_norm": 1.0376712083816528,
+      "learning_rate": 2e-05,
+      "loss": 0.04070775,
+      "step": 18611
+    },
+    {
+      "epoch": 37.224,
+      "grad_norm": 1.1940553188323975,
+      "learning_rate": 2e-05,
+      "loss": 0.03836515,
+      "step": 18612
+    },
+    {
+      "epoch": 37.226,
+      "grad_norm": 0.9701217412948608,
+      "learning_rate": 2e-05,
+      "loss": 0.03590652,
+      "step": 18613
+    },
+    {
+      "epoch": 37.228,
+      "grad_norm": 1.2851488590240479,
+      "learning_rate": 2e-05,
+      "loss": 0.04535402,
+      "step": 18614
+    },
+    {
+      "epoch": 37.23,
+      "grad_norm": 2.4411749839782715,
+      "learning_rate": 2e-05,
+      "loss": 0.06194766,
+      "step": 18615
+    },
+    {
+      "epoch": 37.232,
+      "grad_norm": 1.3536734580993652,
+      "learning_rate": 2e-05,
+      "loss": 0.04599462,
+      "step": 18616
+    },
+    {
+      "epoch": 37.234,
+      "grad_norm": 1.2606672048568726,
+      "learning_rate": 2e-05,
+      "loss": 0.04580709,
+      "step": 18617
+    },
+    {
+      "epoch": 37.236,
+      "grad_norm": 1.3303290605545044,
+      "learning_rate": 2e-05,
+      "loss": 0.04706927,
+      "step": 18618
+    },
+    {
+      "epoch": 37.238,
+      "grad_norm": 1.4099715948104858,
+      "learning_rate": 2e-05,
+      "loss": 0.02947021,
+      "step": 18619
+    },
+    {
+      "epoch": 37.24,
+      "grad_norm": 1.0826631784439087,
+      "learning_rate": 2e-05,
+      "loss": 0.03750202,
+      "step": 18620
+    },
+    {
+      "epoch": 37.242,
+      "grad_norm": 1.1417129039764404,
+      "learning_rate": 2e-05,
+      "loss": 0.05023696,
+      "step": 18621
+    },
+    {
+      "epoch": 37.244,
+      "grad_norm": 1.0746430158615112,
+      "learning_rate": 2e-05,
+      "loss": 0.04541557,
+      "step": 18622
+    },
+    {
+      "epoch": 37.246,
+      "grad_norm": 1.4391956329345703,
+      "learning_rate": 2e-05,
+      "loss": 0.05088585,
+      "step": 18623
+    },
+    {
+      "epoch": 37.248,
+      "grad_norm": 1.7742950916290283,
+      "learning_rate": 2e-05,
+      "loss": 0.04011762,
+      "step": 18624
+    },
+    {
+      "epoch": 37.25,
+      "grad_norm": 2.4555623531341553,
+      "learning_rate": 2e-05,
+      "loss": 0.04191405,
+      "step": 18625
+    },
+    {
+      "epoch": 37.252,
+      "grad_norm": 1.0835415124893188,
+      "learning_rate": 2e-05,
+      "loss": 0.04408552,
+      "step": 18626
+    },
+    {
+      "epoch": 37.254,
+      "grad_norm": 1.494162917137146,
+      "learning_rate": 2e-05,
+      "loss": 0.0393875,
+      "step": 18627
+    },
+    {
+      "epoch": 37.256,
+      "grad_norm": 0.9933038353919983,
+      "learning_rate": 2e-05,
+      "loss": 0.0362031,
+      "step": 18628
+    },
+    {
+      "epoch": 37.258,
+      "grad_norm": 1.1506096124649048,
+      "learning_rate": 2e-05,
+      "loss": 0.0375497,
+      "step": 18629
+    },
+    {
+      "epoch": 37.26,
+      "grad_norm": 1.341207504272461,
+      "learning_rate": 2e-05,
+      "loss": 0.04895544,
+      "step": 18630
+    },
+    {
+      "epoch": 37.262,
+      "grad_norm": 1.4760468006134033,
+      "learning_rate": 2e-05,
+      "loss": 0.04239687,
+      "step": 18631
+    },
+    {
+      "epoch": 37.264,
+      "grad_norm": 1.1099522113800049,
+      "learning_rate": 2e-05,
+      "loss": 0.05547397,
+      "step": 18632
+    },
+    {
+      "epoch": 37.266,
+      "grad_norm": 1.1346218585968018,
+      "learning_rate": 2e-05,
+      "loss": 0.04982274,
+      "step": 18633
+    },
+    {
+      "epoch": 37.268,
+      "grad_norm": 0.9855700731277466,
+      "learning_rate": 2e-05,
+      "loss": 0.03562504,
+      "step": 18634
+    },
+    {
+      "epoch": 37.27,
+      "grad_norm": 1.561249017715454,
+      "learning_rate": 2e-05,
+      "loss": 0.05814129,
+      "step": 18635
+    },
+    {
+      "epoch": 37.272,
+      "grad_norm": 1.0120232105255127,
+      "learning_rate": 2e-05,
+      "loss": 0.0441656,
+      "step": 18636
+    },
+    {
+      "epoch": 37.274,
+      "grad_norm": 3.5975611209869385,
+      "learning_rate": 2e-05,
+      "loss": 0.03757409,
+      "step": 18637
+    },
+    {
+      "epoch": 37.276,
+      "grad_norm": 1.7085171937942505,
+      "learning_rate": 2e-05,
+      "loss": 0.03554228,
+      "step": 18638
+    },
+    {
+      "epoch": 37.278,
+      "grad_norm": 1.2262259721755981,
+      "learning_rate": 2e-05,
+      "loss": 0.02165484,
+      "step": 18639
+    },
+    {
+      "epoch": 37.28,
+      "grad_norm": 1.2670469284057617,
+      "learning_rate": 2e-05,
+      "loss": 0.03860283,
+      "step": 18640
+    },
+    {
+      "epoch": 37.282,
+      "grad_norm": 1.1696661710739136,
+      "learning_rate": 2e-05,
+      "loss": 0.04609601,
+      "step": 18641
+    },
+    {
+      "epoch": 37.284,
+      "grad_norm": 2.9907474517822266,
+      "learning_rate": 2e-05,
+      "loss": 0.05883395,
+      "step": 18642
+    },
+    {
+      "epoch": 37.286,
+      "grad_norm": 1.062591552734375,
+      "learning_rate": 2e-05,
+      "loss": 0.04270145,
+      "step": 18643
+    },
+    {
+      "epoch": 37.288,
+      "grad_norm": 1.022788166999817,
+      "learning_rate": 2e-05,
+      "loss": 0.02788961,
+      "step": 18644
+    },
+    {
+      "epoch": 37.29,
+      "grad_norm": 1.3785815238952637,
+      "learning_rate": 2e-05,
+      "loss": 0.04528718,
+      "step": 18645
+    },
+    {
+      "epoch": 37.292,
+      "grad_norm": 1.0437957048416138,
+      "learning_rate": 2e-05,
+      "loss": 0.03974991,
+      "step": 18646
+    },
+    {
+      "epoch": 37.294,
+      "grad_norm": 1.1609976291656494,
+      "learning_rate": 2e-05,
+      "loss": 0.05013696,
+      "step": 18647
+    },
+    {
+      "epoch": 37.296,
+      "grad_norm": 0.8697720170021057,
+      "learning_rate": 2e-05,
+      "loss": 0.03053538,
+      "step": 18648
+    },
+    {
+      "epoch": 37.298,
+      "grad_norm": 0.9573113918304443,
+      "learning_rate": 2e-05,
+      "loss": 0.03847982,
+      "step": 18649
+    },
+    {
+      "epoch": 37.3,
+      "grad_norm": 1.183646559715271,
+      "learning_rate": 2e-05,
+      "loss": 0.04569777,
+      "step": 18650
+    },
+    {
+      "epoch": 37.302,
+      "grad_norm": 0.8802410364151001,
+      "learning_rate": 2e-05,
+      "loss": 0.02939706,
+      "step": 18651
+    },
+    {
+      "epoch": 37.304,
+      "grad_norm": 1.2559164762496948,
+      "learning_rate": 2e-05,
+      "loss": 0.0539922,
+      "step": 18652
+    },
+    {
+      "epoch": 37.306,
+      "grad_norm": 1.3703943490982056,
+      "learning_rate": 2e-05,
+      "loss": 0.04603429,
+      "step": 18653
+    },
+    {
+      "epoch": 37.308,
+      "grad_norm": 1.1752663850784302,
+      "learning_rate": 2e-05,
+      "loss": 0.05147523,
+      "step": 18654
+    },
+    {
+      "epoch": 37.31,
+      "grad_norm": 1.039732575416565,
+      "learning_rate": 2e-05,
+      "loss": 0.03672284,
+      "step": 18655
+    },
+    {
+      "epoch": 37.312,
+      "grad_norm": 1.0590288639068604,
+      "learning_rate": 2e-05,
+      "loss": 0.04019672,
+      "step": 18656
+    },
+    {
+      "epoch": 37.314,
+      "grad_norm": 1.8295365571975708,
+      "learning_rate": 2e-05,
+      "loss": 0.05494248,
+      "step": 18657
+    },
+    {
+      "epoch": 37.316,
+      "grad_norm": 1.3013070821762085,
+      "learning_rate": 2e-05,
+      "loss": 0.03197949,
+      "step": 18658
+    },
+    {
+      "epoch": 37.318,
+      "grad_norm": 1.4694881439208984,
+      "learning_rate": 2e-05,
+      "loss": 0.02851127,
+      "step": 18659
+    },
+    {
+      "epoch": 37.32,
+      "grad_norm": 1.1061211824417114,
+      "learning_rate": 2e-05,
+      "loss": 0.03776077,
+      "step": 18660
+    },
+    {
+      "epoch": 37.322,
+      "grad_norm": 0.9431129097938538,
+      "learning_rate": 2e-05,
+      "loss": 0.03077835,
+      "step": 18661
+    },
+    {
+      "epoch": 37.324,
+      "grad_norm": 0.9578332901000977,
+      "learning_rate": 2e-05,
+      "loss": 0.02925043,
+      "step": 18662
+    },
+    {
+      "epoch": 37.326,
+      "grad_norm": 1.0546810626983643,
+      "learning_rate": 2e-05,
+      "loss": 0.04426049,
+      "step": 18663
+    },
+    {
+      "epoch": 37.328,
+      "grad_norm": 1.1375724077224731,
+      "learning_rate": 2e-05,
+      "loss": 0.02645444,
+      "step": 18664
+    },
+    {
+      "epoch": 37.33,
+      "grad_norm": 1.3067609071731567,
+      "learning_rate": 2e-05,
+      "loss": 0.04584169,
+      "step": 18665
+    },
+    {
+      "epoch": 37.332,
+      "grad_norm": 2.0529701709747314,
+      "learning_rate": 2e-05,
+      "loss": 0.04706659,
+      "step": 18666
+    },
+    {
+      "epoch": 37.334,
+      "grad_norm": 1.4616862535476685,
+      "learning_rate": 2e-05,
+      "loss": 0.04822695,
+      "step": 18667
+    },
+    {
+      "epoch": 37.336,
+      "grad_norm": 1.1066008806228638,
+      "learning_rate": 2e-05,
+      "loss": 0.03952828,
+      "step": 18668
+    },
+    {
+      "epoch": 37.338,
+      "grad_norm": 2.8666927814483643,
+      "learning_rate": 2e-05,
+      "loss": 0.04431106,
+      "step": 18669
+    },
+    {
+      "epoch": 37.34,
+      "grad_norm": 1.526808500289917,
+      "learning_rate": 2e-05,
+      "loss": 0.04228971,
+      "step": 18670
+    },
+    {
+      "epoch": 37.342,
+      "grad_norm": 1.7227466106414795,
+      "learning_rate": 2e-05,
+      "loss": 0.03985681,
+      "step": 18671
+    },
+    {
+      "epoch": 37.344,
+      "grad_norm": 0.960451602935791,
+      "learning_rate": 2e-05,
+      "loss": 0.02552288,
+      "step": 18672
+    },
+    {
+      "epoch": 37.346,
+      "grad_norm": 1.1226481199264526,
+      "learning_rate": 2e-05,
+      "loss": 0.04143129,
+      "step": 18673
+    },
+    {
+      "epoch": 37.348,
+      "grad_norm": 2.2680718898773193,
+      "learning_rate": 2e-05,
+      "loss": 0.05270278,
+      "step": 18674
+    },
+    {
+      "epoch": 37.35,
+      "grad_norm": 0.8541618585586548,
+      "learning_rate": 2e-05,
+      "loss": 0.0224099,
+      "step": 18675
+    },
+    {
+      "epoch": 37.352,
+      "grad_norm": 1.2834433317184448,
+      "learning_rate": 2e-05,
+      "loss": 0.05758996,
+      "step": 18676
+    },
+    {
+      "epoch": 37.354,
+      "grad_norm": 1.2858184576034546,
+      "learning_rate": 2e-05,
+      "loss": 0.03408124,
+      "step": 18677
+    },
+    {
+      "epoch": 37.356,
+      "grad_norm": 1.2244290113449097,
+      "learning_rate": 2e-05,
+      "loss": 0.04213716,
+      "step": 18678
+    },
+    {
+      "epoch": 37.358,
+      "grad_norm": 1.783223271369934,
+      "learning_rate": 2e-05,
+      "loss": 0.05025373,
+      "step": 18679
+    },
+    {
+      "epoch": 37.36,
+      "grad_norm": 1.0761710405349731,
+      "learning_rate": 2e-05,
+      "loss": 0.04476252,
+      "step": 18680
+    },
+    {
+      "epoch": 37.362,
+      "grad_norm": 0.9174004197120667,
+      "learning_rate": 2e-05,
+      "loss": 0.02644821,
+      "step": 18681
+    },
+    {
+      "epoch": 37.364,
+      "grad_norm": 1.5571218729019165,
+      "learning_rate": 2e-05,
+      "loss": 0.03530311,
+      "step": 18682
+    },
+    {
+      "epoch": 37.366,
+      "grad_norm": 1.3173847198486328,
+      "learning_rate": 2e-05,
+      "loss": 0.0494157,
+      "step": 18683
+    },
+    {
+      "epoch": 37.368,
+      "grad_norm": 3.338667869567871,
+      "learning_rate": 2e-05,
+      "loss": 0.07118287,
+      "step": 18684
+    },
+    {
+      "epoch": 37.37,
+      "grad_norm": 1.157820701599121,
+      "learning_rate": 2e-05,
+      "loss": 0.0302717,
+      "step": 18685
+    },
+    {
+      "epoch": 37.372,
+      "grad_norm": 0.9768419861793518,
+      "learning_rate": 2e-05,
+      "loss": 0.04007539,
+      "step": 18686
+    },
+    {
+      "epoch": 37.374,
+      "grad_norm": 1.2794311046600342,
+      "learning_rate": 2e-05,
+      "loss": 0.04950311,
+      "step": 18687
+    },
+    {
+      "epoch": 37.376,
+      "grad_norm": 1.141780972480774,
+      "learning_rate": 2e-05,
+      "loss": 0.03149645,
+      "step": 18688
+    },
+    {
+      "epoch": 37.378,
+      "grad_norm": 2.4156718254089355,
+      "learning_rate": 2e-05,
+      "loss": 0.04496891,
+      "step": 18689
+    },
+    {
+      "epoch": 37.38,
+      "grad_norm": 0.9722008109092712,
+      "learning_rate": 2e-05,
+      "loss": 0.0443184,
+      "step": 18690
+    },
+    {
+      "epoch": 37.382,
+      "grad_norm": 1.1352218389511108,
+      "learning_rate": 2e-05,
+      "loss": 0.03867435,
+      "step": 18691
+    },
+    {
+      "epoch": 37.384,
+      "grad_norm": 1.627010464668274,
+      "learning_rate": 2e-05,
+      "loss": 0.06134836,
+      "step": 18692
+    },
+    {
+      "epoch": 37.386,
+      "grad_norm": 1.1376895904541016,
+      "learning_rate": 2e-05,
+      "loss": 0.03281328,
+      "step": 18693
+    },
+    {
+      "epoch": 37.388,
+      "grad_norm": 1.209214210510254,
+      "learning_rate": 2e-05,
+      "loss": 0.04794659,
+      "step": 18694
+    },
+    {
+      "epoch": 37.39,
+      "grad_norm": 1.2975291013717651,
+      "learning_rate": 2e-05,
+      "loss": 0.05038759,
+      "step": 18695
+    },
+    {
+      "epoch": 37.392,
+      "grad_norm": 1.3444797992706299,
+      "learning_rate": 2e-05,
+      "loss": 0.03968658,
+      "step": 18696
+    },
+    {
+      "epoch": 37.394,
+      "grad_norm": 1.1733344793319702,
+      "learning_rate": 2e-05,
+      "loss": 0.04530336,
+      "step": 18697
+    },
+    {
+      "epoch": 37.396,
+      "grad_norm": 0.9696661233901978,
+      "learning_rate": 2e-05,
+      "loss": 0.03088088,
+      "step": 18698
+    },
+    {
+      "epoch": 37.398,
+      "grad_norm": 1.0545789003372192,
+      "learning_rate": 2e-05,
+      "loss": 0.03587662,
+      "step": 18699
+    },
+    {
+      "epoch": 37.4,
+      "grad_norm": 1.031237006187439,
+      "learning_rate": 2e-05,
+      "loss": 0.03672967,
+      "step": 18700
+    },
+    {
+      "epoch": 37.402,
+      "grad_norm": 1.3680472373962402,
+      "learning_rate": 2e-05,
+      "loss": 0.05069418,
+      "step": 18701
+    },
+    {
+      "epoch": 37.404,
+      "grad_norm": 1.0634808540344238,
+      "learning_rate": 2e-05,
+      "loss": 0.04009442,
+      "step": 18702
+    },
+    {
+      "epoch": 37.406,
+      "grad_norm": 0.896915078163147,
+      "learning_rate": 2e-05,
+      "loss": 0.02137747,
+      "step": 18703
+    },
+    {
+      "epoch": 37.408,
+      "grad_norm": 1.0462524890899658,
+      "learning_rate": 2e-05,
+      "loss": 0.03831872,
+      "step": 18704
+    },
+    {
+      "epoch": 37.41,
+      "grad_norm": 1.36514151096344,
+      "learning_rate": 2e-05,
+      "loss": 0.05347653,
+      "step": 18705
+    },
+    {
+      "epoch": 37.412,
+      "grad_norm": 1.608876347541809,
+      "learning_rate": 2e-05,
+      "loss": 0.03230177,
+      "step": 18706
+    },
+    {
+      "epoch": 37.414,
+      "grad_norm": 1.5585176944732666,
+      "learning_rate": 2e-05,
+      "loss": 0.04574811,
+      "step": 18707
+    },
+    {
+      "epoch": 37.416,
+      "grad_norm": 1.5784344673156738,
+      "learning_rate": 2e-05,
+      "loss": 0.0467354,
+      "step": 18708
+    },
+    {
+      "epoch": 37.418,
+      "grad_norm": 1.5583195686340332,
+      "learning_rate": 2e-05,
+      "loss": 0.04075438,
+      "step": 18709
+    },
+    {
+      "epoch": 37.42,
+      "grad_norm": 1.0302294492721558,
+      "learning_rate": 2e-05,
+      "loss": 0.04351837,
+      "step": 18710
+    },
+    {
+      "epoch": 37.422,
+      "grad_norm": 1.7316805124282837,
+      "learning_rate": 2e-05,
+      "loss": 0.02402731,
+      "step": 18711
+    },
+    {
+      "epoch": 37.424,
+      "grad_norm": 1.5626541376113892,
+      "learning_rate": 2e-05,
+      "loss": 0.03232317,
+      "step": 18712
+    },
+    {
+      "epoch": 37.426,
+      "grad_norm": 0.9855737686157227,
+      "learning_rate": 2e-05,
+      "loss": 0.03887186,
+      "step": 18713
+    },
+    {
+      "epoch": 37.428,
+      "grad_norm": 1.4292187690734863,
+      "learning_rate": 2e-05,
+      "loss": 0.05129439,
+      "step": 18714
+    },
+    {
+      "epoch": 37.43,
+      "grad_norm": 1.3205820322036743,
+      "learning_rate": 2e-05,
+      "loss": 0.03848224,
+      "step": 18715
+    },
+    {
+      "epoch": 37.432,
+      "grad_norm": 1.474755048751831,
+      "learning_rate": 2e-05,
+      "loss": 0.0503183,
+      "step": 18716
+    },
+    {
+      "epoch": 37.434,
+      "grad_norm": 1.0868511199951172,
+      "learning_rate": 2e-05,
+      "loss": 0.02255207,
+      "step": 18717
+    },
+    {
+      "epoch": 37.436,
+      "grad_norm": 1.2556840181350708,
+      "learning_rate": 2e-05,
+      "loss": 0.05313279,
+      "step": 18718
+    },
+    {
+      "epoch": 37.438,
+      "grad_norm": 1.2861770391464233,
+      "learning_rate": 2e-05,
+      "loss": 0.05083389,
+      "step": 18719
+    },
+    {
+      "epoch": 37.44,
+      "grad_norm": 1.3172714710235596,
+      "learning_rate": 2e-05,
+      "loss": 0.0380227,
+      "step": 18720
+    },
+    {
+      "epoch": 37.442,
+      "grad_norm": 1.6785846948623657,
+      "learning_rate": 2e-05,
+      "loss": 0.05099306,
+      "step": 18721
+    },
+    {
+      "epoch": 37.444,
+      "grad_norm": 1.040389895439148,
+      "learning_rate": 2e-05,
+      "loss": 0.03511147,
+      "step": 18722
+    },
+    {
+      "epoch": 37.446,
+      "grad_norm": 1.0567665100097656,
+      "learning_rate": 2e-05,
+      "loss": 0.02976083,
+      "step": 18723
+    },
+    {
+      "epoch": 37.448,
+      "grad_norm": 0.9617815613746643,
+      "learning_rate": 2e-05,
+      "loss": 0.03670631,
+      "step": 18724
+    },
+    {
+      "epoch": 37.45,
+      "grad_norm": 1.069361925125122,
+      "learning_rate": 2e-05,
+      "loss": 0.03904826,
+      "step": 18725
+    },
+    {
+      "epoch": 37.452,
+      "grad_norm": 1.329677939414978,
+      "learning_rate": 2e-05,
+      "loss": 0.05781191,
+      "step": 18726
+    },
+    {
+      "epoch": 37.454,
+      "grad_norm": 1.146031379699707,
+      "learning_rate": 2e-05,
+      "loss": 0.05027099,
+      "step": 18727
+    },
+    {
+      "epoch": 37.456,
+      "grad_norm": 1.0236581563949585,
+      "learning_rate": 2e-05,
+      "loss": 0.0411206,
+      "step": 18728
+    },
+    {
+      "epoch": 37.458,
+      "grad_norm": 1.2816264629364014,
+      "learning_rate": 2e-05,
+      "loss": 0.04896802,
+      "step": 18729
+    },
+    {
+      "epoch": 37.46,
+      "grad_norm": 1.1948260068893433,
+      "learning_rate": 2e-05,
+      "loss": 0.03480673,
+      "step": 18730
+    },
+    {
+      "epoch": 37.462,
+      "grad_norm": 1.7572070360183716,
+      "learning_rate": 2e-05,
+      "loss": 0.04478199,
+      "step": 18731
+    },
+    {
+      "epoch": 37.464,
+      "grad_norm": 1.0101138353347778,
+      "learning_rate": 2e-05,
+      "loss": 0.044527,
+      "step": 18732
+    },
+    {
+      "epoch": 37.466,
+      "grad_norm": 1.537125825881958,
+      "learning_rate": 2e-05,
+      "loss": 0.04904353,
+      "step": 18733
+    },
+    {
+      "epoch": 37.468,
+      "grad_norm": 2.0057132244110107,
+      "learning_rate": 2e-05,
+      "loss": 0.05284391,
+      "step": 18734
+    },
+    {
+      "epoch": 37.47,
+      "grad_norm": 1.139591932296753,
+      "learning_rate": 2e-05,
+      "loss": 0.03955012,
+      "step": 18735
+    },
+    {
+      "epoch": 37.472,
+      "grad_norm": 1.3117825984954834,
+      "learning_rate": 2e-05,
+      "loss": 0.03998091,
+      "step": 18736
+    },
+    {
+      "epoch": 37.474,
+      "grad_norm": 1.599098563194275,
+      "learning_rate": 2e-05,
+      "loss": 0.0488559,
+      "step": 18737
+    },
+    {
+      "epoch": 37.476,
+      "grad_norm": 1.0199345350265503,
+      "learning_rate": 2e-05,
+      "loss": 0.02837726,
+      "step": 18738
+    },
+    {
+      "epoch": 37.478,
+      "grad_norm": 1.672568678855896,
+      "learning_rate": 2e-05,
+      "loss": 0.05150389,
+      "step": 18739
+    },
+    {
+      "epoch": 37.48,
+      "grad_norm": 0.9359270334243774,
+      "learning_rate": 2e-05,
+      "loss": 0.03139933,
+      "step": 18740
+    },
+    {
+      "epoch": 37.482,
+      "grad_norm": 1.5860284566879272,
+      "learning_rate": 2e-05,
+      "loss": 0.04607878,
+      "step": 18741
+    },
+    {
+      "epoch": 37.484,
+      "grad_norm": 1.0930761098861694,
+      "learning_rate": 2e-05,
+      "loss": 0.03963002,
+      "step": 18742
+    },
+    {
+      "epoch": 37.486,
+      "grad_norm": 1.213152527809143,
+      "learning_rate": 2e-05,
+      "loss": 0.04821583,
+      "step": 18743
+    },
+    {
+      "epoch": 37.488,
+      "grad_norm": 0.9678628444671631,
+      "learning_rate": 2e-05,
+      "loss": 0.03491917,
+      "step": 18744
+    },
+    {
+      "epoch": 37.49,
+      "grad_norm": 1.1579011678695679,
+      "learning_rate": 2e-05,
+      "loss": 0.05182432,
+      "step": 18745
+    },
+    {
+      "epoch": 37.492,
+      "grad_norm": 0.9998135566711426,
+      "learning_rate": 2e-05,
+      "loss": 0.02899381,
+      "step": 18746
+    },
+    {
+      "epoch": 37.494,
+      "grad_norm": 1.365634560585022,
+      "learning_rate": 2e-05,
+      "loss": 0.03647054,
+      "step": 18747
+    },
+    {
+      "epoch": 37.496,
+      "grad_norm": 2.8775148391723633,
+      "learning_rate": 2e-05,
+      "loss": 0.06468409,
+      "step": 18748
+    },
+    {
+      "epoch": 37.498,
+      "grad_norm": 1.3849204778671265,
+      "learning_rate": 2e-05,
+      "loss": 0.05843618,
+      "step": 18749
+    },
+    {
+      "epoch": 37.5,
+      "grad_norm": 1.4979232549667358,
+      "learning_rate": 2e-05,
+      "loss": 0.05652013,
+      "step": 18750
+    },
+    {
+      "epoch": 37.502,
+      "grad_norm": 1.2144519090652466,
+      "learning_rate": 2e-05,
+      "loss": 0.0389149,
+      "step": 18751
+    },
+    {
+      "epoch": 37.504,
+      "grad_norm": 1.0543650388717651,
+      "learning_rate": 2e-05,
+      "loss": 0.0313639,
+      "step": 18752
+    },
+    {
+      "epoch": 37.506,
+      "grad_norm": 1.038210391998291,
+      "learning_rate": 2e-05,
+      "loss": 0.04335963,
+      "step": 18753
+    },
+    {
+      "epoch": 37.508,
+      "grad_norm": 1.0079690217971802,
+      "learning_rate": 2e-05,
+      "loss": 0.03755116,
+      "step": 18754
+    },
+    {
+      "epoch": 37.51,
+      "grad_norm": 1.770876169204712,
+      "learning_rate": 2e-05,
+      "loss": 0.04107982,
+      "step": 18755
+    },
+    {
+      "epoch": 37.512,
+      "grad_norm": 1.6218074560165405,
+      "learning_rate": 2e-05,
+      "loss": 0.04625713,
+      "step": 18756
+    },
+    {
+      "epoch": 37.514,
+      "grad_norm": 1.1750694513320923,
+      "learning_rate": 2e-05,
+      "loss": 0.03950002,
+      "step": 18757
+    },
+    {
+      "epoch": 37.516,
+      "grad_norm": 1.2037353515625,
+      "learning_rate": 2e-05,
+      "loss": 0.04342839,
+      "step": 18758
+    },
+    {
+      "epoch": 37.518,
+      "grad_norm": 1.1445868015289307,
+      "learning_rate": 2e-05,
+      "loss": 0.04752634,
+      "step": 18759
+    },
+    {
+      "epoch": 37.52,
+      "grad_norm": 1.0486931800842285,
+      "learning_rate": 2e-05,
+      "loss": 0.04497886,
+      "step": 18760
+    },
+    {
+      "epoch": 37.522,
+      "grad_norm": 0.9221504330635071,
+      "learning_rate": 2e-05,
+      "loss": 0.0370011,
+      "step": 18761
+    },
+    {
+      "epoch": 37.524,
+      "grad_norm": 0.8999439477920532,
+      "learning_rate": 2e-05,
+      "loss": 0.03294341,
+      "step": 18762
+    },
+    {
+      "epoch": 37.526,
+      "grad_norm": 1.1722791194915771,
+      "learning_rate": 2e-05,
+      "loss": 0.03934911,
+      "step": 18763
+    },
+    {
+      "epoch": 37.528,
+      "grad_norm": 1.2961126565933228,
+      "learning_rate": 2e-05,
+      "loss": 0.04197449,
+      "step": 18764
+    },
+    {
+      "epoch": 37.53,
+      "grad_norm": 0.9199429154396057,
+      "learning_rate": 2e-05,
+      "loss": 0.02904882,
+      "step": 18765
+    },
+    {
+      "epoch": 37.532,
+      "grad_norm": 0.9045437574386597,
+      "learning_rate": 2e-05,
+      "loss": 0.03960499,
+      "step": 18766
+    },
+    {
+      "epoch": 37.534,
+      "grad_norm": 0.9374881982803345,
+      "learning_rate": 2e-05,
+      "loss": 0.03096394,
+      "step": 18767
+    },
+    {
+      "epoch": 37.536,
+      "grad_norm": 1.2470446825027466,
+      "learning_rate": 2e-05,
+      "loss": 0.04021191,
+      "step": 18768
+    },
+    {
+      "epoch": 37.538,
+      "grad_norm": 1.3754534721374512,
+      "learning_rate": 2e-05,
+      "loss": 0.0446906,
+      "step": 18769
+    },
+    {
+      "epoch": 37.54,
+      "grad_norm": 1.608644962310791,
+      "learning_rate": 2e-05,
+      "loss": 0.04486812,
+      "step": 18770
+    },
+    {
+      "epoch": 37.542,
+      "grad_norm": 1.4814962148666382,
+      "learning_rate": 2e-05,
+      "loss": 0.03845561,
+      "step": 18771
+    },
+    {
+      "epoch": 37.544,
+      "grad_norm": 0.8970831036567688,
+      "learning_rate": 2e-05,
+      "loss": 0.03752285,
+      "step": 18772
+    },
+    {
+      "epoch": 37.546,
+      "grad_norm": 1.1656574010849,
+      "learning_rate": 2e-05,
+      "loss": 0.03882036,
+      "step": 18773
+    },
+    {
+      "epoch": 37.548,
+      "grad_norm": 0.8688967823982239,
+      "learning_rate": 2e-05,
+      "loss": 0.01826111,
+      "step": 18774
+    },
+    {
+      "epoch": 37.55,
+      "grad_norm": 1.0525118112564087,
+      "learning_rate": 2e-05,
+      "loss": 0.04012558,
+      "step": 18775
+    },
+    {
+      "epoch": 37.552,
+      "grad_norm": 1.1830229759216309,
+      "learning_rate": 2e-05,
+      "loss": 0.05035875,
+      "step": 18776
+    },
+    {
+      "epoch": 37.554,
+      "grad_norm": 1.1931424140930176,
+      "learning_rate": 2e-05,
+      "loss": 0.04494492,
+      "step": 18777
+    },
+    {
+      "epoch": 37.556,
+      "grad_norm": 1.3106215000152588,
+      "learning_rate": 2e-05,
+      "loss": 0.04554956,
+      "step": 18778
+    },
+    {
+      "epoch": 37.558,
+      "grad_norm": 1.2222228050231934,
+      "learning_rate": 2e-05,
+      "loss": 0.04136209,
+      "step": 18779
+    },
+    {
+      "epoch": 37.56,
+      "grad_norm": 2.1237621307373047,
+      "learning_rate": 2e-05,
+      "loss": 0.04145623,
+      "step": 18780
+    },
+    {
+      "epoch": 37.562,
+      "grad_norm": 1.1739251613616943,
+      "learning_rate": 2e-05,
+      "loss": 0.04080192,
+      "step": 18781
+    },
+    {
+      "epoch": 37.564,
+      "grad_norm": 2.191302537918091,
+      "learning_rate": 2e-05,
+      "loss": 0.06097953,
+      "step": 18782
+    },
+    {
+      "epoch": 37.566,
+      "grad_norm": 0.9671980738639832,
+      "learning_rate": 2e-05,
+      "loss": 0.03338189,
+      "step": 18783
+    },
+    {
+      "epoch": 37.568,
+      "grad_norm": 1.8375569581985474,
+      "learning_rate": 2e-05,
+      "loss": 0.04796746,
+      "step": 18784
+    },
+    {
+      "epoch": 37.57,
+      "grad_norm": 1.0496113300323486,
+      "learning_rate": 2e-05,
+      "loss": 0.02443472,
+      "step": 18785
+    },
+    {
+      "epoch": 37.572,
+      "grad_norm": 2.503190755844116,
+      "learning_rate": 2e-05,
+      "loss": 0.05313979,
+      "step": 18786
+    },
+    {
+      "epoch": 37.574,
+      "grad_norm": 0.965300977230072,
+      "learning_rate": 2e-05,
+      "loss": 0.03321171,
+      "step": 18787
+    },
+    {
+      "epoch": 37.576,
+      "grad_norm": 1.3297483921051025,
+      "learning_rate": 2e-05,
+      "loss": 0.03972304,
+      "step": 18788
+    },
+    {
+      "epoch": 37.578,
+      "grad_norm": 1.1176469326019287,
+      "learning_rate": 2e-05,
+      "loss": 0.03872283,
+      "step": 18789
+    },
+    {
+      "epoch": 37.58,
+      "grad_norm": 1.2309982776641846,
+      "learning_rate": 2e-05,
+      "loss": 0.05587587,
+      "step": 18790
+    },
+    {
+      "epoch": 37.582,
+      "grad_norm": 1.7290147542953491,
+      "learning_rate": 2e-05,
+      "loss": 0.04843904,
+      "step": 18791
+    },
+    {
+      "epoch": 37.584,
+      "grad_norm": 1.3155032396316528,
+      "learning_rate": 2e-05,
+      "loss": 0.04053347,
+      "step": 18792
+    },
+    {
+      "epoch": 37.586,
+      "grad_norm": 1.361668586730957,
+      "learning_rate": 2e-05,
+      "loss": 0.03224905,
+      "step": 18793
+    },
+    {
+      "epoch": 37.588,
+      "grad_norm": 1.1360480785369873,
+      "learning_rate": 2e-05,
+      "loss": 0.03972386,
+      "step": 18794
+    },
+    {
+      "epoch": 37.59,
+      "grad_norm": 1.3317919969558716,
+      "learning_rate": 2e-05,
+      "loss": 0.05043165,
+      "step": 18795
+    },
+    {
+      "epoch": 37.592,
+      "grad_norm": 1.1488837003707886,
+      "learning_rate": 2e-05,
+      "loss": 0.03248937,
+      "step": 18796
+    },
+    {
+      "epoch": 37.594,
+      "grad_norm": 1.4023540019989014,
+      "learning_rate": 2e-05,
+      "loss": 0.03502064,
+      "step": 18797
+    },
+    {
+      "epoch": 37.596,
+      "grad_norm": 1.0766652822494507,
+      "learning_rate": 2e-05,
+      "loss": 0.03397478,
+      "step": 18798
+    },
+    {
+      "epoch": 37.598,
+      "grad_norm": 1.0824081897735596,
+      "learning_rate": 2e-05,
+      "loss": 0.04138397,
+      "step": 18799
+    },
+    {
+      "epoch": 37.6,
+      "grad_norm": 1.0459606647491455,
+      "learning_rate": 2e-05,
+      "loss": 0.04331833,
+      "step": 18800
+    },
+    {
+      "epoch": 37.602,
+      "grad_norm": 1.3006701469421387,
+      "learning_rate": 2e-05,
+      "loss": 0.06298004,
+      "step": 18801
+    },
+    {
+      "epoch": 37.604,
+      "grad_norm": 0.9166358113288879,
+      "learning_rate": 2e-05,
+      "loss": 0.02415272,
+      "step": 18802
+    },
+    {
+      "epoch": 37.606,
+      "grad_norm": 1.091423511505127,
+      "learning_rate": 2e-05,
+      "loss": 0.03883186,
+      "step": 18803
+    },
+    {
+      "epoch": 37.608,
+      "grad_norm": 1.242665410041809,
+      "learning_rate": 2e-05,
+      "loss": 0.04210323,
+      "step": 18804
+    },
+    {
+      "epoch": 37.61,
+      "grad_norm": 1.4640295505523682,
+      "learning_rate": 2e-05,
+      "loss": 0.05143019,
+      "step": 18805
+    },
+    {
+      "epoch": 37.612,
+      "grad_norm": 1.2153027057647705,
+      "learning_rate": 2e-05,
+      "loss": 0.05529271,
+      "step": 18806
+    },
+    {
+      "epoch": 37.614,
+      "grad_norm": 2.2144391536712646,
+      "learning_rate": 2e-05,
+      "loss": 0.04179824,
+      "step": 18807
+    },
+    {
+      "epoch": 37.616,
+      "grad_norm": 1.0586007833480835,
+      "learning_rate": 2e-05,
+      "loss": 0.0338909,
+      "step": 18808
+    },
+    {
+      "epoch": 37.618,
+      "grad_norm": 1.1268335580825806,
+      "learning_rate": 2e-05,
+      "loss": 0.03529718,
+      "step": 18809
+    },
+    {
+      "epoch": 37.62,
+      "grad_norm": 1.4889854192733765,
+      "learning_rate": 2e-05,
+      "loss": 0.0585743,
+      "step": 18810
+    },
+    {
+      "epoch": 37.622,
+      "grad_norm": 1.6791647672653198,
+      "learning_rate": 2e-05,
+      "loss": 0.05393176,
+      "step": 18811
+    },
+    {
+      "epoch": 37.624,
+      "grad_norm": 1.1075037717819214,
+      "learning_rate": 2e-05,
+      "loss": 0.02849288,
+      "step": 18812
+    },
+    {
+      "epoch": 37.626,
+      "grad_norm": 1.3440752029418945,
+      "learning_rate": 2e-05,
+      "loss": 0.04735216,
+      "step": 18813
+    },
+    {
+      "epoch": 37.628,
+      "grad_norm": 1.2121247053146362,
+      "learning_rate": 2e-05,
+      "loss": 0.03134248,
+      "step": 18814
+    },
+    {
+      "epoch": 37.63,
+      "grad_norm": 1.0615363121032715,
+      "learning_rate": 2e-05,
+      "loss": 0.03442275,
+      "step": 18815
+    },
+    {
+      "epoch": 37.632,
+      "grad_norm": 1.183545708656311,
+      "learning_rate": 2e-05,
+      "loss": 0.03669387,
+      "step": 18816
+    },
+    {
+      "epoch": 37.634,
+      "grad_norm": 0.9101192951202393,
+      "learning_rate": 2e-05,
+      "loss": 0.03575148,
+      "step": 18817
+    },
+    {
+      "epoch": 37.636,
+      "grad_norm": 1.5718504190444946,
+      "learning_rate": 2e-05,
+      "loss": 0.05680436,
+      "step": 18818
+    },
+    {
+      "epoch": 37.638,
+      "grad_norm": 1.2223116159439087,
+      "learning_rate": 2e-05,
+      "loss": 0.04046562,
+      "step": 18819
+    },
+    {
+      "epoch": 37.64,
+      "grad_norm": 1.1684964895248413,
+      "learning_rate": 2e-05,
+      "loss": 0.0525642,
+      "step": 18820
+    },
+    {
+      "epoch": 37.642,
+      "grad_norm": 1.0936496257781982,
+      "learning_rate": 2e-05,
+      "loss": 0.04073422,
+      "step": 18821
+    },
+    {
+      "epoch": 37.644,
+      "grad_norm": 1.0097864866256714,
+      "learning_rate": 2e-05,
+      "loss": 0.03097866,
+      "step": 18822
+    },
+    {
+      "epoch": 37.646,
+      "grad_norm": 1.2851216793060303,
+      "learning_rate": 2e-05,
+      "loss": 0.05580614,
+      "step": 18823
+    },
+    {
+      "epoch": 37.648,
+      "grad_norm": 1.9229702949523926,
+      "learning_rate": 2e-05,
+      "loss": 0.03958568,
+      "step": 18824
+    },
+    {
+      "epoch": 37.65,
+      "grad_norm": 1.3668385744094849,
+      "learning_rate": 2e-05,
+      "loss": 0.05524768,
+      "step": 18825
+    },
+    {
+      "epoch": 37.652,
+      "grad_norm": 1.417112946510315,
+      "learning_rate": 2e-05,
+      "loss": 0.04026988,
+      "step": 18826
+    },
+    {
+      "epoch": 37.654,
+      "grad_norm": 1.154919147491455,
+      "learning_rate": 2e-05,
+      "loss": 0.04654554,
+      "step": 18827
+    },
+    {
+      "epoch": 37.656,
+      "grad_norm": 0.9886472821235657,
+      "learning_rate": 2e-05,
+      "loss": 0.03133095,
+      "step": 18828
+    },
+    {
+      "epoch": 37.658,
+      "grad_norm": 1.2836391925811768,
+      "learning_rate": 2e-05,
+      "loss": 0.03864687,
+      "step": 18829
+    },
+    {
+      "epoch": 37.66,
+      "grad_norm": 1.2053261995315552,
+      "learning_rate": 2e-05,
+      "loss": 0.05016092,
+      "step": 18830
+    },
+    {
+      "epoch": 37.662,
+      "grad_norm": 1.1073130369186401,
+      "learning_rate": 2e-05,
+      "loss": 0.03775865,
+      "step": 18831
+    },
+    {
+      "epoch": 37.664,
+      "grad_norm": 1.2172409296035767,
+      "learning_rate": 2e-05,
+      "loss": 0.05411567,
+      "step": 18832
+    },
+    {
+      "epoch": 37.666,
+      "grad_norm": 1.1823009252548218,
+      "learning_rate": 2e-05,
+      "loss": 0.0415944,
+      "step": 18833
+    },
+    {
+      "epoch": 37.668,
+      "grad_norm": 1.0593645572662354,
+      "learning_rate": 2e-05,
+      "loss": 0.03887521,
+      "step": 18834
+    },
+    {
+      "epoch": 37.67,
+      "grad_norm": 1.0737974643707275,
+      "learning_rate": 2e-05,
+      "loss": 0.03494824,
+      "step": 18835
+    },
+    {
+      "epoch": 37.672,
+      "grad_norm": 1.4756815433502197,
+      "learning_rate": 2e-05,
+      "loss": 0.05688382,
+      "step": 18836
+    },
+    {
+      "epoch": 37.674,
+      "grad_norm": 1.2232595682144165,
+      "learning_rate": 2e-05,
+      "loss": 0.05065658,
+      "step": 18837
+    },
+    {
+      "epoch": 37.676,
+      "grad_norm": 1.5146193504333496,
+      "learning_rate": 2e-05,
+      "loss": 0.04606935,
+      "step": 18838
+    },
+    {
+      "epoch": 37.678,
+      "grad_norm": 1.1179404258728027,
+      "learning_rate": 2e-05,
+      "loss": 0.04785991,
+      "step": 18839
+    },
+    {
+      "epoch": 37.68,
+      "grad_norm": 1.7704997062683105,
+      "learning_rate": 2e-05,
+      "loss": 0.03499816,
+      "step": 18840
+    },
+    {
+      "epoch": 37.682,
+      "grad_norm": 2.1645781993865967,
+      "learning_rate": 2e-05,
+      "loss": 0.06860285,
+      "step": 18841
+    },
+    {
+      "epoch": 37.684,
+      "grad_norm": 1.4239698648452759,
+      "learning_rate": 2e-05,
+      "loss": 0.05423044,
+      "step": 18842
+    },
+    {
+      "epoch": 37.686,
+      "grad_norm": 1.2061432600021362,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 18843
+    },
+    {
+      "epoch": 37.688,
+      "grad_norm": 1.434126615524292,
+      "learning_rate": 2e-05,
+      "loss": 0.05467582,
+      "step": 18844
+    },
+    {
+      "epoch": 37.69,
+      "grad_norm": 1.0668646097183228,
+      "learning_rate": 2e-05,
+      "loss": 0.04183818,
+      "step": 18845
+    },
+    {
+      "epoch": 37.692,
+      "grad_norm": 0.9105406999588013,
+      "learning_rate": 2e-05,
+      "loss": 0.0302218,
+      "step": 18846
+    },
+    {
+      "epoch": 37.694,
+      "grad_norm": 2.7572875022888184,
+      "learning_rate": 2e-05,
+      "loss": 0.04500583,
+      "step": 18847
+    },
+    {
+      "epoch": 37.696,
+      "grad_norm": 2.284607172012329,
+      "learning_rate": 2e-05,
+      "loss": 0.04284734,
+      "step": 18848
+    },
+    {
+      "epoch": 37.698,
+      "grad_norm": 2.073329210281372,
+      "learning_rate": 2e-05,
+      "loss": 0.03520484,
+      "step": 18849
+    },
+    {
+      "epoch": 37.7,
+      "grad_norm": 3.604762077331543,
+      "learning_rate": 2e-05,
+      "loss": 0.06385957,
+      "step": 18850
+    },
+    {
+      "epoch": 37.702,
+      "grad_norm": 1.1945736408233643,
+      "learning_rate": 2e-05,
+      "loss": 0.03985525,
+      "step": 18851
+    },
+    {
+      "epoch": 37.704,
+      "grad_norm": 1.2185908555984497,
+      "learning_rate": 2e-05,
+      "loss": 0.03318772,
+      "step": 18852
+    },
+    {
+      "epoch": 37.706,
+      "grad_norm": 1.2688206434249878,
+      "learning_rate": 2e-05,
+      "loss": 0.05495746,
+      "step": 18853
+    },
+    {
+      "epoch": 37.708,
+      "grad_norm": 1.0381929874420166,
+      "learning_rate": 2e-05,
+      "loss": 0.03600042,
+      "step": 18854
+    },
+    {
+      "epoch": 37.71,
+      "grad_norm": 0.952168345451355,
+      "learning_rate": 2e-05,
+      "loss": 0.02752374,
+      "step": 18855
+    },
+    {
+      "epoch": 37.712,
+      "grad_norm": 1.2322078943252563,
+      "learning_rate": 2e-05,
+      "loss": 0.04587801,
+      "step": 18856
+    },
+    {
+      "epoch": 37.714,
+      "grad_norm": 1.2536684274673462,
+      "learning_rate": 2e-05,
+      "loss": 0.03789248,
+      "step": 18857
+    },
+    {
+      "epoch": 37.716,
+      "grad_norm": 2.647052526473999,
+      "learning_rate": 2e-05,
+      "loss": 0.07668023,
+      "step": 18858
+    },
+    {
+      "epoch": 37.718,
+      "grad_norm": 2.344609022140503,
+      "learning_rate": 2e-05,
+      "loss": 0.06175537,
+      "step": 18859
+    },
+    {
+      "epoch": 37.72,
+      "grad_norm": 1.710271954536438,
+      "learning_rate": 2e-05,
+      "loss": 0.07425263,
+      "step": 18860
+    },
+    {
+      "epoch": 37.722,
+      "grad_norm": 0.9385952353477478,
+      "learning_rate": 2e-05,
+      "loss": 0.03988194,
+      "step": 18861
+    },
+    {
+      "epoch": 37.724,
+      "grad_norm": 1.0760703086853027,
+      "learning_rate": 2e-05,
+      "loss": 0.03247745,
+      "step": 18862
+    },
+    {
+      "epoch": 37.726,
+      "grad_norm": 1.346384882926941,
+      "learning_rate": 2e-05,
+      "loss": 0.0547515,
+      "step": 18863
+    },
+    {
+      "epoch": 37.728,
+      "grad_norm": 1.0319280624389648,
+      "learning_rate": 2e-05,
+      "loss": 0.03719486,
+      "step": 18864
+    },
+    {
+      "epoch": 37.73,
+      "grad_norm": 1.327825665473938,
+      "learning_rate": 2e-05,
+      "loss": 0.02865359,
+      "step": 18865
+    },
+    {
+      "epoch": 37.732,
+      "grad_norm": 1.7671387195587158,
+      "learning_rate": 2e-05,
+      "loss": 0.03048402,
+      "step": 18866
+    },
+    {
+      "epoch": 37.734,
+      "grad_norm": 1.0499013662338257,
+      "learning_rate": 2e-05,
+      "loss": 0.03205416,
+      "step": 18867
+    },
+    {
+      "epoch": 37.736,
+      "grad_norm": 0.9944764971733093,
+      "learning_rate": 2e-05,
+      "loss": 0.03722709,
+      "step": 18868
+    },
+    {
+      "epoch": 37.738,
+      "grad_norm": 1.3089152574539185,
+      "learning_rate": 2e-05,
+      "loss": 0.04756982,
+      "step": 18869
+    },
+    {
+      "epoch": 37.74,
+      "grad_norm": 1.5196123123168945,
+      "learning_rate": 2e-05,
+      "loss": 0.06203455,
+      "step": 18870
+    },
+    {
+      "epoch": 37.742,
+      "grad_norm": 1.11556077003479,
+      "learning_rate": 2e-05,
+      "loss": 0.03433929,
+      "step": 18871
+    },
+    {
+      "epoch": 37.744,
+      "grad_norm": 2.001180410385132,
+      "learning_rate": 2e-05,
+      "loss": 0.046538,
+      "step": 18872
+    },
+    {
+      "epoch": 37.746,
+      "grad_norm": 1.2591460943222046,
+      "learning_rate": 2e-05,
+      "loss": 0.05003956,
+      "step": 18873
+    },
+    {
+      "epoch": 37.748,
+      "grad_norm": 1.6820541620254517,
+      "learning_rate": 2e-05,
+      "loss": 0.043096,
+      "step": 18874
+    },
+    {
+      "epoch": 37.75,
+      "grad_norm": 1.2268869876861572,
+      "learning_rate": 2e-05,
+      "loss": 0.04136671,
+      "step": 18875
+    },
+    {
+      "epoch": 37.752,
+      "grad_norm": 1.2533190250396729,
+      "learning_rate": 2e-05,
+      "loss": 0.0366976,
+      "step": 18876
+    },
+    {
+      "epoch": 37.754,
+      "grad_norm": 1.4866986274719238,
+      "learning_rate": 2e-05,
+      "loss": 0.03636875,
+      "step": 18877
+    },
+    {
+      "epoch": 37.756,
+      "grad_norm": 1.5848642587661743,
+      "learning_rate": 2e-05,
+      "loss": 0.04737707,
+      "step": 18878
+    },
+    {
+      "epoch": 37.758,
+      "grad_norm": 1.353268027305603,
+      "learning_rate": 2e-05,
+      "loss": 0.03133782,
+      "step": 18879
+    },
+    {
+      "epoch": 37.76,
+      "grad_norm": 0.9636505842208862,
+      "learning_rate": 2e-05,
+      "loss": 0.0301811,
+      "step": 18880
+    },
+    {
+      "epoch": 37.762,
+      "grad_norm": 1.2497457265853882,
+      "learning_rate": 2e-05,
+      "loss": 0.05001786,
+      "step": 18881
+    },
+    {
+      "epoch": 37.764,
+      "grad_norm": 1.0673328638076782,
+      "learning_rate": 2e-05,
+      "loss": 0.03139454,
+      "step": 18882
+    },
+    {
+      "epoch": 37.766,
+      "grad_norm": 1.557485580444336,
+      "learning_rate": 2e-05,
+      "loss": 0.0464134,
+      "step": 18883
+    },
+    {
+      "epoch": 37.768,
+      "grad_norm": 0.973440408706665,
+      "learning_rate": 2e-05,
+      "loss": 0.03469043,
+      "step": 18884
+    },
+    {
+      "epoch": 37.77,
+      "grad_norm": 1.1748510599136353,
+      "learning_rate": 2e-05,
+      "loss": 0.03730968,
+      "step": 18885
+    },
+    {
+      "epoch": 37.772,
+      "grad_norm": 1.0944515466690063,
+      "learning_rate": 2e-05,
+      "loss": 0.03888043,
+      "step": 18886
+    },
+    {
+      "epoch": 37.774,
+      "grad_norm": 3.6084587574005127,
+      "learning_rate": 2e-05,
+      "loss": 0.05079032,
+      "step": 18887
+    },
+    {
+      "epoch": 37.776,
+      "grad_norm": 1.7033201456069946,
+      "learning_rate": 2e-05,
+      "loss": 0.0442396,
+      "step": 18888
+    },
+    {
+      "epoch": 37.778,
+      "grad_norm": 1.3825198411941528,
+      "learning_rate": 2e-05,
+      "loss": 0.05911713,
+      "step": 18889
+    },
+    {
+      "epoch": 37.78,
+      "grad_norm": 1.3935472965240479,
+      "learning_rate": 2e-05,
+      "loss": 0.04597136,
+      "step": 18890
+    },
+    {
+      "epoch": 37.782,
+      "grad_norm": 1.515001893043518,
+      "learning_rate": 2e-05,
+      "loss": 0.04437057,
+      "step": 18891
+    },
+    {
+      "epoch": 37.784,
+      "grad_norm": 1.1268634796142578,
+      "learning_rate": 2e-05,
+      "loss": 0.03411634,
+      "step": 18892
+    },
+    {
+      "epoch": 37.786,
+      "grad_norm": 0.9976750016212463,
+      "learning_rate": 2e-05,
+      "loss": 0.03444566,
+      "step": 18893
+    },
+    {
+      "epoch": 37.788,
+      "grad_norm": 1.5224945545196533,
+      "learning_rate": 2e-05,
+      "loss": 0.0607723,
+      "step": 18894
+    },
+    {
+      "epoch": 37.79,
+      "grad_norm": 1.3955048322677612,
+      "learning_rate": 2e-05,
+      "loss": 0.05213226,
+      "step": 18895
+    },
+    {
+      "epoch": 37.792,
+      "grad_norm": 1.1549447774887085,
+      "learning_rate": 2e-05,
+      "loss": 0.03736443,
+      "step": 18896
+    },
+    {
+      "epoch": 37.794,
+      "grad_norm": 0.8485814929008484,
+      "learning_rate": 2e-05,
+      "loss": 0.02462177,
+      "step": 18897
+    },
+    {
+      "epoch": 37.796,
+      "grad_norm": 1.3781180381774902,
+      "learning_rate": 2e-05,
+      "loss": 0.05236533,
+      "step": 18898
+    },
+    {
+      "epoch": 37.798,
+      "grad_norm": 1.4734482765197754,
+      "learning_rate": 2e-05,
+      "loss": 0.05728751,
+      "step": 18899
+    },
+    {
+      "epoch": 37.8,
+      "grad_norm": 1.1258692741394043,
+      "learning_rate": 2e-05,
+      "loss": 0.03739987,
+      "step": 18900
+    },
+    {
+      "epoch": 37.802,
+      "grad_norm": 1.2827837467193604,
+      "learning_rate": 2e-05,
+      "loss": 0.05210217,
+      "step": 18901
+    },
+    {
+      "epoch": 37.804,
+      "grad_norm": 1.1557444334030151,
+      "learning_rate": 2e-05,
+      "loss": 0.0390267,
+      "step": 18902
+    },
+    {
+      "epoch": 37.806,
+      "grad_norm": 1.1604247093200684,
+      "learning_rate": 2e-05,
+      "loss": 0.04971418,
+      "step": 18903
+    },
+    {
+      "epoch": 37.808,
+      "grad_norm": 1.0621238946914673,
+      "learning_rate": 2e-05,
+      "loss": 0.03640627,
+      "step": 18904
+    },
+    {
+      "epoch": 37.81,
+      "grad_norm": 0.9951264262199402,
+      "learning_rate": 2e-05,
+      "loss": 0.03904339,
+      "step": 18905
+    },
+    {
+      "epoch": 37.812,
+      "grad_norm": 1.199965476989746,
+      "learning_rate": 2e-05,
+      "loss": 0.04364878,
+      "step": 18906
+    },
+    {
+      "epoch": 37.814,
+      "grad_norm": 3.0767629146575928,
+      "learning_rate": 2e-05,
+      "loss": 0.04365045,
+      "step": 18907
+    },
+    {
+      "epoch": 37.816,
+      "grad_norm": 0.8647134304046631,
+      "learning_rate": 2e-05,
+      "loss": 0.0334196,
+      "step": 18908
+    },
+    {
+      "epoch": 37.818,
+      "grad_norm": 1.1921844482421875,
+      "learning_rate": 2e-05,
+      "loss": 0.03833631,
+      "step": 18909
+    },
+    {
+      "epoch": 37.82,
+      "grad_norm": 1.0273867845535278,
+      "learning_rate": 2e-05,
+      "loss": 0.04143481,
+      "step": 18910
+    },
+    {
+      "epoch": 37.822,
+      "grad_norm": 1.125104308128357,
+      "learning_rate": 2e-05,
+      "loss": 0.05544262,
+      "step": 18911
+    },
+    {
+      "epoch": 37.824,
+      "grad_norm": 1.0411149263381958,
+      "learning_rate": 2e-05,
+      "loss": 0.04289424,
+      "step": 18912
+    },
+    {
+      "epoch": 37.826,
+      "grad_norm": 1.2323323488235474,
+      "learning_rate": 2e-05,
+      "loss": 0.05162044,
+      "step": 18913
+    },
+    {
+      "epoch": 37.828,
+      "grad_norm": 1.2034226655960083,
+      "learning_rate": 2e-05,
+      "loss": 0.05214075,
+      "step": 18914
+    },
+    {
+      "epoch": 37.83,
+      "grad_norm": 1.3405710458755493,
+      "learning_rate": 2e-05,
+      "loss": 0.05168641,
+      "step": 18915
+    },
+    {
+      "epoch": 37.832,
+      "grad_norm": 1.2981528043746948,
+      "learning_rate": 2e-05,
+      "loss": 0.04356202,
+      "step": 18916
+    },
+    {
+      "epoch": 37.834,
+      "grad_norm": 0.9416662454605103,
+      "learning_rate": 2e-05,
+      "loss": 0.03926639,
+      "step": 18917
+    },
+    {
+      "epoch": 37.836,
+      "grad_norm": 1.080146312713623,
+      "learning_rate": 2e-05,
+      "loss": 0.03623664,
+      "step": 18918
+    },
+    {
+      "epoch": 37.838,
+      "grad_norm": 1.164093255996704,
+      "learning_rate": 2e-05,
+      "loss": 0.04676047,
+      "step": 18919
+    },
+    {
+      "epoch": 37.84,
+      "grad_norm": 1.0014550685882568,
+      "learning_rate": 2e-05,
+      "loss": 0.03417875,
+      "step": 18920
+    },
+    {
+      "epoch": 37.842,
+      "grad_norm": 1.2077052593231201,
+      "learning_rate": 2e-05,
+      "loss": 0.03918765,
+      "step": 18921
+    },
+    {
+      "epoch": 37.844,
+      "grad_norm": 1.121488094329834,
+      "learning_rate": 2e-05,
+      "loss": 0.04199528,
+      "step": 18922
+    },
+    {
+      "epoch": 37.846,
+      "grad_norm": 2.043963670730591,
+      "learning_rate": 2e-05,
+      "loss": 0.06683844,
+      "step": 18923
+    },
+    {
+      "epoch": 37.848,
+      "grad_norm": 1.0892078876495361,
+      "learning_rate": 2e-05,
+      "loss": 0.04384585,
+      "step": 18924
+    },
+    {
+      "epoch": 37.85,
+      "grad_norm": 1.333020567893982,
+      "learning_rate": 2e-05,
+      "loss": 0.04862942,
+      "step": 18925
+    },
+    {
+      "epoch": 37.852,
+      "grad_norm": 1.0511001348495483,
+      "learning_rate": 2e-05,
+      "loss": 0.03976622,
+      "step": 18926
+    },
+    {
+      "epoch": 37.854,
+      "grad_norm": 1.2701404094696045,
+      "learning_rate": 2e-05,
+      "loss": 0.04149718,
+      "step": 18927
+    },
+    {
+      "epoch": 37.856,
+      "grad_norm": 1.0651763677597046,
+      "learning_rate": 2e-05,
+      "loss": 0.03586752,
+      "step": 18928
+    },
+    {
+      "epoch": 37.858,
+      "grad_norm": 1.0089900493621826,
+      "learning_rate": 2e-05,
+      "loss": 0.03525446,
+      "step": 18929
+    },
+    {
+      "epoch": 37.86,
+      "grad_norm": 1.1876775026321411,
+      "learning_rate": 2e-05,
+      "loss": 0.04068927,
+      "step": 18930
+    },
+    {
+      "epoch": 37.862,
+      "grad_norm": 1.7308181524276733,
+      "learning_rate": 2e-05,
+      "loss": 0.04104917,
+      "step": 18931
+    },
+    {
+      "epoch": 37.864,
+      "grad_norm": 1.0084033012390137,
+      "learning_rate": 2e-05,
+      "loss": 0.03396938,
+      "step": 18932
+    },
+    {
+      "epoch": 37.866,
+      "grad_norm": 1.4454501867294312,
+      "learning_rate": 2e-05,
+      "loss": 0.05074462,
+      "step": 18933
+    },
+    {
+      "epoch": 37.868,
+      "grad_norm": 1.103598952293396,
+      "learning_rate": 2e-05,
+      "loss": 0.04150467,
+      "step": 18934
+    },
+    {
+      "epoch": 37.87,
+      "grad_norm": 1.0260006189346313,
+      "learning_rate": 2e-05,
+      "loss": 0.03909385,
+      "step": 18935
+    },
+    {
+      "epoch": 37.872,
+      "grad_norm": 1.1484661102294922,
+      "learning_rate": 2e-05,
+      "loss": 0.0446529,
+      "step": 18936
+    },
+    {
+      "epoch": 37.874,
+      "grad_norm": 0.9371235966682434,
+      "learning_rate": 2e-05,
+      "loss": 0.03469496,
+      "step": 18937
+    },
+    {
+      "epoch": 37.876,
+      "grad_norm": 2.441399097442627,
+      "learning_rate": 2e-05,
+      "loss": 0.04239362,
+      "step": 18938
+    },
+    {
+      "epoch": 37.878,
+      "grad_norm": 0.9973140358924866,
+      "learning_rate": 2e-05,
+      "loss": 0.03371353,
+      "step": 18939
+    },
+    {
+      "epoch": 37.88,
+      "grad_norm": 1.1159844398498535,
+      "learning_rate": 2e-05,
+      "loss": 0.04418877,
+      "step": 18940
+    },
+    {
+      "epoch": 37.882,
+      "grad_norm": 1.0986323356628418,
+      "learning_rate": 2e-05,
+      "loss": 0.0396911,
+      "step": 18941
+    },
+    {
+      "epoch": 37.884,
+      "grad_norm": 2.239150285720825,
+      "learning_rate": 2e-05,
+      "loss": 0.05810123,
+      "step": 18942
+    },
+    {
+      "epoch": 37.886,
+      "grad_norm": 1.1507476568222046,
+      "learning_rate": 2e-05,
+      "loss": 0.04050256,
+      "step": 18943
+    },
+    {
+      "epoch": 37.888,
+      "grad_norm": 1.643411636352539,
+      "learning_rate": 2e-05,
+      "loss": 0.07198939,
+      "step": 18944
+    },
+    {
+      "epoch": 37.89,
+      "grad_norm": 1.7802282571792603,
+      "learning_rate": 2e-05,
+      "loss": 0.0361943,
+      "step": 18945
+    },
+    {
+      "epoch": 37.892,
+      "grad_norm": 1.0710393190383911,
+      "learning_rate": 2e-05,
+      "loss": 0.0276309,
+      "step": 18946
+    },
+    {
+      "epoch": 37.894,
+      "grad_norm": 1.3028510808944702,
+      "learning_rate": 2e-05,
+      "loss": 0.04614664,
+      "step": 18947
+    },
+    {
+      "epoch": 37.896,
+      "grad_norm": 1.6930269002914429,
+      "learning_rate": 2e-05,
+      "loss": 0.05176745,
+      "step": 18948
+    },
+    {
+      "epoch": 37.898,
+      "grad_norm": 1.3994868993759155,
+      "learning_rate": 2e-05,
+      "loss": 0.04597143,
+      "step": 18949
+    },
+    {
+      "epoch": 37.9,
+      "grad_norm": 1.300676703453064,
+      "learning_rate": 2e-05,
+      "loss": 0.0538766,
+      "step": 18950
+    },
+    {
+      "epoch": 37.902,
+      "grad_norm": 1.897986888885498,
+      "learning_rate": 2e-05,
+      "loss": 0.03771758,
+      "step": 18951
+    },
+    {
+      "epoch": 37.904,
+      "grad_norm": 1.1213583946228027,
+      "learning_rate": 2e-05,
+      "loss": 0.04674072,
+      "step": 18952
+    },
+    {
+      "epoch": 37.906,
+      "grad_norm": 1.3618868589401245,
+      "learning_rate": 2e-05,
+      "loss": 0.06035518,
+      "step": 18953
+    },
+    {
+      "epoch": 37.908,
+      "grad_norm": 1.2731349468231201,
+      "learning_rate": 2e-05,
+      "loss": 0.05713828,
+      "step": 18954
+    },
+    {
+      "epoch": 37.91,
+      "grad_norm": 1.0878880023956299,
+      "learning_rate": 2e-05,
+      "loss": 0.05491067,
+      "step": 18955
+    },
+    {
+      "epoch": 37.912,
+      "grad_norm": 1.5778956413269043,
+      "learning_rate": 2e-05,
+      "loss": 0.05252867,
+      "step": 18956
+    },
+    {
+      "epoch": 37.914,
+      "grad_norm": 0.9078095555305481,
+      "learning_rate": 2e-05,
+      "loss": 0.03131845,
+      "step": 18957
+    },
+    {
+      "epoch": 37.916,
+      "grad_norm": 1.0328174829483032,
+      "learning_rate": 2e-05,
+      "loss": 0.03666805,
+      "step": 18958
+    },
+    {
+      "epoch": 37.918,
+      "grad_norm": 1.363809585571289,
+      "learning_rate": 2e-05,
+      "loss": 0.05573125,
+      "step": 18959
+    },
+    {
+      "epoch": 37.92,
+      "grad_norm": 1.0942089557647705,
+      "learning_rate": 2e-05,
+      "loss": 0.05161095,
+      "step": 18960
+    },
+    {
+      "epoch": 37.922,
+      "grad_norm": 1.2652561664581299,
+      "learning_rate": 2e-05,
+      "loss": 0.05439752,
+      "step": 18961
+    },
+    {
+      "epoch": 37.924,
+      "grad_norm": 1.4360697269439697,
+      "learning_rate": 2e-05,
+      "loss": 0.04909812,
+      "step": 18962
+    },
+    {
+      "epoch": 37.926,
+      "grad_norm": 1.5713309049606323,
+      "learning_rate": 2e-05,
+      "loss": 0.05065858,
+      "step": 18963
+    },
+    {
+      "epoch": 37.928,
+      "grad_norm": 1.3560004234313965,
+      "learning_rate": 2e-05,
+      "loss": 0.05886821,
+      "step": 18964
+    },
+    {
+      "epoch": 37.93,
+      "grad_norm": 1.1961268186569214,
+      "learning_rate": 2e-05,
+      "loss": 0.04319284,
+      "step": 18965
+    },
+    {
+      "epoch": 37.932,
+      "grad_norm": 1.2720478773117065,
+      "learning_rate": 2e-05,
+      "loss": 0.07183904,
+      "step": 18966
+    },
+    {
+      "epoch": 37.934,
+      "grad_norm": 1.26603102684021,
+      "learning_rate": 2e-05,
+      "loss": 0.04299184,
+      "step": 18967
+    },
+    {
+      "epoch": 37.936,
+      "grad_norm": 1.0896682739257812,
+      "learning_rate": 2e-05,
+      "loss": 0.04101091,
+      "step": 18968
+    },
+    {
+      "epoch": 37.938,
+      "grad_norm": 1.333804726600647,
+      "learning_rate": 2e-05,
+      "loss": 0.03814703,
+      "step": 18969
+    },
+    {
+      "epoch": 37.94,
+      "grad_norm": 1.6315276622772217,
+      "learning_rate": 2e-05,
+      "loss": 0.02696334,
+      "step": 18970
+    },
+    {
+      "epoch": 37.942,
+      "grad_norm": 1.2752485275268555,
+      "learning_rate": 2e-05,
+      "loss": 0.04623385,
+      "step": 18971
+    },
+    {
+      "epoch": 37.944,
+      "grad_norm": 1.2738232612609863,
+      "learning_rate": 2e-05,
+      "loss": 0.05119524,
+      "step": 18972
+    },
+    {
+      "epoch": 37.946,
+      "grad_norm": 1.229262351989746,
+      "learning_rate": 2e-05,
+      "loss": 0.03859371,
+      "step": 18973
+    },
+    {
+      "epoch": 37.948,
+      "grad_norm": 1.292868971824646,
+      "learning_rate": 2e-05,
+      "loss": 0.03332417,
+      "step": 18974
+    },
+    {
+      "epoch": 37.95,
+      "grad_norm": 0.9618489742279053,
+      "learning_rate": 2e-05,
+      "loss": 0.03035795,
+      "step": 18975
+    },
+    {
+      "epoch": 37.952,
+      "grad_norm": 0.9348632097244263,
+      "learning_rate": 2e-05,
+      "loss": 0.03629586,
+      "step": 18976
+    },
+    {
+      "epoch": 37.954,
+      "grad_norm": 1.0433769226074219,
+      "learning_rate": 2e-05,
+      "loss": 0.04461107,
+      "step": 18977
+    },
+    {
+      "epoch": 37.956,
+      "grad_norm": 1.8684062957763672,
+      "learning_rate": 2e-05,
+      "loss": 0.04248746,
+      "step": 18978
+    },
+    {
+      "epoch": 37.958,
+      "grad_norm": 1.2349817752838135,
+      "learning_rate": 2e-05,
+      "loss": 0.04595061,
+      "step": 18979
+    },
+    {
+      "epoch": 37.96,
+      "grad_norm": 1.1329549551010132,
+      "learning_rate": 2e-05,
+      "loss": 0.05171024,
+      "step": 18980
+    },
+    {
+      "epoch": 37.962,
+      "grad_norm": 1.4046157598495483,
+      "learning_rate": 2e-05,
+      "loss": 0.04267633,
+      "step": 18981
+    },
+    {
+      "epoch": 37.964,
+      "grad_norm": 1.4222880601882935,
+      "learning_rate": 2e-05,
+      "loss": 0.03729562,
+      "step": 18982
+    },
+    {
+      "epoch": 37.966,
+      "grad_norm": 1.1951045989990234,
+      "learning_rate": 2e-05,
+      "loss": 0.04684995,
+      "step": 18983
+    },
+    {
+      "epoch": 37.968,
+      "grad_norm": 1.7370728254318237,
+      "learning_rate": 2e-05,
+      "loss": 0.0354794,
+      "step": 18984
+    },
+    {
+      "epoch": 37.97,
+      "grad_norm": 1.2649239301681519,
+      "learning_rate": 2e-05,
+      "loss": 0.03856641,
+      "step": 18985
+    },
+    {
+      "epoch": 37.972,
+      "grad_norm": 1.2516887187957764,
+      "learning_rate": 2e-05,
+      "loss": 0.03826671,
+      "step": 18986
+    },
+    {
+      "epoch": 37.974,
+      "grad_norm": 1.4137239456176758,
+      "learning_rate": 2e-05,
+      "loss": 0.04873248,
+      "step": 18987
+    },
+    {
+      "epoch": 37.976,
+      "grad_norm": 1.205417275428772,
+      "learning_rate": 2e-05,
+      "loss": 0.03422677,
+      "step": 18988
+    },
+    {
+      "epoch": 37.978,
+      "grad_norm": 1.3862630128860474,
+      "learning_rate": 2e-05,
+      "loss": 0.03816392,
+      "step": 18989
+    },
+    {
+      "epoch": 37.98,
+      "grad_norm": 1.0804787874221802,
+      "learning_rate": 2e-05,
+      "loss": 0.04000157,
+      "step": 18990
+    },
+    {
+      "epoch": 37.982,
+      "grad_norm": 1.2574692964553833,
+      "learning_rate": 2e-05,
+      "loss": 0.04244877,
+      "step": 18991
+    },
+    {
+      "epoch": 37.984,
+      "grad_norm": 0.8911446928977966,
+      "learning_rate": 2e-05,
+      "loss": 0.02988572,
+      "step": 18992
+    },
+    {
+      "epoch": 37.986,
+      "grad_norm": 1.157371997833252,
+      "learning_rate": 2e-05,
+      "loss": 0.03278516,
+      "step": 18993
+    },
+    {
+      "epoch": 37.988,
+      "grad_norm": 1.2546195983886719,
+      "learning_rate": 2e-05,
+      "loss": 0.04516232,
+      "step": 18994
+    },
+    {
+      "epoch": 37.99,
+      "grad_norm": 1.0913975238800049,
+      "learning_rate": 2e-05,
+      "loss": 0.03805106,
+      "step": 18995
+    },
+    {
+      "epoch": 37.992,
+      "grad_norm": 1.3969191312789917,
+      "learning_rate": 2e-05,
+      "loss": 0.04859225,
+      "step": 18996
+    },
+    {
+      "epoch": 37.994,
+      "grad_norm": 1.4129071235656738,
+      "learning_rate": 2e-05,
+      "loss": 0.03075989,
+      "step": 18997
+    },
+    {
+      "epoch": 37.996,
+      "grad_norm": 1.4782291650772095,
+      "learning_rate": 2e-05,
+      "loss": 0.05404112,
+      "step": 18998
+    },
+    {
+      "epoch": 37.998,
+      "grad_norm": 2.0971972942352295,
+      "learning_rate": 2e-05,
+      "loss": 0.06696081,
+      "step": 18999
+    },
+    {
+      "epoch": 38.0,
+      "grad_norm": 1.0580739974975586,
+      "learning_rate": 2e-05,
+      "loss": 0.03982267,
+      "step": 19000
+    },
+    {
+      "epoch": 38.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9900199600798403,
+        "Equal_1": 0.998,
+        "Equal_2": 0.9820359281437125,
+        "Equal_3": 0.9840319361277445,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 0.998003992015968,
+        "LineComparison_3": 1.0,
+        "Parallel_1": 0.9919839679358717,
+        "Parallel_2": 0.9959919839679359,
+        "Parallel_3": 0.992,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 1.0,
+        "Perpendicular_3": 0.8977955911823647,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 1.0,
+        "PointLiesOnCircle_3": 0.9956,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9939879759519038,
+        "PointLiesOnLine_3": 0.9840319361277445
+      },
+      "eval_runtime": 319.6496,
+      "eval_samples_per_second": 32.848,
+      "eval_steps_per_second": 0.657,
+      "step": 19000
+    },
+    {
+      "epoch": 38.002,
+      "grad_norm": 1.2653032541275024,
+      "learning_rate": 2e-05,
+      "loss": 0.05122709,
+      "step": 19001
+    },
+    {
+      "epoch": 38.004,
+      "grad_norm": 1.1466037034988403,
+      "learning_rate": 2e-05,
+      "loss": 0.03442731,
+      "step": 19002
+    },
+    {
+      "epoch": 38.006,
+      "grad_norm": 2.500974178314209,
+      "learning_rate": 2e-05,
+      "loss": 0.05683993,
+      "step": 19003
+    },
+    {
+      "epoch": 38.008,
+      "grad_norm": 1.1632379293441772,
+      "learning_rate": 2e-05,
+      "loss": 0.04220579,
+      "step": 19004
+    },
+    {
+      "epoch": 38.01,
+      "grad_norm": 1.1766409873962402,
+      "learning_rate": 2e-05,
+      "loss": 0.0437943,
+      "step": 19005
+    },
+    {
+      "epoch": 38.012,
+      "grad_norm": 1.2098218202590942,
+      "learning_rate": 2e-05,
+      "loss": 0.04391178,
+      "step": 19006
+    },
+    {
+      "epoch": 38.014,
+      "grad_norm": 1.236122965812683,
+      "learning_rate": 2e-05,
+      "loss": 0.05343842,
+      "step": 19007
+    },
+    {
+      "epoch": 38.016,
+      "grad_norm": 1.6261658668518066,
+      "learning_rate": 2e-05,
+      "loss": 0.06093828,
+      "step": 19008
+    },
+    {
+      "epoch": 38.018,
+      "grad_norm": 1.1033320426940918,
+      "learning_rate": 2e-05,
+      "loss": 0.03286046,
+      "step": 19009
+    },
+    {
+      "epoch": 38.02,
+      "grad_norm": 2.3395283222198486,
+      "learning_rate": 2e-05,
+      "loss": 0.04223632,
+      "step": 19010
+    },
+    {
+      "epoch": 38.022,
+      "grad_norm": 1.4763344526290894,
+      "learning_rate": 2e-05,
+      "loss": 0.05102969,
+      "step": 19011
+    },
+    {
+      "epoch": 38.024,
+      "grad_norm": 1.1282676458358765,
+      "learning_rate": 2e-05,
+      "loss": 0.04249762,
+      "step": 19012
+    },
+    {
+      "epoch": 38.026,
+      "grad_norm": 1.2232784032821655,
+      "learning_rate": 2e-05,
+      "loss": 0.03590241,
+      "step": 19013
+    },
+    {
+      "epoch": 38.028,
+      "grad_norm": 1.187089204788208,
+      "learning_rate": 2e-05,
+      "loss": 0.05354986,
+      "step": 19014
+    },
+    {
+      "epoch": 38.03,
+      "grad_norm": 1.0039637088775635,
+      "learning_rate": 2e-05,
+      "loss": 0.03929551,
+      "step": 19015
+    },
+    {
+      "epoch": 38.032,
+      "grad_norm": 1.1649113893508911,
+      "learning_rate": 2e-05,
+      "loss": 0.04089183,
+      "step": 19016
+    },
+    {
+      "epoch": 38.034,
+      "grad_norm": 1.1364166736602783,
+      "learning_rate": 2e-05,
+      "loss": 0.0420451,
+      "step": 19017
+    },
+    {
+      "epoch": 38.036,
+      "grad_norm": 1.5335687398910522,
+      "learning_rate": 2e-05,
+      "loss": 0.04455702,
+      "step": 19018
+    },
+    {
+      "epoch": 38.038,
+      "grad_norm": 1.4168477058410645,
+      "learning_rate": 2e-05,
+      "loss": 0.05238315,
+      "step": 19019
+    },
+    {
+      "epoch": 38.04,
+      "grad_norm": 1.2343907356262207,
+      "learning_rate": 2e-05,
+      "loss": 0.04423891,
+      "step": 19020
+    },
+    {
+      "epoch": 38.042,
+      "grad_norm": 1.1726269721984863,
+      "learning_rate": 2e-05,
+      "loss": 0.05387194,
+      "step": 19021
+    },
+    {
+      "epoch": 38.044,
+      "grad_norm": 2.447615146636963,
+      "learning_rate": 2e-05,
+      "loss": 0.05135684,
+      "step": 19022
+    },
+    {
+      "epoch": 38.046,
+      "grad_norm": 1.5104469060897827,
+      "learning_rate": 2e-05,
+      "loss": 0.043165,
+      "step": 19023
+    },
+    {
+      "epoch": 38.048,
+      "grad_norm": 1.3275978565216064,
+      "learning_rate": 2e-05,
+      "loss": 0.04400131,
+      "step": 19024
+    },
+    {
+      "epoch": 38.05,
+      "grad_norm": 1.2232415676116943,
+      "learning_rate": 2e-05,
+      "loss": 0.04481234,
+      "step": 19025
+    },
+    {
+      "epoch": 38.052,
+      "grad_norm": 1.1241319179534912,
+      "learning_rate": 2e-05,
+      "loss": 0.04850707,
+      "step": 19026
+    },
+    {
+      "epoch": 38.054,
+      "grad_norm": 1.286199688911438,
+      "learning_rate": 2e-05,
+      "loss": 0.04397567,
+      "step": 19027
+    },
+    {
+      "epoch": 38.056,
+      "grad_norm": 1.2442455291748047,
+      "learning_rate": 2e-05,
+      "loss": 0.04638599,
+      "step": 19028
+    },
+    {
+      "epoch": 38.058,
+      "grad_norm": 1.691164493560791,
+      "learning_rate": 2e-05,
+      "loss": 0.05108991,
+      "step": 19029
+    },
+    {
+      "epoch": 38.06,
+      "grad_norm": 1.2171449661254883,
+      "learning_rate": 2e-05,
+      "loss": 0.04477761,
+      "step": 19030
+    },
+    {
+      "epoch": 38.062,
+      "grad_norm": 1.1396671533584595,
+      "learning_rate": 2e-05,
+      "loss": 0.05528203,
+      "step": 19031
+    },
+    {
+      "epoch": 38.064,
+      "grad_norm": 1.8548091650009155,
+      "learning_rate": 2e-05,
+      "loss": 0.05561259,
+      "step": 19032
+    },
+    {
+      "epoch": 38.066,
+      "grad_norm": 1.0096694231033325,
+      "learning_rate": 2e-05,
+      "loss": 0.03706703,
+      "step": 19033
+    },
+    {
+      "epoch": 38.068,
+      "grad_norm": 1.2540130615234375,
+      "learning_rate": 2e-05,
+      "loss": 0.0528646,
+      "step": 19034
+    },
+    {
+      "epoch": 38.07,
+      "grad_norm": 1.436498999595642,
+      "learning_rate": 2e-05,
+      "loss": 0.05726215,
+      "step": 19035
+    },
+    {
+      "epoch": 38.072,
+      "grad_norm": 1.2053461074829102,
+      "learning_rate": 2e-05,
+      "loss": 0.03862919,
+      "step": 19036
+    },
+    {
+      "epoch": 38.074,
+      "grad_norm": 1.2847776412963867,
+      "learning_rate": 2e-05,
+      "loss": 0.04858331,
+      "step": 19037
+    },
+    {
+      "epoch": 38.076,
+      "grad_norm": 2.5267117023468018,
+      "learning_rate": 2e-05,
+      "loss": 0.04583256,
+      "step": 19038
+    },
+    {
+      "epoch": 38.078,
+      "grad_norm": 1.9496853351593018,
+      "learning_rate": 2e-05,
+      "loss": 0.05861641,
+      "step": 19039
+    },
+    {
+      "epoch": 38.08,
+      "grad_norm": 1.204428791999817,
+      "learning_rate": 2e-05,
+      "loss": 0.04567886,
+      "step": 19040
+    },
+    {
+      "epoch": 38.082,
+      "grad_norm": 1.1530473232269287,
+      "learning_rate": 2e-05,
+      "loss": 0.03973469,
+      "step": 19041
+    },
+    {
+      "epoch": 38.084,
+      "grad_norm": 1.2695164680480957,
+      "learning_rate": 2e-05,
+      "loss": 0.05841737,
+      "step": 19042
+    },
+    {
+      "epoch": 38.086,
+      "grad_norm": 1.5974242687225342,
+      "learning_rate": 2e-05,
+      "loss": 0.04161192,
+      "step": 19043
+    },
+    {
+      "epoch": 38.088,
+      "grad_norm": 1.1564900875091553,
+      "learning_rate": 2e-05,
+      "loss": 0.05044547,
+      "step": 19044
+    },
+    {
+      "epoch": 38.09,
+      "grad_norm": 1.2367781400680542,
+      "learning_rate": 2e-05,
+      "loss": 0.03753296,
+      "step": 19045
+    },
+    {
+      "epoch": 38.092,
+      "grad_norm": 1.287509560585022,
+      "learning_rate": 2e-05,
+      "loss": 0.04154661,
+      "step": 19046
+    },
+    {
+      "epoch": 38.094,
+      "grad_norm": 1.1446880102157593,
+      "learning_rate": 2e-05,
+      "loss": 0.03765508,
+      "step": 19047
+    },
+    {
+      "epoch": 38.096,
+      "grad_norm": 1.177966594696045,
+      "learning_rate": 2e-05,
+      "loss": 0.05387244,
+      "step": 19048
+    },
+    {
+      "epoch": 38.098,
+      "grad_norm": 1.4184907674789429,
+      "learning_rate": 2e-05,
+      "loss": 0.03067533,
+      "step": 19049
+    },
+    {
+      "epoch": 38.1,
+      "grad_norm": 1.4943373203277588,
+      "learning_rate": 2e-05,
+      "loss": 0.05171797,
+      "step": 19050
+    },
+    {
+      "epoch": 38.102,
+      "grad_norm": 1.4353896379470825,
+      "learning_rate": 2e-05,
+      "loss": 0.06505428,
+      "step": 19051
+    },
+    {
+      "epoch": 38.104,
+      "grad_norm": 1.2791118621826172,
+      "learning_rate": 2e-05,
+      "loss": 0.0403844,
+      "step": 19052
+    },
+    {
+      "epoch": 38.106,
+      "grad_norm": 1.1676243543624878,
+      "learning_rate": 2e-05,
+      "loss": 0.05121243,
+      "step": 19053
+    },
+    {
+      "epoch": 38.108,
+      "grad_norm": 1.0785608291625977,
+      "learning_rate": 2e-05,
+      "loss": 0.04726262,
+      "step": 19054
+    },
+    {
+      "epoch": 38.11,
+      "grad_norm": 0.9102175235748291,
+      "learning_rate": 2e-05,
+      "loss": 0.02758621,
+      "step": 19055
+    },
+    {
+      "epoch": 38.112,
+      "grad_norm": 1.3264470100402832,
+      "learning_rate": 2e-05,
+      "loss": 0.05410461,
+      "step": 19056
+    },
+    {
+      "epoch": 38.114,
+      "grad_norm": 1.18822181224823,
+      "learning_rate": 2e-05,
+      "loss": 0.04347523,
+      "step": 19057
+    },
+    {
+      "epoch": 38.116,
+      "grad_norm": 1.4534834623336792,
+      "learning_rate": 2e-05,
+      "loss": 0.04533903,
+      "step": 19058
+    },
+    {
+      "epoch": 38.118,
+      "grad_norm": 1.193938136100769,
+      "learning_rate": 2e-05,
+      "loss": 0.04152393,
+      "step": 19059
+    },
+    {
+      "epoch": 38.12,
+      "grad_norm": 0.9936333298683167,
+      "learning_rate": 2e-05,
+      "loss": 0.02456021,
+      "step": 19060
+    },
+    {
+      "epoch": 38.122,
+      "grad_norm": 1.4597140550613403,
+      "learning_rate": 2e-05,
+      "loss": 0.04886616,
+      "step": 19061
+    },
+    {
+      "epoch": 38.124,
+      "grad_norm": 1.4201385974884033,
+      "learning_rate": 2e-05,
+      "loss": 0.05720758,
+      "step": 19062
+    },
+    {
+      "epoch": 38.126,
+      "grad_norm": 1.197838544845581,
+      "learning_rate": 2e-05,
+      "loss": 0.04132029,
+      "step": 19063
+    },
+    {
+      "epoch": 38.128,
+      "grad_norm": 2.9631714820861816,
+      "learning_rate": 2e-05,
+      "loss": 0.05553018,
+      "step": 19064
+    },
+    {
+      "epoch": 38.13,
+      "grad_norm": 1.32180655002594,
+      "learning_rate": 2e-05,
+      "loss": 0.03765612,
+      "step": 19065
+    },
+    {
+      "epoch": 38.132,
+      "grad_norm": 1.2723172903060913,
+      "learning_rate": 2e-05,
+      "loss": 0.05038822,
+      "step": 19066
+    },
+    {
+      "epoch": 38.134,
+      "grad_norm": 1.2453999519348145,
+      "learning_rate": 2e-05,
+      "loss": 0.04746665,
+      "step": 19067
+    },
+    {
+      "epoch": 38.136,
+      "grad_norm": 1.2901151180267334,
+      "learning_rate": 2e-05,
+      "loss": 0.05191978,
+      "step": 19068
+    },
+    {
+      "epoch": 38.138,
+      "grad_norm": 1.393450379371643,
+      "learning_rate": 2e-05,
+      "loss": 0.04058621,
+      "step": 19069
+    },
+    {
+      "epoch": 38.14,
+      "grad_norm": 3.79284405708313,
+      "learning_rate": 2e-05,
+      "loss": 0.089285,
+      "step": 19070
+    },
+    {
+      "epoch": 38.142,
+      "grad_norm": 1.231909990310669,
+      "learning_rate": 2e-05,
+      "loss": 0.04528357,
+      "step": 19071
+    },
+    {
+      "epoch": 38.144,
+      "grad_norm": 0.9689210057258606,
+      "learning_rate": 2e-05,
+      "loss": 0.03057509,
+      "step": 19072
+    },
+    {
+      "epoch": 38.146,
+      "grad_norm": 1.169950246810913,
+      "learning_rate": 2e-05,
+      "loss": 0.04367265,
+      "step": 19073
+    },
+    {
+      "epoch": 38.148,
+      "grad_norm": 1.002368688583374,
+      "learning_rate": 2e-05,
+      "loss": 0.0362767,
+      "step": 19074
+    },
+    {
+      "epoch": 38.15,
+      "grad_norm": 1.0449949502944946,
+      "learning_rate": 2e-05,
+      "loss": 0.03191894,
+      "step": 19075
+    },
+    {
+      "epoch": 38.152,
+      "grad_norm": 1.23591947555542,
+      "learning_rate": 2e-05,
+      "loss": 0.0624435,
+      "step": 19076
+    },
+    {
+      "epoch": 38.154,
+      "grad_norm": 1.3809751272201538,
+      "learning_rate": 2e-05,
+      "loss": 0.04096212,
+      "step": 19077
+    },
+    {
+      "epoch": 38.156,
+      "grad_norm": 1.3251644372940063,
+      "learning_rate": 2e-05,
+      "loss": 0.04161641,
+      "step": 19078
+    },
+    {
+      "epoch": 38.158,
+      "grad_norm": 1.9493850469589233,
+      "learning_rate": 2e-05,
+      "loss": 0.04814357,
+      "step": 19079
+    },
+    {
+      "epoch": 38.16,
+      "grad_norm": 1.3206309080123901,
+      "learning_rate": 2e-05,
+      "loss": 0.05091274,
+      "step": 19080
+    },
+    {
+      "epoch": 38.162,
+      "grad_norm": 1.2875269651412964,
+      "learning_rate": 2e-05,
+      "loss": 0.04496816,
+      "step": 19081
+    },
+    {
+      "epoch": 38.164,
+      "grad_norm": 0.9897270202636719,
+      "learning_rate": 2e-05,
+      "loss": 0.03702825,
+      "step": 19082
+    },
+    {
+      "epoch": 38.166,
+      "grad_norm": 1.2053303718566895,
+      "learning_rate": 2e-05,
+      "loss": 0.04418178,
+      "step": 19083
+    },
+    {
+      "epoch": 38.168,
+      "grad_norm": 1.2316017150878906,
+      "learning_rate": 2e-05,
+      "loss": 0.05090442,
+      "step": 19084
+    },
+    {
+      "epoch": 38.17,
+      "grad_norm": 1.5326013565063477,
+      "learning_rate": 2e-05,
+      "loss": 0.0485841,
+      "step": 19085
+    },
+    {
+      "epoch": 38.172,
+      "grad_norm": 1.6336983442306519,
+      "learning_rate": 2e-05,
+      "loss": 0.04477915,
+      "step": 19086
+    },
+    {
+      "epoch": 38.174,
+      "grad_norm": 1.2164212465286255,
+      "learning_rate": 2e-05,
+      "loss": 0.0437744,
+      "step": 19087
+    },
+    {
+      "epoch": 38.176,
+      "grad_norm": 1.078202724456787,
+      "learning_rate": 2e-05,
+      "loss": 0.04401769,
+      "step": 19088
+    },
+    {
+      "epoch": 38.178,
+      "grad_norm": 1.8545653820037842,
+      "learning_rate": 2e-05,
+      "loss": 0.05269055,
+      "step": 19089
+    },
+    {
+      "epoch": 38.18,
+      "grad_norm": 1.920038104057312,
+      "learning_rate": 2e-05,
+      "loss": 0.05271509,
+      "step": 19090
+    },
+    {
+      "epoch": 38.182,
+      "grad_norm": 1.242497205734253,
+      "learning_rate": 2e-05,
+      "loss": 0.0489504,
+      "step": 19091
+    },
+    {
+      "epoch": 38.184,
+      "grad_norm": 1.3824882507324219,
+      "learning_rate": 2e-05,
+      "loss": 0.04757221,
+      "step": 19092
+    },
+    {
+      "epoch": 38.186,
+      "grad_norm": 1.007732629776001,
+      "learning_rate": 2e-05,
+      "loss": 0.03562904,
+      "step": 19093
+    },
+    {
+      "epoch": 38.188,
+      "grad_norm": 1.1626834869384766,
+      "learning_rate": 2e-05,
+      "loss": 0.04818817,
+      "step": 19094
+    },
+    {
+      "epoch": 38.19,
+      "grad_norm": 2.008382558822632,
+      "learning_rate": 2e-05,
+      "loss": 0.05179016,
+      "step": 19095
+    },
+    {
+      "epoch": 38.192,
+      "grad_norm": 1.0385682582855225,
+      "learning_rate": 2e-05,
+      "loss": 0.03288983,
+      "step": 19096
+    },
+    {
+      "epoch": 38.194,
+      "grad_norm": 1.3801738023757935,
+      "learning_rate": 2e-05,
+      "loss": 0.05888782,
+      "step": 19097
+    },
+    {
+      "epoch": 38.196,
+      "grad_norm": 1.1773533821105957,
+      "learning_rate": 2e-05,
+      "loss": 0.04121737,
+      "step": 19098
+    },
+    {
+      "epoch": 38.198,
+      "grad_norm": 1.0565184354782104,
+      "learning_rate": 2e-05,
+      "loss": 0.03833776,
+      "step": 19099
+    },
+    {
+      "epoch": 38.2,
+      "grad_norm": 1.225994348526001,
+      "learning_rate": 2e-05,
+      "loss": 0.04604607,
+      "step": 19100
+    },
+    {
+      "epoch": 38.202,
+      "grad_norm": 1.1828093528747559,
+      "learning_rate": 2e-05,
+      "loss": 0.0390872,
+      "step": 19101
+    },
+    {
+      "epoch": 38.204,
+      "grad_norm": 1.1246726512908936,
+      "learning_rate": 2e-05,
+      "loss": 0.04674789,
+      "step": 19102
+    },
+    {
+      "epoch": 38.206,
+      "grad_norm": 1.1224784851074219,
+      "learning_rate": 2e-05,
+      "loss": 0.04065286,
+      "step": 19103
+    },
+    {
+      "epoch": 38.208,
+      "grad_norm": 1.1293448209762573,
+      "learning_rate": 2e-05,
+      "loss": 0.03387726,
+      "step": 19104
+    },
+    {
+      "epoch": 38.21,
+      "grad_norm": 1.6799935102462769,
+      "learning_rate": 2e-05,
+      "loss": 0.04963982,
+      "step": 19105
+    },
+    {
+      "epoch": 38.212,
+      "grad_norm": 0.9128696322441101,
+      "learning_rate": 2e-05,
+      "loss": 0.03350312,
+      "step": 19106
+    },
+    {
+      "epoch": 38.214,
+      "grad_norm": 1.3261280059814453,
+      "learning_rate": 2e-05,
+      "loss": 0.04866079,
+      "step": 19107
+    },
+    {
+      "epoch": 38.216,
+      "grad_norm": 1.1664577722549438,
+      "learning_rate": 2e-05,
+      "loss": 0.03899412,
+      "step": 19108
+    },
+    {
+      "epoch": 38.218,
+      "grad_norm": 1.4886510372161865,
+      "learning_rate": 2e-05,
+      "loss": 0.05114747,
+      "step": 19109
+    },
+    {
+      "epoch": 38.22,
+      "grad_norm": 1.2345974445343018,
+      "learning_rate": 2e-05,
+      "loss": 0.05016024,
+      "step": 19110
+    },
+    {
+      "epoch": 38.222,
+      "grad_norm": 3.129671812057495,
+      "learning_rate": 2e-05,
+      "loss": 0.05178972,
+      "step": 19111
+    },
+    {
+      "epoch": 38.224,
+      "grad_norm": 1.3265200853347778,
+      "learning_rate": 2e-05,
+      "loss": 0.04992515,
+      "step": 19112
+    },
+    {
+      "epoch": 38.226,
+      "grad_norm": 1.359765887260437,
+      "learning_rate": 2e-05,
+      "loss": 0.05563714,
+      "step": 19113
+    },
+    {
+      "epoch": 38.228,
+      "grad_norm": 1.2336148023605347,
+      "learning_rate": 2e-05,
+      "loss": 0.03507271,
+      "step": 19114
+    },
+    {
+      "epoch": 38.23,
+      "grad_norm": 1.2638157606124878,
+      "learning_rate": 2e-05,
+      "loss": 0.06374335,
+      "step": 19115
+    },
+    {
+      "epoch": 38.232,
+      "grad_norm": 1.4349949359893799,
+      "learning_rate": 2e-05,
+      "loss": 0.05200667,
+      "step": 19116
+    },
+    {
+      "epoch": 38.234,
+      "grad_norm": 2.013317584991455,
+      "learning_rate": 2e-05,
+      "loss": 0.03400422,
+      "step": 19117
+    },
+    {
+      "epoch": 38.236,
+      "grad_norm": 1.2802153825759888,
+      "learning_rate": 2e-05,
+      "loss": 0.03919921,
+      "step": 19118
+    },
+    {
+      "epoch": 38.238,
+      "grad_norm": 0.9960325360298157,
+      "learning_rate": 2e-05,
+      "loss": 0.03037401,
+      "step": 19119
+    },
+    {
+      "epoch": 38.24,
+      "grad_norm": 2.0125842094421387,
+      "learning_rate": 2e-05,
+      "loss": 0.05429566,
+      "step": 19120
+    },
+    {
+      "epoch": 38.242,
+      "grad_norm": 1.291225552558899,
+      "learning_rate": 2e-05,
+      "loss": 0.05072563,
+      "step": 19121
+    },
+    {
+      "epoch": 38.244,
+      "grad_norm": 0.9041070938110352,
+      "learning_rate": 2e-05,
+      "loss": 0.03640107,
+      "step": 19122
+    },
+    {
+      "epoch": 38.246,
+      "grad_norm": 2.185776710510254,
+      "learning_rate": 2e-05,
+      "loss": 0.04717995,
+      "step": 19123
+    },
+    {
+      "epoch": 38.248,
+      "grad_norm": 1.4118475914001465,
+      "learning_rate": 2e-05,
+      "loss": 0.03875816,
+      "step": 19124
+    },
+    {
+      "epoch": 38.25,
+      "grad_norm": 1.0311429500579834,
+      "learning_rate": 2e-05,
+      "loss": 0.05210552,
+      "step": 19125
+    },
+    {
+      "epoch": 38.252,
+      "grad_norm": 0.9929075241088867,
+      "learning_rate": 2e-05,
+      "loss": 0.04484344,
+      "step": 19126
+    },
+    {
+      "epoch": 38.254,
+      "grad_norm": 1.266503930091858,
+      "learning_rate": 2e-05,
+      "loss": 0.05835935,
+      "step": 19127
+    },
+    {
+      "epoch": 38.256,
+      "grad_norm": 1.0839781761169434,
+      "learning_rate": 2e-05,
+      "loss": 0.04155935,
+      "step": 19128
+    },
+    {
+      "epoch": 38.258,
+      "grad_norm": 1.6843171119689941,
+      "learning_rate": 2e-05,
+      "loss": 0.05540327,
+      "step": 19129
+    },
+    {
+      "epoch": 38.26,
+      "grad_norm": 1.0930052995681763,
+      "learning_rate": 2e-05,
+      "loss": 0.04292402,
+      "step": 19130
+    },
+    {
+      "epoch": 38.262,
+      "grad_norm": 1.0712100267410278,
+      "learning_rate": 2e-05,
+      "loss": 0.0455328,
+      "step": 19131
+    },
+    {
+      "epoch": 38.264,
+      "grad_norm": 1.5520873069763184,
+      "learning_rate": 2e-05,
+      "loss": 0.05642373,
+      "step": 19132
+    },
+    {
+      "epoch": 38.266,
+      "grad_norm": 1.319484829902649,
+      "learning_rate": 2e-05,
+      "loss": 0.05045997,
+      "step": 19133
+    },
+    {
+      "epoch": 38.268,
+      "grad_norm": 1.1289286613464355,
+      "learning_rate": 2e-05,
+      "loss": 0.03494524,
+      "step": 19134
+    },
+    {
+      "epoch": 38.27,
+      "grad_norm": 2.3391904830932617,
+      "learning_rate": 2e-05,
+      "loss": 0.05442752,
+      "step": 19135
+    },
+    {
+      "epoch": 38.272,
+      "grad_norm": 1.3560073375701904,
+      "learning_rate": 2e-05,
+      "loss": 0.0621442,
+      "step": 19136
+    },
+    {
+      "epoch": 38.274,
+      "grad_norm": 1.094653844833374,
+      "learning_rate": 2e-05,
+      "loss": 0.04534341,
+      "step": 19137
+    },
+    {
+      "epoch": 38.276,
+      "grad_norm": 1.2565248012542725,
+      "learning_rate": 2e-05,
+      "loss": 0.042792,
+      "step": 19138
+    },
+    {
+      "epoch": 38.278,
+      "grad_norm": 1.1656410694122314,
+      "learning_rate": 2e-05,
+      "loss": 0.0445144,
+      "step": 19139
+    },
+    {
+      "epoch": 38.28,
+      "grad_norm": 1.1632707118988037,
+      "learning_rate": 2e-05,
+      "loss": 0.04515301,
+      "step": 19140
+    },
+    {
+      "epoch": 38.282,
+      "grad_norm": 1.305557370185852,
+      "learning_rate": 2e-05,
+      "loss": 0.04854187,
+      "step": 19141
+    },
+    {
+      "epoch": 38.284,
+      "grad_norm": 1.1423861980438232,
+      "learning_rate": 2e-05,
+      "loss": 0.04725296,
+      "step": 19142
+    },
+    {
+      "epoch": 38.286,
+      "grad_norm": 1.8827954530715942,
+      "learning_rate": 2e-05,
+      "loss": 0.05522726,
+      "step": 19143
+    },
+    {
+      "epoch": 38.288,
+      "grad_norm": 1.2071250677108765,
+      "learning_rate": 2e-05,
+      "loss": 0.03631373,
+      "step": 19144
+    },
+    {
+      "epoch": 38.29,
+      "grad_norm": 1.5091750621795654,
+      "learning_rate": 2e-05,
+      "loss": 0.04798866,
+      "step": 19145
+    },
+    {
+      "epoch": 38.292,
+      "grad_norm": 1.3303464651107788,
+      "learning_rate": 2e-05,
+      "loss": 0.0500569,
+      "step": 19146
+    },
+    {
+      "epoch": 38.294,
+      "grad_norm": 1.247398018836975,
+      "learning_rate": 2e-05,
+      "loss": 0.04999819,
+      "step": 19147
+    },
+    {
+      "epoch": 38.296,
+      "grad_norm": 1.242234230041504,
+      "learning_rate": 2e-05,
+      "loss": 0.0521092,
+      "step": 19148
+    },
+    {
+      "epoch": 38.298,
+      "grad_norm": 1.7514032125473022,
+      "learning_rate": 2e-05,
+      "loss": 0.04611446,
+      "step": 19149
+    },
+    {
+      "epoch": 38.3,
+      "grad_norm": 2.0535500049591064,
+      "learning_rate": 2e-05,
+      "loss": 0.04753162,
+      "step": 19150
+    },
+    {
+      "epoch": 38.302,
+      "grad_norm": 1.1260801553726196,
+      "learning_rate": 2e-05,
+      "loss": 0.03980887,
+      "step": 19151
+    },
+    {
+      "epoch": 38.304,
+      "grad_norm": 1.2702425718307495,
+      "learning_rate": 2e-05,
+      "loss": 0.0628961,
+      "step": 19152
+    },
+    {
+      "epoch": 38.306,
+      "grad_norm": 1.7497895956039429,
+      "learning_rate": 2e-05,
+      "loss": 0.04326449,
+      "step": 19153
+    },
+    {
+      "epoch": 38.308,
+      "grad_norm": 1.2419395446777344,
+      "learning_rate": 2e-05,
+      "loss": 0.06270743,
+      "step": 19154
+    },
+    {
+      "epoch": 38.31,
+      "grad_norm": 2.1101601123809814,
+      "learning_rate": 2e-05,
+      "loss": 0.04829663,
+      "step": 19155
+    },
+    {
+      "epoch": 38.312,
+      "grad_norm": 1.3105263710021973,
+      "learning_rate": 2e-05,
+      "loss": 0.05578191,
+      "step": 19156
+    },
+    {
+      "epoch": 38.314,
+      "grad_norm": 1.2866804599761963,
+      "learning_rate": 2e-05,
+      "loss": 0.04316871,
+      "step": 19157
+    },
+    {
+      "epoch": 38.316,
+      "grad_norm": 1.2355014085769653,
+      "learning_rate": 2e-05,
+      "loss": 0.06234382,
+      "step": 19158
+    },
+    {
+      "epoch": 38.318,
+      "grad_norm": 1.1769856214523315,
+      "learning_rate": 2e-05,
+      "loss": 0.02845784,
+      "step": 19159
+    },
+    {
+      "epoch": 38.32,
+      "grad_norm": 1.315886378288269,
+      "learning_rate": 2e-05,
+      "loss": 0.05123859,
+      "step": 19160
+    },
+    {
+      "epoch": 38.322,
+      "grad_norm": 1.1736537218093872,
+      "learning_rate": 2e-05,
+      "loss": 0.05540813,
+      "step": 19161
+    },
+    {
+      "epoch": 38.324,
+      "grad_norm": 1.1168758869171143,
+      "learning_rate": 2e-05,
+      "loss": 0.04026729,
+      "step": 19162
+    },
+    {
+      "epoch": 38.326,
+      "grad_norm": 1.267259120941162,
+      "learning_rate": 2e-05,
+      "loss": 0.05570924,
+      "step": 19163
+    },
+    {
+      "epoch": 38.328,
+      "grad_norm": 0.9030328989028931,
+      "learning_rate": 2e-05,
+      "loss": 0.03402457,
+      "step": 19164
+    },
+    {
+      "epoch": 38.33,
+      "grad_norm": 1.660753846168518,
+      "learning_rate": 2e-05,
+      "loss": 0.06479599,
+      "step": 19165
+    },
+    {
+      "epoch": 38.332,
+      "grad_norm": 1.6835894584655762,
+      "learning_rate": 2e-05,
+      "loss": 0.05633393,
+      "step": 19166
+    },
+    {
+      "epoch": 38.334,
+      "grad_norm": 1.3103505373001099,
+      "learning_rate": 2e-05,
+      "loss": 0.05098883,
+      "step": 19167
+    },
+    {
+      "epoch": 38.336,
+      "grad_norm": 1.1868908405303955,
+      "learning_rate": 2e-05,
+      "loss": 0.04712537,
+      "step": 19168
+    },
+    {
+      "epoch": 38.338,
+      "grad_norm": 1.0892060995101929,
+      "learning_rate": 2e-05,
+      "loss": 0.05058462,
+      "step": 19169
+    },
+    {
+      "epoch": 38.34,
+      "grad_norm": 1.2633219957351685,
+      "learning_rate": 2e-05,
+      "loss": 0.05579946,
+      "step": 19170
+    },
+    {
+      "epoch": 38.342,
+      "grad_norm": 1.1790961027145386,
+      "learning_rate": 2e-05,
+      "loss": 0.04451218,
+      "step": 19171
+    },
+    {
+      "epoch": 38.344,
+      "grad_norm": 1.1820404529571533,
+      "learning_rate": 2e-05,
+      "loss": 0.04325217,
+      "step": 19172
+    },
+    {
+      "epoch": 38.346,
+      "grad_norm": 1.4797415733337402,
+      "learning_rate": 2e-05,
+      "loss": 0.04494851,
+      "step": 19173
+    },
+    {
+      "epoch": 38.348,
+      "grad_norm": 1.6756718158721924,
+      "learning_rate": 2e-05,
+      "loss": 0.04998035,
+      "step": 19174
+    },
+    {
+      "epoch": 38.35,
+      "grad_norm": 1.0109578371047974,
+      "learning_rate": 2e-05,
+      "loss": 0.04028856,
+      "step": 19175
+    },
+    {
+      "epoch": 38.352,
+      "grad_norm": 1.486900806427002,
+      "learning_rate": 2e-05,
+      "loss": 0.04508706,
+      "step": 19176
+    },
+    {
+      "epoch": 38.354,
+      "grad_norm": 1.2009479999542236,
+      "learning_rate": 2e-05,
+      "loss": 0.05440203,
+      "step": 19177
+    },
+    {
+      "epoch": 38.356,
+      "grad_norm": 1.3738353252410889,
+      "learning_rate": 2e-05,
+      "loss": 0.04805979,
+      "step": 19178
+    },
+    {
+      "epoch": 38.358,
+      "grad_norm": 1.2725608348846436,
+      "learning_rate": 2e-05,
+      "loss": 0.05642783,
+      "step": 19179
+    },
+    {
+      "epoch": 38.36,
+      "grad_norm": 1.2467677593231201,
+      "learning_rate": 2e-05,
+      "loss": 0.07003997,
+      "step": 19180
+    },
+    {
+      "epoch": 38.362,
+      "grad_norm": 0.9856001734733582,
+      "learning_rate": 2e-05,
+      "loss": 0.0346032,
+      "step": 19181
+    },
+    {
+      "epoch": 38.364,
+      "grad_norm": 1.4484895467758179,
+      "learning_rate": 2e-05,
+      "loss": 0.0645525,
+      "step": 19182
+    },
+    {
+      "epoch": 38.366,
+      "grad_norm": 3.5123982429504395,
+      "learning_rate": 2e-05,
+      "loss": 0.03936257,
+      "step": 19183
+    },
+    {
+      "epoch": 38.368,
+      "grad_norm": 1.4622162580490112,
+      "learning_rate": 2e-05,
+      "loss": 0.03753219,
+      "step": 19184
+    },
+    {
+      "epoch": 38.37,
+      "grad_norm": 1.103468656539917,
+      "learning_rate": 2e-05,
+      "loss": 0.04038829,
+      "step": 19185
+    },
+    {
+      "epoch": 38.372,
+      "grad_norm": 1.114812970161438,
+      "learning_rate": 2e-05,
+      "loss": 0.04476861,
+      "step": 19186
+    },
+    {
+      "epoch": 38.374,
+      "grad_norm": 1.14467453956604,
+      "learning_rate": 2e-05,
+      "loss": 0.03987251,
+      "step": 19187
+    },
+    {
+      "epoch": 38.376,
+      "grad_norm": 1.4186773300170898,
+      "learning_rate": 2e-05,
+      "loss": 0.05686907,
+      "step": 19188
+    },
+    {
+      "epoch": 38.378,
+      "grad_norm": 2.0654118061065674,
+      "learning_rate": 2e-05,
+      "loss": 0.05757586,
+      "step": 19189
+    },
+    {
+      "epoch": 38.38,
+      "grad_norm": 4.5302414894104,
+      "learning_rate": 2e-05,
+      "loss": 0.04920984,
+      "step": 19190
+    },
+    {
+      "epoch": 38.382,
+      "grad_norm": 1.2781339883804321,
+      "learning_rate": 2e-05,
+      "loss": 0.04979723,
+      "step": 19191
+    },
+    {
+      "epoch": 38.384,
+      "grad_norm": 1.0826224088668823,
+      "learning_rate": 2e-05,
+      "loss": 0.03591222,
+      "step": 19192
+    },
+    {
+      "epoch": 38.386,
+      "grad_norm": 1.9421765804290771,
+      "learning_rate": 2e-05,
+      "loss": 0.05677667,
+      "step": 19193
+    },
+    {
+      "epoch": 38.388,
+      "grad_norm": 1.4869078397750854,
+      "learning_rate": 2e-05,
+      "loss": 0.05089599,
+      "step": 19194
+    },
+    {
+      "epoch": 38.39,
+      "grad_norm": 0.8969262838363647,
+      "learning_rate": 2e-05,
+      "loss": 0.02550054,
+      "step": 19195
+    },
+    {
+      "epoch": 38.392,
+      "grad_norm": 2.2101852893829346,
+      "learning_rate": 2e-05,
+      "loss": 0.07438694,
+      "step": 19196
+    },
+    {
+      "epoch": 38.394,
+      "grad_norm": 1.3274297714233398,
+      "learning_rate": 2e-05,
+      "loss": 0.03078259,
+      "step": 19197
+    },
+    {
+      "epoch": 38.396,
+      "grad_norm": 1.6053683757781982,
+      "learning_rate": 2e-05,
+      "loss": 0.06297421,
+      "step": 19198
+    },
+    {
+      "epoch": 38.398,
+      "grad_norm": 1.0665053129196167,
+      "learning_rate": 2e-05,
+      "loss": 0.03194698,
+      "step": 19199
+    },
+    {
+      "epoch": 38.4,
+      "grad_norm": 1.062880277633667,
+      "learning_rate": 2e-05,
+      "loss": 0.03726966,
+      "step": 19200
+    },
+    {
+      "epoch": 38.402,
+      "grad_norm": 1.6072918176651,
+      "learning_rate": 2e-05,
+      "loss": 0.04004788,
+      "step": 19201
+    },
+    {
+      "epoch": 38.404,
+      "grad_norm": 1.026965856552124,
+      "learning_rate": 2e-05,
+      "loss": 0.04006281,
+      "step": 19202
+    },
+    {
+      "epoch": 38.406,
+      "grad_norm": 1.131760597229004,
+      "learning_rate": 2e-05,
+      "loss": 0.04987944,
+      "step": 19203
+    },
+    {
+      "epoch": 38.408,
+      "grad_norm": 1.1853435039520264,
+      "learning_rate": 2e-05,
+      "loss": 0.04732582,
+      "step": 19204
+    },
+    {
+      "epoch": 38.41,
+      "grad_norm": 1.2606431245803833,
+      "learning_rate": 2e-05,
+      "loss": 0.04184794,
+      "step": 19205
+    },
+    {
+      "epoch": 38.412,
+      "grad_norm": 1.06600022315979,
+      "learning_rate": 2e-05,
+      "loss": 0.04323639,
+      "step": 19206
+    },
+    {
+      "epoch": 38.414,
+      "grad_norm": 1.5144490003585815,
+      "learning_rate": 2e-05,
+      "loss": 0.05575287,
+      "step": 19207
+    },
+    {
+      "epoch": 38.416,
+      "grad_norm": 1.3510489463806152,
+      "learning_rate": 2e-05,
+      "loss": 0.05808666,
+      "step": 19208
+    },
+    {
+      "epoch": 38.418,
+      "grad_norm": 1.2126901149749756,
+      "learning_rate": 2e-05,
+      "loss": 0.04525844,
+      "step": 19209
+    },
+    {
+      "epoch": 38.42,
+      "grad_norm": 1.9178175926208496,
+      "learning_rate": 2e-05,
+      "loss": 0.04605162,
+      "step": 19210
+    },
+    {
+      "epoch": 38.422,
+      "grad_norm": 1.0999705791473389,
+      "learning_rate": 2e-05,
+      "loss": 0.03449193,
+      "step": 19211
+    },
+    {
+      "epoch": 38.424,
+      "grad_norm": 1.1063787937164307,
+      "learning_rate": 2e-05,
+      "loss": 0.04200126,
+      "step": 19212
+    },
+    {
+      "epoch": 38.426,
+      "grad_norm": 1.1517763137817383,
+      "learning_rate": 2e-05,
+      "loss": 0.03924587,
+      "step": 19213
+    },
+    {
+      "epoch": 38.428,
+      "grad_norm": 1.4197931289672852,
+      "learning_rate": 2e-05,
+      "loss": 0.06597655,
+      "step": 19214
+    },
+    {
+      "epoch": 38.43,
+      "grad_norm": 1.267537236213684,
+      "learning_rate": 2e-05,
+      "loss": 0.04415938,
+      "step": 19215
+    },
+    {
+      "epoch": 38.432,
+      "grad_norm": 1.7044909000396729,
+      "learning_rate": 2e-05,
+      "loss": 0.05466124,
+      "step": 19216
+    },
+    {
+      "epoch": 38.434,
+      "grad_norm": 1.201332449913025,
+      "learning_rate": 2e-05,
+      "loss": 0.04956157,
+      "step": 19217
+    },
+    {
+      "epoch": 38.436,
+      "grad_norm": 1.2084065675735474,
+      "learning_rate": 2e-05,
+      "loss": 0.05828648,
+      "step": 19218
+    },
+    {
+      "epoch": 38.438,
+      "grad_norm": 0.9206735491752625,
+      "learning_rate": 2e-05,
+      "loss": 0.0277809,
+      "step": 19219
+    },
+    {
+      "epoch": 38.44,
+      "grad_norm": 1.2664018869400024,
+      "learning_rate": 2e-05,
+      "loss": 0.05090235,
+      "step": 19220
+    },
+    {
+      "epoch": 38.442,
+      "grad_norm": 1.4958431720733643,
+      "learning_rate": 2e-05,
+      "loss": 0.03930274,
+      "step": 19221
+    },
+    {
+      "epoch": 38.444,
+      "grad_norm": 1.0052601099014282,
+      "learning_rate": 2e-05,
+      "loss": 0.03331514,
+      "step": 19222
+    },
+    {
+      "epoch": 38.446,
+      "grad_norm": 1.4637385606765747,
+      "learning_rate": 2e-05,
+      "loss": 0.03536751,
+      "step": 19223
+    },
+    {
+      "epoch": 38.448,
+      "grad_norm": 2.041100263595581,
+      "learning_rate": 2e-05,
+      "loss": 0.05582372,
+      "step": 19224
+    },
+    {
+      "epoch": 38.45,
+      "grad_norm": 2.045063018798828,
+      "learning_rate": 2e-05,
+      "loss": 0.05946861,
+      "step": 19225
+    },
+    {
+      "epoch": 38.452,
+      "grad_norm": 1.763791799545288,
+      "learning_rate": 2e-05,
+      "loss": 0.03987541,
+      "step": 19226
+    },
+    {
+      "epoch": 38.454,
+      "grad_norm": 1.3544304370880127,
+      "learning_rate": 2e-05,
+      "loss": 0.04928677,
+      "step": 19227
+    },
+    {
+      "epoch": 38.456,
+      "grad_norm": 1.0796279907226562,
+      "learning_rate": 2e-05,
+      "loss": 0.04805015,
+      "step": 19228
+    },
+    {
+      "epoch": 38.458,
+      "grad_norm": 1.1521120071411133,
+      "learning_rate": 2e-05,
+      "loss": 0.0500932,
+      "step": 19229
+    },
+    {
+      "epoch": 38.46,
+      "grad_norm": 1.0432672500610352,
+      "learning_rate": 2e-05,
+      "loss": 0.0380001,
+      "step": 19230
+    },
+    {
+      "epoch": 38.462,
+      "grad_norm": 1.323717474937439,
+      "learning_rate": 2e-05,
+      "loss": 0.04665473,
+      "step": 19231
+    },
+    {
+      "epoch": 38.464,
+      "grad_norm": 1.5594381093978882,
+      "learning_rate": 2e-05,
+      "loss": 0.0410453,
+      "step": 19232
+    },
+    {
+      "epoch": 38.466,
+      "grad_norm": 1.2671560049057007,
+      "learning_rate": 2e-05,
+      "loss": 0.04604392,
+      "step": 19233
+    },
+    {
+      "epoch": 38.468,
+      "grad_norm": 2.227236032485962,
+      "learning_rate": 2e-05,
+      "loss": 0.06509386,
+      "step": 19234
+    },
+    {
+      "epoch": 38.47,
+      "grad_norm": 1.537432074546814,
+      "learning_rate": 2e-05,
+      "loss": 0.05568233,
+      "step": 19235
+    },
+    {
+      "epoch": 38.472,
+      "grad_norm": 1.411525011062622,
+      "learning_rate": 2e-05,
+      "loss": 0.05239115,
+      "step": 19236
+    },
+    {
+      "epoch": 38.474,
+      "grad_norm": 1.2611944675445557,
+      "learning_rate": 2e-05,
+      "loss": 0.03326932,
+      "step": 19237
+    },
+    {
+      "epoch": 38.476,
+      "grad_norm": 1.2077014446258545,
+      "learning_rate": 2e-05,
+      "loss": 0.05218486,
+      "step": 19238
+    },
+    {
+      "epoch": 38.478,
+      "grad_norm": 1.6436618566513062,
+      "learning_rate": 2e-05,
+      "loss": 0.05138172,
+      "step": 19239
+    },
+    {
+      "epoch": 38.48,
+      "grad_norm": 1.2838315963745117,
+      "learning_rate": 2e-05,
+      "loss": 0.04669505,
+      "step": 19240
+    },
+    {
+      "epoch": 38.482,
+      "grad_norm": 1.3253170251846313,
+      "learning_rate": 2e-05,
+      "loss": 0.05545774,
+      "step": 19241
+    },
+    {
+      "epoch": 38.484,
+      "grad_norm": 1.3036000728607178,
+      "learning_rate": 2e-05,
+      "loss": 0.05103177,
+      "step": 19242
+    },
+    {
+      "epoch": 38.486,
+      "grad_norm": 1.1623722314834595,
+      "learning_rate": 2e-05,
+      "loss": 0.04469373,
+      "step": 19243
+    },
+    {
+      "epoch": 38.488,
+      "grad_norm": 1.3823708295822144,
+      "learning_rate": 2e-05,
+      "loss": 0.05241356,
+      "step": 19244
+    },
+    {
+      "epoch": 38.49,
+      "grad_norm": 1.8385578393936157,
+      "learning_rate": 2e-05,
+      "loss": 0.05396533,
+      "step": 19245
+    },
+    {
+      "epoch": 38.492,
+      "grad_norm": 1.2699154615402222,
+      "learning_rate": 2e-05,
+      "loss": 0.05115933,
+      "step": 19246
+    },
+    {
+      "epoch": 38.494,
+      "grad_norm": 1.23024582862854,
+      "learning_rate": 2e-05,
+      "loss": 0.03833042,
+      "step": 19247
+    },
+    {
+      "epoch": 38.496,
+      "grad_norm": 1.3315479755401611,
+      "learning_rate": 2e-05,
+      "loss": 0.0449665,
+      "step": 19248
+    },
+    {
+      "epoch": 38.498,
+      "grad_norm": 1.5424014329910278,
+      "learning_rate": 2e-05,
+      "loss": 0.04347795,
+      "step": 19249
+    },
+    {
+      "epoch": 38.5,
+      "grad_norm": 1.245294213294983,
+      "learning_rate": 2e-05,
+      "loss": 0.05182352,
+      "step": 19250
+    },
+    {
+      "epoch": 38.502,
+      "grad_norm": 2.529345750808716,
+      "learning_rate": 2e-05,
+      "loss": 0.04279209,
+      "step": 19251
+    },
+    {
+      "epoch": 38.504,
+      "grad_norm": 1.083033800125122,
+      "learning_rate": 2e-05,
+      "loss": 0.04436203,
+      "step": 19252
+    },
+    {
+      "epoch": 38.506,
+      "grad_norm": 1.3808947801589966,
+      "learning_rate": 2e-05,
+      "loss": 0.06541286,
+      "step": 19253
+    },
+    {
+      "epoch": 38.508,
+      "grad_norm": 1.0596139430999756,
+      "learning_rate": 2e-05,
+      "loss": 0.03044969,
+      "step": 19254
+    },
+    {
+      "epoch": 38.51,
+      "grad_norm": 1.435849666595459,
+      "learning_rate": 2e-05,
+      "loss": 0.06559452,
+      "step": 19255
+    },
+    {
+      "epoch": 38.512,
+      "grad_norm": 1.3231841325759888,
+      "learning_rate": 2e-05,
+      "loss": 0.05220275,
+      "step": 19256
+    },
+    {
+      "epoch": 38.514,
+      "grad_norm": 1.2284640073776245,
+      "learning_rate": 2e-05,
+      "loss": 0.05421726,
+      "step": 19257
+    },
+    {
+      "epoch": 38.516,
+      "grad_norm": 1.0423113107681274,
+      "learning_rate": 2e-05,
+      "loss": 0.0366753,
+      "step": 19258
+    },
+    {
+      "epoch": 38.518,
+      "grad_norm": 1.5633624792099,
+      "learning_rate": 2e-05,
+      "loss": 0.05434642,
+      "step": 19259
+    },
+    {
+      "epoch": 38.52,
+      "grad_norm": 2.032729148864746,
+      "learning_rate": 2e-05,
+      "loss": 0.04470177,
+      "step": 19260
+    },
+    {
+      "epoch": 38.522,
+      "grad_norm": 1.3375166654586792,
+      "learning_rate": 2e-05,
+      "loss": 0.04499324,
+      "step": 19261
+    },
+    {
+      "epoch": 38.524,
+      "grad_norm": 1.552940011024475,
+      "learning_rate": 2e-05,
+      "loss": 0.05302621,
+      "step": 19262
+    },
+    {
+      "epoch": 38.526,
+      "grad_norm": 1.896203637123108,
+      "learning_rate": 2e-05,
+      "loss": 0.05639236,
+      "step": 19263
+    },
+    {
+      "epoch": 38.528,
+      "grad_norm": 1.3673580884933472,
+      "learning_rate": 2e-05,
+      "loss": 0.05501523,
+      "step": 19264
+    },
+    {
+      "epoch": 38.53,
+      "grad_norm": 1.3088792562484741,
+      "learning_rate": 2e-05,
+      "loss": 0.04335637,
+      "step": 19265
+    },
+    {
+      "epoch": 38.532,
+      "grad_norm": 1.1292967796325684,
+      "learning_rate": 2e-05,
+      "loss": 0.04752572,
+      "step": 19266
+    },
+    {
+      "epoch": 38.534,
+      "grad_norm": 1.1001681089401245,
+      "learning_rate": 2e-05,
+      "loss": 0.04105626,
+      "step": 19267
+    },
+    {
+      "epoch": 38.536,
+      "grad_norm": 1.1444545984268188,
+      "learning_rate": 2e-05,
+      "loss": 0.03688351,
+      "step": 19268
+    },
+    {
+      "epoch": 38.538,
+      "grad_norm": 1.626502275466919,
+      "learning_rate": 2e-05,
+      "loss": 0.0603931,
+      "step": 19269
+    },
+    {
+      "epoch": 38.54,
+      "grad_norm": 2.7609822750091553,
+      "learning_rate": 2e-05,
+      "loss": 0.04431499,
+      "step": 19270
+    },
+    {
+      "epoch": 38.542,
+      "grad_norm": 1.9874058961868286,
+      "learning_rate": 2e-05,
+      "loss": 0.04629558,
+      "step": 19271
+    },
+    {
+      "epoch": 38.544,
+      "grad_norm": 1.3202643394470215,
+      "learning_rate": 2e-05,
+      "loss": 0.0395207,
+      "step": 19272
+    },
+    {
+      "epoch": 38.546,
+      "grad_norm": 3.389218807220459,
+      "learning_rate": 2e-05,
+      "loss": 0.03313368,
+      "step": 19273
+    },
+    {
+      "epoch": 38.548,
+      "grad_norm": 1.8491859436035156,
+      "learning_rate": 2e-05,
+      "loss": 0.05389631,
+      "step": 19274
+    },
+    {
+      "epoch": 38.55,
+      "grad_norm": 1.5536370277404785,
+      "learning_rate": 2e-05,
+      "loss": 0.04745733,
+      "step": 19275
+    },
+    {
+      "epoch": 38.552,
+      "grad_norm": 1.2798070907592773,
+      "learning_rate": 2e-05,
+      "loss": 0.04839146,
+      "step": 19276
+    },
+    {
+      "epoch": 38.554,
+      "grad_norm": 1.372031807899475,
+      "learning_rate": 2e-05,
+      "loss": 0.04501195,
+      "step": 19277
+    },
+    {
+      "epoch": 38.556,
+      "grad_norm": 2.0433499813079834,
+      "learning_rate": 2e-05,
+      "loss": 0.04445722,
+      "step": 19278
+    },
+    {
+      "epoch": 38.558,
+      "grad_norm": 2.050704002380371,
+      "learning_rate": 2e-05,
+      "loss": 0.06163099,
+      "step": 19279
+    },
+    {
+      "epoch": 38.56,
+      "grad_norm": 1.116773247718811,
+      "learning_rate": 2e-05,
+      "loss": 0.04683295,
+      "step": 19280
+    },
+    {
+      "epoch": 38.562,
+      "grad_norm": 1.1566107273101807,
+      "learning_rate": 2e-05,
+      "loss": 0.04381316,
+      "step": 19281
+    },
+    {
+      "epoch": 38.564,
+      "grad_norm": 1.1835641860961914,
+      "learning_rate": 2e-05,
+      "loss": 0.05227757,
+      "step": 19282
+    },
+    {
+      "epoch": 38.566,
+      "grad_norm": 1.1514893770217896,
+      "learning_rate": 2e-05,
+      "loss": 0.04288551,
+      "step": 19283
+    },
+    {
+      "epoch": 38.568,
+      "grad_norm": 0.9600772857666016,
+      "learning_rate": 2e-05,
+      "loss": 0.03298654,
+      "step": 19284
+    },
+    {
+      "epoch": 38.57,
+      "grad_norm": 1.2036709785461426,
+      "learning_rate": 2e-05,
+      "loss": 0.050576,
+      "step": 19285
+    },
+    {
+      "epoch": 38.572,
+      "grad_norm": 3.4883370399475098,
+      "learning_rate": 2e-05,
+      "loss": 0.05142732,
+      "step": 19286
+    },
+    {
+      "epoch": 38.574,
+      "grad_norm": 1.1536023616790771,
+      "learning_rate": 2e-05,
+      "loss": 0.04423029,
+      "step": 19287
+    },
+    {
+      "epoch": 38.576,
+      "grad_norm": 1.2535980939865112,
+      "learning_rate": 2e-05,
+      "loss": 0.04615848,
+      "step": 19288
+    },
+    {
+      "epoch": 38.578,
+      "grad_norm": 1.3373180627822876,
+      "learning_rate": 2e-05,
+      "loss": 0.0535438,
+      "step": 19289
+    },
+    {
+      "epoch": 38.58,
+      "grad_norm": 1.3583929538726807,
+      "learning_rate": 2e-05,
+      "loss": 0.04743036,
+      "step": 19290
+    },
+    {
+      "epoch": 38.582,
+      "grad_norm": 1.2381739616394043,
+      "learning_rate": 2e-05,
+      "loss": 0.04024129,
+      "step": 19291
+    },
+    {
+      "epoch": 38.584,
+      "grad_norm": 1.3908512592315674,
+      "learning_rate": 2e-05,
+      "loss": 0.03594154,
+      "step": 19292
+    },
+    {
+      "epoch": 38.586,
+      "grad_norm": 2.500706911087036,
+      "learning_rate": 2e-05,
+      "loss": 0.0516097,
+      "step": 19293
+    },
+    {
+      "epoch": 38.588,
+      "grad_norm": 1.6076672077178955,
+      "learning_rate": 2e-05,
+      "loss": 0.04902358,
+      "step": 19294
+    },
+    {
+      "epoch": 38.59,
+      "grad_norm": 1.1396098136901855,
+      "learning_rate": 2e-05,
+      "loss": 0.04243046,
+      "step": 19295
+    },
+    {
+      "epoch": 38.592,
+      "grad_norm": 1.1097677946090698,
+      "learning_rate": 2e-05,
+      "loss": 0.04250278,
+      "step": 19296
+    },
+    {
+      "epoch": 38.594,
+      "grad_norm": 1.5975714921951294,
+      "learning_rate": 2e-05,
+      "loss": 0.05274806,
+      "step": 19297
+    },
+    {
+      "epoch": 38.596,
+      "grad_norm": 1.1447633504867554,
+      "learning_rate": 2e-05,
+      "loss": 0.03151781,
+      "step": 19298
+    },
+    {
+      "epoch": 38.598,
+      "grad_norm": 0.938985288143158,
+      "learning_rate": 2e-05,
+      "loss": 0.03843331,
+      "step": 19299
+    },
+    {
+      "epoch": 38.6,
+      "grad_norm": 2.130866050720215,
+      "learning_rate": 2e-05,
+      "loss": 0.06860761,
+      "step": 19300
+    },
+    {
+      "epoch": 38.602,
+      "grad_norm": 1.0700651407241821,
+      "learning_rate": 2e-05,
+      "loss": 0.03487083,
+      "step": 19301
+    },
+    {
+      "epoch": 38.604,
+      "grad_norm": 1.744493007659912,
+      "learning_rate": 2e-05,
+      "loss": 0.06696406,
+      "step": 19302
+    },
+    {
+      "epoch": 38.606,
+      "grad_norm": 1.0361813306808472,
+      "learning_rate": 2e-05,
+      "loss": 0.04734627,
+      "step": 19303
+    },
+    {
+      "epoch": 38.608,
+      "grad_norm": 2.7028353214263916,
+      "learning_rate": 2e-05,
+      "loss": 0.0526311,
+      "step": 19304
+    },
+    {
+      "epoch": 38.61,
+      "grad_norm": 1.9693485498428345,
+      "learning_rate": 2e-05,
+      "loss": 0.04218089,
+      "step": 19305
+    },
+    {
+      "epoch": 38.612,
+      "grad_norm": 1.2493609189987183,
+      "learning_rate": 2e-05,
+      "loss": 0.05733298,
+      "step": 19306
+    },
+    {
+      "epoch": 38.614,
+      "grad_norm": 1.406848669052124,
+      "learning_rate": 2e-05,
+      "loss": 0.04433504,
+      "step": 19307
+    },
+    {
+      "epoch": 38.616,
+      "grad_norm": 1.199169397354126,
+      "learning_rate": 2e-05,
+      "loss": 0.04781777,
+      "step": 19308
+    },
+    {
+      "epoch": 38.618,
+      "grad_norm": 1.0775588750839233,
+      "learning_rate": 2e-05,
+      "loss": 0.0368211,
+      "step": 19309
+    },
+    {
+      "epoch": 38.62,
+      "grad_norm": 1.0789852142333984,
+      "learning_rate": 2e-05,
+      "loss": 0.05344281,
+      "step": 19310
+    },
+    {
+      "epoch": 38.622,
+      "grad_norm": 1.7068625688552856,
+      "learning_rate": 2e-05,
+      "loss": 0.03457305,
+      "step": 19311
+    },
+    {
+      "epoch": 38.624,
+      "grad_norm": 1.6094670295715332,
+      "learning_rate": 2e-05,
+      "loss": 0.04848836,
+      "step": 19312
+    },
+    {
+      "epoch": 38.626,
+      "grad_norm": 1.0214130878448486,
+      "learning_rate": 2e-05,
+      "loss": 0.03269782,
+      "step": 19313
+    },
+    {
+      "epoch": 38.628,
+      "grad_norm": 2.0523335933685303,
+      "learning_rate": 2e-05,
+      "loss": 0.04664942,
+      "step": 19314
+    },
+    {
+      "epoch": 38.63,
+      "grad_norm": 1.2456196546554565,
+      "learning_rate": 2e-05,
+      "loss": 0.03976301,
+      "step": 19315
+    },
+    {
+      "epoch": 38.632,
+      "grad_norm": 1.076731562614441,
+      "learning_rate": 2e-05,
+      "loss": 0.03696981,
+      "step": 19316
+    },
+    {
+      "epoch": 38.634,
+      "grad_norm": 1.4856189489364624,
+      "learning_rate": 2e-05,
+      "loss": 0.05270289,
+      "step": 19317
+    },
+    {
+      "epoch": 38.636,
+      "grad_norm": 1.0082216262817383,
+      "learning_rate": 2e-05,
+      "loss": 0.03996618,
+      "step": 19318
+    },
+    {
+      "epoch": 38.638,
+      "grad_norm": 1.2000486850738525,
+      "learning_rate": 2e-05,
+      "loss": 0.05204409,
+      "step": 19319
+    },
+    {
+      "epoch": 38.64,
+      "grad_norm": 1.0704103708267212,
+      "learning_rate": 2e-05,
+      "loss": 0.04252416,
+      "step": 19320
+    },
+    {
+      "epoch": 38.642,
+      "grad_norm": 1.4279651641845703,
+      "learning_rate": 2e-05,
+      "loss": 0.05704008,
+      "step": 19321
+    },
+    {
+      "epoch": 38.644,
+      "grad_norm": 1.1782804727554321,
+      "learning_rate": 2e-05,
+      "loss": 0.04499515,
+      "step": 19322
+    },
+    {
+      "epoch": 38.646,
+      "grad_norm": 1.00314199924469,
+      "learning_rate": 2e-05,
+      "loss": 0.03433149,
+      "step": 19323
+    },
+    {
+      "epoch": 38.648,
+      "grad_norm": 1.2116609811782837,
+      "learning_rate": 2e-05,
+      "loss": 0.04214825,
+      "step": 19324
+    },
+    {
+      "epoch": 38.65,
+      "grad_norm": 1.176286220550537,
+      "learning_rate": 2e-05,
+      "loss": 0.04334562,
+      "step": 19325
+    },
+    {
+      "epoch": 38.652,
+      "grad_norm": 1.6107664108276367,
+      "learning_rate": 2e-05,
+      "loss": 0.04943176,
+      "step": 19326
+    },
+    {
+      "epoch": 38.654,
+      "grad_norm": 1.3699370622634888,
+      "learning_rate": 2e-05,
+      "loss": 0.05509346,
+      "step": 19327
+    },
+    {
+      "epoch": 38.656,
+      "grad_norm": 2.1541030406951904,
+      "learning_rate": 2e-05,
+      "loss": 0.0614479,
+      "step": 19328
+    },
+    {
+      "epoch": 38.658,
+      "grad_norm": 1.3070214986801147,
+      "learning_rate": 2e-05,
+      "loss": 0.04495452,
+      "step": 19329
+    },
+    {
+      "epoch": 38.66,
+      "grad_norm": 1.4150477647781372,
+      "learning_rate": 2e-05,
+      "loss": 0.04712527,
+      "step": 19330
+    },
+    {
+      "epoch": 38.662,
+      "grad_norm": 2.080083131790161,
+      "learning_rate": 2e-05,
+      "loss": 0.04044828,
+      "step": 19331
+    },
+    {
+      "epoch": 38.664,
+      "grad_norm": 1.225448489189148,
+      "learning_rate": 2e-05,
+      "loss": 0.05349731,
+      "step": 19332
+    },
+    {
+      "epoch": 38.666,
+      "grad_norm": 1.4030938148498535,
+      "learning_rate": 2e-05,
+      "loss": 0.04746981,
+      "step": 19333
+    },
+    {
+      "epoch": 38.668,
+      "grad_norm": 1.162904143333435,
+      "learning_rate": 2e-05,
+      "loss": 0.05157799,
+      "step": 19334
+    },
+    {
+      "epoch": 38.67,
+      "grad_norm": 1.4816622734069824,
+      "learning_rate": 2e-05,
+      "loss": 0.04534583,
+      "step": 19335
+    },
+    {
+      "epoch": 38.672,
+      "grad_norm": 3.349114179611206,
+      "learning_rate": 2e-05,
+      "loss": 0.04592524,
+      "step": 19336
+    },
+    {
+      "epoch": 38.674,
+      "grad_norm": 1.2360336780548096,
+      "learning_rate": 2e-05,
+      "loss": 0.04279222,
+      "step": 19337
+    },
+    {
+      "epoch": 38.676,
+      "grad_norm": 2.137404441833496,
+      "learning_rate": 2e-05,
+      "loss": 0.04810013,
+      "step": 19338
+    },
+    {
+      "epoch": 38.678,
+      "grad_norm": 1.7355518341064453,
+      "learning_rate": 2e-05,
+      "loss": 0.0334636,
+      "step": 19339
+    },
+    {
+      "epoch": 38.68,
+      "grad_norm": 1.9419608116149902,
+      "learning_rate": 2e-05,
+      "loss": 0.05887587,
+      "step": 19340
+    },
+    {
+      "epoch": 38.682,
+      "grad_norm": 1.431899905204773,
+      "learning_rate": 2e-05,
+      "loss": 0.05057277,
+      "step": 19341
+    },
+    {
+      "epoch": 38.684,
+      "grad_norm": 1.284608244895935,
+      "learning_rate": 2e-05,
+      "loss": 0.04898845,
+      "step": 19342
+    },
+    {
+      "epoch": 38.686,
+      "grad_norm": 1.426267385482788,
+      "learning_rate": 2e-05,
+      "loss": 0.05257843,
+      "step": 19343
+    },
+    {
+      "epoch": 38.688,
+      "grad_norm": 1.2423020601272583,
+      "learning_rate": 2e-05,
+      "loss": 0.05175748,
+      "step": 19344
+    },
+    {
+      "epoch": 38.69,
+      "grad_norm": 1.3101670742034912,
+      "learning_rate": 2e-05,
+      "loss": 0.04395001,
+      "step": 19345
+    },
+    {
+      "epoch": 38.692,
+      "grad_norm": 1.2051000595092773,
+      "learning_rate": 2e-05,
+      "loss": 0.04174311,
+      "step": 19346
+    },
+    {
+      "epoch": 38.694,
+      "grad_norm": 1.1244724988937378,
+      "learning_rate": 2e-05,
+      "loss": 0.04402897,
+      "step": 19347
+    },
+    {
+      "epoch": 38.696,
+      "grad_norm": 1.2426832914352417,
+      "learning_rate": 2e-05,
+      "loss": 0.04704241,
+      "step": 19348
+    },
+    {
+      "epoch": 38.698,
+      "grad_norm": 1.4777288436889648,
+      "learning_rate": 2e-05,
+      "loss": 0.06701204,
+      "step": 19349
+    },
+    {
+      "epoch": 38.7,
+      "grad_norm": 2.267500162124634,
+      "learning_rate": 2e-05,
+      "loss": 0.05487394,
+      "step": 19350
+    },
+    {
+      "epoch": 38.702,
+      "grad_norm": 1.73554527759552,
+      "learning_rate": 2e-05,
+      "loss": 0.05015863,
+      "step": 19351
+    },
+    {
+      "epoch": 38.704,
+      "grad_norm": 2.264845371246338,
+      "learning_rate": 2e-05,
+      "loss": 0.03342777,
+      "step": 19352
+    },
+    {
+      "epoch": 38.706,
+      "grad_norm": 1.1378469467163086,
+      "learning_rate": 2e-05,
+      "loss": 0.04409365,
+      "step": 19353
+    },
+    {
+      "epoch": 38.708,
+      "grad_norm": 1.3917489051818848,
+      "learning_rate": 2e-05,
+      "loss": 0.04507478,
+      "step": 19354
+    },
+    {
+      "epoch": 38.71,
+      "grad_norm": 1.3060683012008667,
+      "learning_rate": 2e-05,
+      "loss": 0.05471382,
+      "step": 19355
+    },
+    {
+      "epoch": 38.712,
+      "grad_norm": 1.3260350227355957,
+      "learning_rate": 2e-05,
+      "loss": 0.04912686,
+      "step": 19356
+    },
+    {
+      "epoch": 38.714,
+      "grad_norm": 2.3827006816864014,
+      "learning_rate": 2e-05,
+      "loss": 0.05015289,
+      "step": 19357
+    },
+    {
+      "epoch": 38.716,
+      "grad_norm": 1.5596778392791748,
+      "learning_rate": 2e-05,
+      "loss": 0.0522569,
+      "step": 19358
+    },
+    {
+      "epoch": 38.718,
+      "grad_norm": 2.026498556137085,
+      "learning_rate": 2e-05,
+      "loss": 0.04310253,
+      "step": 19359
+    },
+    {
+      "epoch": 38.72,
+      "grad_norm": 1.090510606765747,
+      "learning_rate": 2e-05,
+      "loss": 0.03988282,
+      "step": 19360
+    },
+    {
+      "epoch": 38.722,
+      "grad_norm": 1.136027455329895,
+      "learning_rate": 2e-05,
+      "loss": 0.04197805,
+      "step": 19361
+    },
+    {
+      "epoch": 38.724,
+      "grad_norm": 1.0448952913284302,
+      "learning_rate": 2e-05,
+      "loss": 0.04863923,
+      "step": 19362
+    },
+    {
+      "epoch": 38.726,
+      "grad_norm": 1.3965747356414795,
+      "learning_rate": 2e-05,
+      "loss": 0.05056681,
+      "step": 19363
+    },
+    {
+      "epoch": 38.728,
+      "grad_norm": 1.705721378326416,
+      "learning_rate": 2e-05,
+      "loss": 0.05423497,
+      "step": 19364
+    },
+    {
+      "epoch": 38.73,
+      "grad_norm": 2.248636245727539,
+      "learning_rate": 2e-05,
+      "loss": 0.06463277,
+      "step": 19365
+    },
+    {
+      "epoch": 38.732,
+      "grad_norm": 1.4248642921447754,
+      "learning_rate": 2e-05,
+      "loss": 0.04670746,
+      "step": 19366
+    },
+    {
+      "epoch": 38.734,
+      "grad_norm": 1.1761763095855713,
+      "learning_rate": 2e-05,
+      "loss": 0.05271841,
+      "step": 19367
+    },
+    {
+      "epoch": 38.736,
+      "grad_norm": 1.2061562538146973,
+      "learning_rate": 2e-05,
+      "loss": 0.04730795,
+      "step": 19368
+    },
+    {
+      "epoch": 38.738,
+      "grad_norm": 1.3923081159591675,
+      "learning_rate": 2e-05,
+      "loss": 0.05840427,
+      "step": 19369
+    },
+    {
+      "epoch": 38.74,
+      "grad_norm": 1.250772476196289,
+      "learning_rate": 2e-05,
+      "loss": 0.04998203,
+      "step": 19370
+    },
+    {
+      "epoch": 38.742,
+      "grad_norm": 1.2058557271957397,
+      "learning_rate": 2e-05,
+      "loss": 0.04659589,
+      "step": 19371
+    },
+    {
+      "epoch": 38.744,
+      "grad_norm": 1.7007390260696411,
+      "learning_rate": 2e-05,
+      "loss": 0.05928213,
+      "step": 19372
+    },
+    {
+      "epoch": 38.746,
+      "grad_norm": 2.651254415512085,
+      "learning_rate": 2e-05,
+      "loss": 0.05975854,
+      "step": 19373
+    },
+    {
+      "epoch": 38.748,
+      "grad_norm": 3.1047842502593994,
+      "learning_rate": 2e-05,
+      "loss": 0.06387729,
+      "step": 19374
+    },
+    {
+      "epoch": 38.75,
+      "grad_norm": 5.55999755859375,
+      "learning_rate": 2e-05,
+      "loss": 0.06207875,
+      "step": 19375
+    },
+    {
+      "epoch": 38.752,
+      "grad_norm": 1.1087514162063599,
+      "learning_rate": 2e-05,
+      "loss": 0.03848689,
+      "step": 19376
+    },
+    {
+      "epoch": 38.754,
+      "grad_norm": 1.1218056678771973,
+      "learning_rate": 2e-05,
+      "loss": 0.045539,
+      "step": 19377
+    },
+    {
+      "epoch": 38.756,
+      "grad_norm": 1.1395851373672485,
+      "learning_rate": 2e-05,
+      "loss": 0.05567363,
+      "step": 19378
+    },
+    {
+      "epoch": 38.758,
+      "grad_norm": 1.1848268508911133,
+      "learning_rate": 2e-05,
+      "loss": 0.04403425,
+      "step": 19379
+    },
+    {
+      "epoch": 38.76,
+      "grad_norm": 1.0772857666015625,
+      "learning_rate": 2e-05,
+      "loss": 0.04982436,
+      "step": 19380
+    },
+    {
+      "epoch": 38.762,
+      "grad_norm": 1.0320247411727905,
+      "learning_rate": 2e-05,
+      "loss": 0.03512707,
+      "step": 19381
+    },
+    {
+      "epoch": 38.764,
+      "grad_norm": 1.3068004846572876,
+      "learning_rate": 2e-05,
+      "loss": 0.04124657,
+      "step": 19382
+    },
+    {
+      "epoch": 38.766,
+      "grad_norm": 1.5414135456085205,
+      "learning_rate": 2e-05,
+      "loss": 0.05403698,
+      "step": 19383
+    },
+    {
+      "epoch": 38.768,
+      "grad_norm": 1.0777575969696045,
+      "learning_rate": 2e-05,
+      "loss": 0.05617758,
+      "step": 19384
+    },
+    {
+      "epoch": 38.77,
+      "grad_norm": 1.1199369430541992,
+      "learning_rate": 2e-05,
+      "loss": 0.04268428,
+      "step": 19385
+    },
+    {
+      "epoch": 38.772,
+      "grad_norm": 1.1600946187973022,
+      "learning_rate": 2e-05,
+      "loss": 0.0405513,
+      "step": 19386
+    },
+    {
+      "epoch": 38.774,
+      "grad_norm": 3.984130382537842,
+      "learning_rate": 2e-05,
+      "loss": 0.05990764,
+      "step": 19387
+    },
+    {
+      "epoch": 38.776,
+      "grad_norm": 1.1094694137573242,
+      "learning_rate": 2e-05,
+      "loss": 0.03420043,
+      "step": 19388
+    },
+    {
+      "epoch": 38.778,
+      "grad_norm": 2.123666286468506,
+      "learning_rate": 2e-05,
+      "loss": 0.05570301,
+      "step": 19389
+    },
+    {
+      "epoch": 38.78,
+      "grad_norm": 1.096800446510315,
+      "learning_rate": 2e-05,
+      "loss": 0.04737834,
+      "step": 19390
+    },
+    {
+      "epoch": 38.782,
+      "grad_norm": 1.0921696424484253,
+      "learning_rate": 2e-05,
+      "loss": 0.04886656,
+      "step": 19391
+    },
+    {
+      "epoch": 38.784,
+      "grad_norm": 1.10837984085083,
+      "learning_rate": 2e-05,
+      "loss": 0.03572984,
+      "step": 19392
+    },
+    {
+      "epoch": 38.786,
+      "grad_norm": 0.9516921639442444,
+      "learning_rate": 2e-05,
+      "loss": 0.03601223,
+      "step": 19393
+    },
+    {
+      "epoch": 38.788,
+      "grad_norm": 1.0108002424240112,
+      "learning_rate": 2e-05,
+      "loss": 0.04009066,
+      "step": 19394
+    },
+    {
+      "epoch": 38.79,
+      "grad_norm": 1.7825038433074951,
+      "learning_rate": 2e-05,
+      "loss": 0.05417824,
+      "step": 19395
+    },
+    {
+      "epoch": 38.792,
+      "grad_norm": 1.7519702911376953,
+      "learning_rate": 2e-05,
+      "loss": 0.05254176,
+      "step": 19396
+    },
+    {
+      "epoch": 38.794,
+      "grad_norm": 1.3303686380386353,
+      "learning_rate": 2e-05,
+      "loss": 0.05503142,
+      "step": 19397
+    },
+    {
+      "epoch": 38.796,
+      "grad_norm": 1.049363136291504,
+      "learning_rate": 2e-05,
+      "loss": 0.0418441,
+      "step": 19398
+    },
+    {
+      "epoch": 38.798,
+      "grad_norm": 1.388549566268921,
+      "learning_rate": 2e-05,
+      "loss": 0.04791044,
+      "step": 19399
+    },
+    {
+      "epoch": 38.8,
+      "grad_norm": 1.4452887773513794,
+      "learning_rate": 2e-05,
+      "loss": 0.04734738,
+      "step": 19400
+    },
+    {
+      "epoch": 38.802,
+      "grad_norm": 1.2454729080200195,
+      "learning_rate": 2e-05,
+      "loss": 0.05771025,
+      "step": 19401
+    },
+    {
+      "epoch": 38.804,
+      "grad_norm": 1.1745954751968384,
+      "learning_rate": 2e-05,
+      "loss": 0.04313898,
+      "step": 19402
+    },
+    {
+      "epoch": 38.806,
+      "grad_norm": 1.3334699869155884,
+      "learning_rate": 2e-05,
+      "loss": 0.06793031,
+      "step": 19403
+    },
+    {
+      "epoch": 38.808,
+      "grad_norm": 1.42832350730896,
+      "learning_rate": 2e-05,
+      "loss": 0.03323733,
+      "step": 19404
+    },
+    {
+      "epoch": 38.81,
+      "grad_norm": 1.0687870979309082,
+      "learning_rate": 2e-05,
+      "loss": 0.0360961,
+      "step": 19405
+    },
+    {
+      "epoch": 38.812,
+      "grad_norm": 1.741623878479004,
+      "learning_rate": 2e-05,
+      "loss": 0.06625956,
+      "step": 19406
+    },
+    {
+      "epoch": 38.814,
+      "grad_norm": 1.6357783079147339,
+      "learning_rate": 2e-05,
+      "loss": 0.05045918,
+      "step": 19407
+    },
+    {
+      "epoch": 38.816,
+      "grad_norm": 1.15137779712677,
+      "learning_rate": 2e-05,
+      "loss": 0.03751816,
+      "step": 19408
+    },
+    {
+      "epoch": 38.818,
+      "grad_norm": 1.2414789199829102,
+      "learning_rate": 2e-05,
+      "loss": 0.04433636,
+      "step": 19409
+    },
+    {
+      "epoch": 38.82,
+      "grad_norm": 1.303000569343567,
+      "learning_rate": 2e-05,
+      "loss": 0.05153903,
+      "step": 19410
+    },
+    {
+      "epoch": 38.822,
+      "grad_norm": 2.2832419872283936,
+      "learning_rate": 2e-05,
+      "loss": 0.05765641,
+      "step": 19411
+    },
+    {
+      "epoch": 38.824,
+      "grad_norm": 1.1722428798675537,
+      "learning_rate": 2e-05,
+      "loss": 0.04241635,
+      "step": 19412
+    },
+    {
+      "epoch": 38.826,
+      "grad_norm": 1.613328218460083,
+      "learning_rate": 2e-05,
+      "loss": 0.03948274,
+      "step": 19413
+    },
+    {
+      "epoch": 38.828,
+      "grad_norm": 1.2444097995758057,
+      "learning_rate": 2e-05,
+      "loss": 0.04656627,
+      "step": 19414
+    },
+    {
+      "epoch": 38.83,
+      "grad_norm": 1.2532914876937866,
+      "learning_rate": 2e-05,
+      "loss": 0.0406023,
+      "step": 19415
+    },
+    {
+      "epoch": 38.832,
+      "grad_norm": 1.4562067985534668,
+      "learning_rate": 2e-05,
+      "loss": 0.06386394,
+      "step": 19416
+    },
+    {
+      "epoch": 38.834,
+      "grad_norm": 1.2982302904129028,
+      "learning_rate": 2e-05,
+      "loss": 0.04366529,
+      "step": 19417
+    },
+    {
+      "epoch": 38.836,
+      "grad_norm": 1.3933048248291016,
+      "learning_rate": 2e-05,
+      "loss": 0.05839304,
+      "step": 19418
+    },
+    {
+      "epoch": 38.838,
+      "grad_norm": 1.2442266941070557,
+      "learning_rate": 2e-05,
+      "loss": 0.04477119,
+      "step": 19419
+    },
+    {
+      "epoch": 38.84,
+      "grad_norm": 1.2520731687545776,
+      "learning_rate": 2e-05,
+      "loss": 0.05134234,
+      "step": 19420
+    },
+    {
+      "epoch": 38.842,
+      "grad_norm": 1.3517646789550781,
+      "learning_rate": 2e-05,
+      "loss": 0.04959063,
+      "step": 19421
+    },
+    {
+      "epoch": 38.844,
+      "grad_norm": 1.1629074811935425,
+      "learning_rate": 2e-05,
+      "loss": 0.04994313,
+      "step": 19422
+    },
+    {
+      "epoch": 38.846,
+      "grad_norm": 4.365392684936523,
+      "learning_rate": 2e-05,
+      "loss": 0.0617961,
+      "step": 19423
+    },
+    {
+      "epoch": 38.848,
+      "grad_norm": 2.191051959991455,
+      "learning_rate": 2e-05,
+      "loss": 0.04884791,
+      "step": 19424
+    },
+    {
+      "epoch": 38.85,
+      "grad_norm": 1.9126558303833008,
+      "learning_rate": 2e-05,
+      "loss": 0.04287531,
+      "step": 19425
+    },
+    {
+      "epoch": 38.852,
+      "grad_norm": 1.3821977376937866,
+      "learning_rate": 2e-05,
+      "loss": 0.0512047,
+      "step": 19426
+    },
+    {
+      "epoch": 38.854,
+      "grad_norm": 2.440401554107666,
+      "learning_rate": 2e-05,
+      "loss": 0.06993075,
+      "step": 19427
+    },
+    {
+      "epoch": 38.856,
+      "grad_norm": 1.3591678142547607,
+      "learning_rate": 2e-05,
+      "loss": 0.05418681,
+      "step": 19428
+    },
+    {
+      "epoch": 38.858,
+      "grad_norm": 1.1129034757614136,
+      "learning_rate": 2e-05,
+      "loss": 0.04099159,
+      "step": 19429
+    },
+    {
+      "epoch": 38.86,
+      "grad_norm": 0.9613949656486511,
+      "learning_rate": 2e-05,
+      "loss": 0.02941245,
+      "step": 19430
+    },
+    {
+      "epoch": 38.862,
+      "grad_norm": 0.8730504512786865,
+      "learning_rate": 2e-05,
+      "loss": 0.02738763,
+      "step": 19431
+    },
+    {
+      "epoch": 38.864,
+      "grad_norm": 1.091715931892395,
+      "learning_rate": 2e-05,
+      "loss": 0.03717642,
+      "step": 19432
+    },
+    {
+      "epoch": 38.866,
+      "grad_norm": 1.1144849061965942,
+      "learning_rate": 2e-05,
+      "loss": 0.04412191,
+      "step": 19433
+    },
+    {
+      "epoch": 38.868,
+      "grad_norm": 1.2194790840148926,
+      "learning_rate": 2e-05,
+      "loss": 0.04663887,
+      "step": 19434
+    },
+    {
+      "epoch": 38.87,
+      "grad_norm": 1.3206437826156616,
+      "learning_rate": 2e-05,
+      "loss": 0.05822754,
+      "step": 19435
+    },
+    {
+      "epoch": 38.872,
+      "grad_norm": 1.0756211280822754,
+      "learning_rate": 2e-05,
+      "loss": 0.03367351,
+      "step": 19436
+    },
+    {
+      "epoch": 38.874,
+      "grad_norm": 1.2603015899658203,
+      "learning_rate": 2e-05,
+      "loss": 0.04884558,
+      "step": 19437
+    },
+    {
+      "epoch": 38.876,
+      "grad_norm": 1.3238136768341064,
+      "learning_rate": 2e-05,
+      "loss": 0.0366739,
+      "step": 19438
+    },
+    {
+      "epoch": 38.878,
+      "grad_norm": 1.2530521154403687,
+      "learning_rate": 2e-05,
+      "loss": 0.05306315,
+      "step": 19439
+    },
+    {
+      "epoch": 38.88,
+      "grad_norm": 1.3359322547912598,
+      "learning_rate": 2e-05,
+      "loss": 0.04838687,
+      "step": 19440
+    },
+    {
+      "epoch": 38.882,
+      "grad_norm": 1.201737880706787,
+      "learning_rate": 2e-05,
+      "loss": 0.05184001,
+      "step": 19441
+    },
+    {
+      "epoch": 38.884,
+      "grad_norm": 1.2076383829116821,
+      "learning_rate": 2e-05,
+      "loss": 0.04534627,
+      "step": 19442
+    },
+    {
+      "epoch": 38.886,
+      "grad_norm": 0.982481062412262,
+      "learning_rate": 2e-05,
+      "loss": 0.03846693,
+      "step": 19443
+    },
+    {
+      "epoch": 38.888,
+      "grad_norm": 1.109921932220459,
+      "learning_rate": 2e-05,
+      "loss": 0.04263131,
+      "step": 19444
+    },
+    {
+      "epoch": 38.89,
+      "grad_norm": 1.115616798400879,
+      "learning_rate": 2e-05,
+      "loss": 0.05047696,
+      "step": 19445
+    },
+    {
+      "epoch": 38.892,
+      "grad_norm": 1.1365337371826172,
+      "learning_rate": 2e-05,
+      "loss": 0.0466363,
+      "step": 19446
+    },
+    {
+      "epoch": 38.894,
+      "grad_norm": 1.30055832862854,
+      "learning_rate": 2e-05,
+      "loss": 0.06107081,
+      "step": 19447
+    },
+    {
+      "epoch": 38.896,
+      "grad_norm": 1.2148528099060059,
+      "learning_rate": 2e-05,
+      "loss": 0.04618006,
+      "step": 19448
+    },
+    {
+      "epoch": 38.898,
+      "grad_norm": 1.095479130744934,
+      "learning_rate": 2e-05,
+      "loss": 0.04486157,
+      "step": 19449
+    },
+    {
+      "epoch": 38.9,
+      "grad_norm": 1.2125225067138672,
+      "learning_rate": 2e-05,
+      "loss": 0.04702915,
+      "step": 19450
+    },
+    {
+      "epoch": 38.902,
+      "grad_norm": 1.1719056367874146,
+      "learning_rate": 2e-05,
+      "loss": 0.05031497,
+      "step": 19451
+    },
+    {
+      "epoch": 38.904,
+      "grad_norm": 1.9016114473342896,
+      "learning_rate": 2e-05,
+      "loss": 0.06579206,
+      "step": 19452
+    },
+    {
+      "epoch": 38.906,
+      "grad_norm": 2.084883213043213,
+      "learning_rate": 2e-05,
+      "loss": 0.04503134,
+      "step": 19453
+    },
+    {
+      "epoch": 38.908,
+      "grad_norm": 3.5525739192962646,
+      "learning_rate": 2e-05,
+      "loss": 0.06399915,
+      "step": 19454
+    },
+    {
+      "epoch": 38.91,
+      "grad_norm": 1.238961935043335,
+      "learning_rate": 2e-05,
+      "loss": 0.05072328,
+      "step": 19455
+    },
+    {
+      "epoch": 38.912,
+      "grad_norm": 1.1139464378356934,
+      "learning_rate": 2e-05,
+      "loss": 0.05063874,
+      "step": 19456
+    },
+    {
+      "epoch": 38.914,
+      "grad_norm": 1.591186285018921,
+      "learning_rate": 2e-05,
+      "loss": 0.04680803,
+      "step": 19457
+    },
+    {
+      "epoch": 38.916,
+      "grad_norm": 1.3666948080062866,
+      "learning_rate": 2e-05,
+      "loss": 0.05156454,
+      "step": 19458
+    },
+    {
+      "epoch": 38.918,
+      "grad_norm": 1.5260530710220337,
+      "learning_rate": 2e-05,
+      "loss": 0.07067595,
+      "step": 19459
+    },
+    {
+      "epoch": 38.92,
+      "grad_norm": 1.2484623193740845,
+      "learning_rate": 2e-05,
+      "loss": 0.03980666,
+      "step": 19460
+    },
+    {
+      "epoch": 38.922,
+      "grad_norm": 1.2105351686477661,
+      "learning_rate": 2e-05,
+      "loss": 0.05022258,
+      "step": 19461
+    },
+    {
+      "epoch": 38.924,
+      "grad_norm": 0.9624425172805786,
+      "learning_rate": 2e-05,
+      "loss": 0.02262569,
+      "step": 19462
+    },
+    {
+      "epoch": 38.926,
+      "grad_norm": 1.1966172456741333,
+      "learning_rate": 2e-05,
+      "loss": 0.04677077,
+      "step": 19463
+    },
+    {
+      "epoch": 38.928,
+      "grad_norm": 2.7761287689208984,
+      "learning_rate": 2e-05,
+      "loss": 0.04095355,
+      "step": 19464
+    },
+    {
+      "epoch": 38.93,
+      "grad_norm": 1.3439223766326904,
+      "learning_rate": 2e-05,
+      "loss": 0.05076102,
+      "step": 19465
+    },
+    {
+      "epoch": 38.932,
+      "grad_norm": 1.281233310699463,
+      "learning_rate": 2e-05,
+      "loss": 0.05176672,
+      "step": 19466
+    },
+    {
+      "epoch": 38.934,
+      "grad_norm": 1.2068181037902832,
+      "learning_rate": 2e-05,
+      "loss": 0.03931053,
+      "step": 19467
+    },
+    {
+      "epoch": 38.936,
+      "grad_norm": 2.3821423053741455,
+      "learning_rate": 2e-05,
+      "loss": 0.05391637,
+      "step": 19468
+    },
+    {
+      "epoch": 38.938,
+      "grad_norm": 0.8565077781677246,
+      "learning_rate": 2e-05,
+      "loss": 0.02748447,
+      "step": 19469
+    },
+    {
+      "epoch": 38.94,
+      "grad_norm": 1.593507170677185,
+      "learning_rate": 2e-05,
+      "loss": 0.05532858,
+      "step": 19470
+    },
+    {
+      "epoch": 38.942,
+      "grad_norm": 1.2847896814346313,
+      "learning_rate": 2e-05,
+      "loss": 0.052665,
+      "step": 19471
+    },
+    {
+      "epoch": 38.944,
+      "grad_norm": 3.365082263946533,
+      "learning_rate": 2e-05,
+      "loss": 0.04390424,
+      "step": 19472
+    },
+    {
+      "epoch": 38.946,
+      "grad_norm": 1.228043556213379,
+      "learning_rate": 2e-05,
+      "loss": 0.03701095,
+      "step": 19473
+    },
+    {
+      "epoch": 38.948,
+      "grad_norm": 1.2872579097747803,
+      "learning_rate": 2e-05,
+      "loss": 0.06041116,
+      "step": 19474
+    },
+    {
+      "epoch": 38.95,
+      "grad_norm": 5.743537902832031,
+      "learning_rate": 2e-05,
+      "loss": 0.05942594,
+      "step": 19475
+    },
+    {
+      "epoch": 38.952,
+      "grad_norm": 1.4819161891937256,
+      "learning_rate": 2e-05,
+      "loss": 0.05315714,
+      "step": 19476
+    },
+    {
+      "epoch": 38.954,
+      "grad_norm": 4.263488292694092,
+      "learning_rate": 2e-05,
+      "loss": 0.05684969,
+      "step": 19477
+    },
+    {
+      "epoch": 38.956,
+      "grad_norm": 1.293532133102417,
+      "learning_rate": 2e-05,
+      "loss": 0.05474854,
+      "step": 19478
+    },
+    {
+      "epoch": 38.958,
+      "grad_norm": 1.5321731567382812,
+      "learning_rate": 2e-05,
+      "loss": 0.0521361,
+      "step": 19479
+    },
+    {
+      "epoch": 38.96,
+      "grad_norm": 1.5575649738311768,
+      "learning_rate": 2e-05,
+      "loss": 0.0335691,
+      "step": 19480
+    },
+    {
+      "epoch": 38.962,
+      "grad_norm": 1.2537888288497925,
+      "learning_rate": 2e-05,
+      "loss": 0.04636952,
+      "step": 19481
+    },
+    {
+      "epoch": 38.964,
+      "grad_norm": 1.7497940063476562,
+      "learning_rate": 2e-05,
+      "loss": 0.06122085,
+      "step": 19482
+    },
+    {
+      "epoch": 38.966,
+      "grad_norm": 0.9865143299102783,
+      "learning_rate": 2e-05,
+      "loss": 0.03592449,
+      "step": 19483
+    },
+    {
+      "epoch": 38.968,
+      "grad_norm": 1.0101889371871948,
+      "learning_rate": 2e-05,
+      "loss": 0.04233185,
+      "step": 19484
+    },
+    {
+      "epoch": 38.97,
+      "grad_norm": 1.1617642641067505,
+      "learning_rate": 2e-05,
+      "loss": 0.04329985,
+      "step": 19485
+    },
+    {
+      "epoch": 38.972,
+      "grad_norm": 1.1280710697174072,
+      "learning_rate": 2e-05,
+      "loss": 0.03594364,
+      "step": 19486
+    },
+    {
+      "epoch": 38.974,
+      "grad_norm": 1.4211704730987549,
+      "learning_rate": 2e-05,
+      "loss": 0.06367813,
+      "step": 19487
+    },
+    {
+      "epoch": 38.976,
+      "grad_norm": 1.0155389308929443,
+      "learning_rate": 2e-05,
+      "loss": 0.04221959,
+      "step": 19488
+    },
+    {
+      "epoch": 38.978,
+      "grad_norm": 0.9778318405151367,
+      "learning_rate": 2e-05,
+      "loss": 0.0305156,
+      "step": 19489
+    },
+    {
+      "epoch": 38.98,
+      "grad_norm": 1.2039105892181396,
+      "learning_rate": 2e-05,
+      "loss": 0.04294903,
+      "step": 19490
+    },
+    {
+      "epoch": 38.982,
+      "grad_norm": 0.9735546112060547,
+      "learning_rate": 2e-05,
+      "loss": 0.03935603,
+      "step": 19491
+    },
+    {
+      "epoch": 38.984,
+      "grad_norm": 2.8653101921081543,
+      "learning_rate": 2e-05,
+      "loss": 0.05370308,
+      "step": 19492
+    },
+    {
+      "epoch": 38.986,
+      "grad_norm": 1.1377127170562744,
+      "learning_rate": 2e-05,
+      "loss": 0.04588469,
+      "step": 19493
+    },
+    {
+      "epoch": 38.988,
+      "grad_norm": 1.1004186868667603,
+      "learning_rate": 2e-05,
+      "loss": 0.05202518,
+      "step": 19494
+    },
+    {
+      "epoch": 38.99,
+      "grad_norm": 1.7122163772583008,
+      "learning_rate": 2e-05,
+      "loss": 0.03624396,
+      "step": 19495
+    },
+    {
+      "epoch": 38.992,
+      "grad_norm": 2.000995635986328,
+      "learning_rate": 2e-05,
+      "loss": 0.04281428,
+      "step": 19496
+    },
+    {
+      "epoch": 38.994,
+      "grad_norm": 1.3111525774002075,
+      "learning_rate": 2e-05,
+      "loss": 0.05325833,
+      "step": 19497
+    },
+    {
+      "epoch": 38.996,
+      "grad_norm": 1.5315814018249512,
+      "learning_rate": 2e-05,
+      "loss": 0.06504035,
+      "step": 19498
+    },
+    {
+      "epoch": 38.998,
+      "grad_norm": 1.1684447526931763,
+      "learning_rate": 2e-05,
+      "loss": 0.04492722,
+      "step": 19499
+    },
+    {
+      "epoch": 39.0,
+      "grad_norm": 1.3859971761703491,
+      "learning_rate": 2e-05,
+      "loss": 0.06541097,
+      "step": 19500
+    },
+    {
+      "epoch": 39.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.998,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9920159680638723,
+        "Equal_1": 1.0,
+        "Equal_2": 0.9860279441117764,
+        "Equal_3": 0.9900199600798403,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 1.0,
+        "Parallel_1": 0.9879759519038076,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 0.996,
+        "Perpendicular_2": 0.992,
+        "Perpendicular_3": 0.8847695390781564,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 0.9976666666666667,
+        "PointLiesOnCircle_3": 0.9972666666666666,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9840319361277445
+      },
+      "eval_runtime": 320.8325,
+      "eval_samples_per_second": 32.727,
+      "eval_steps_per_second": 0.655,
+      "step": 19500
+    },
+    {
+      "epoch": 39.002,
+      "grad_norm": 6.220664024353027,
+      "learning_rate": 2e-05,
+      "loss": 0.04667213,
+      "step": 19501
+    },
+    {
+      "epoch": 39.004,
+      "grad_norm": 1.135877013206482,
+      "learning_rate": 2e-05,
+      "loss": 0.03528913,
+      "step": 19502
+    },
+    {
+      "epoch": 39.006,
+      "grad_norm": 1.918533205986023,
+      "learning_rate": 2e-05,
+      "loss": 0.05770627,
+      "step": 19503
+    },
+    {
+      "epoch": 39.008,
+      "grad_norm": 1.3888102769851685,
+      "learning_rate": 2e-05,
+      "loss": 0.05147287,
+      "step": 19504
+    },
+    {
+      "epoch": 39.01,
+      "grad_norm": 1.2796348333358765,
+      "learning_rate": 2e-05,
+      "loss": 0.04314057,
+      "step": 19505
+    },
+    {
+      "epoch": 39.012,
+      "grad_norm": 1.2352646589279175,
+      "learning_rate": 2e-05,
+      "loss": 0.04238914,
+      "step": 19506
+    },
+    {
+      "epoch": 39.014,
+      "grad_norm": 1.5327287912368774,
+      "learning_rate": 2e-05,
+      "loss": 0.06135549,
+      "step": 19507
+    },
+    {
+      "epoch": 39.016,
+      "grad_norm": 1.9634387493133545,
+      "learning_rate": 2e-05,
+      "loss": 0.04630491,
+      "step": 19508
+    },
+    {
+      "epoch": 39.018,
+      "grad_norm": 1.3204143047332764,
+      "learning_rate": 2e-05,
+      "loss": 0.04790439,
+      "step": 19509
+    },
+    {
+      "epoch": 39.02,
+      "grad_norm": 1.5740340948104858,
+      "learning_rate": 2e-05,
+      "loss": 0.04070012,
+      "step": 19510
+    },
+    {
+      "epoch": 39.022,
+      "grad_norm": 1.3499553203582764,
+      "learning_rate": 2e-05,
+      "loss": 0.05047411,
+      "step": 19511
+    },
+    {
+      "epoch": 39.024,
+      "grad_norm": 0.9350709915161133,
+      "learning_rate": 2e-05,
+      "loss": 0.02976116,
+      "step": 19512
+    },
+    {
+      "epoch": 39.026,
+      "grad_norm": 1.0683033466339111,
+      "learning_rate": 2e-05,
+      "loss": 0.03930287,
+      "step": 19513
+    },
+    {
+      "epoch": 39.028,
+      "grad_norm": 1.1646449565887451,
+      "learning_rate": 2e-05,
+      "loss": 0.04475568,
+      "step": 19514
+    },
+    {
+      "epoch": 39.03,
+      "grad_norm": 1.3507591485977173,
+      "learning_rate": 2e-05,
+      "loss": 0.05256594,
+      "step": 19515
+    },
+    {
+      "epoch": 39.032,
+      "grad_norm": 1.2592138051986694,
+      "learning_rate": 2e-05,
+      "loss": 0.04029159,
+      "step": 19516
+    },
+    {
+      "epoch": 39.034,
+      "grad_norm": 1.0979448556900024,
+      "learning_rate": 2e-05,
+      "loss": 0.05017549,
+      "step": 19517
+    },
+    {
+      "epoch": 39.036,
+      "grad_norm": 1.0506963729858398,
+      "learning_rate": 2e-05,
+      "loss": 0.04544584,
+      "step": 19518
+    },
+    {
+      "epoch": 39.038,
+      "grad_norm": 1.2208514213562012,
+      "learning_rate": 2e-05,
+      "loss": 0.04804792,
+      "step": 19519
+    },
+    {
+      "epoch": 39.04,
+      "grad_norm": 1.086912989616394,
+      "learning_rate": 2e-05,
+      "loss": 0.04290297,
+      "step": 19520
+    },
+    {
+      "epoch": 39.042,
+      "grad_norm": 2.351929187774658,
+      "learning_rate": 2e-05,
+      "loss": 0.0460132,
+      "step": 19521
+    },
+    {
+      "epoch": 39.044,
+      "grad_norm": 1.1975494623184204,
+      "learning_rate": 2e-05,
+      "loss": 0.04094232,
+      "step": 19522
+    },
+    {
+      "epoch": 39.046,
+      "grad_norm": 1.4654606580734253,
+      "learning_rate": 2e-05,
+      "loss": 0.0614688,
+      "step": 19523
+    },
+    {
+      "epoch": 39.048,
+      "grad_norm": 1.3602313995361328,
+      "learning_rate": 2e-05,
+      "loss": 0.05025218,
+      "step": 19524
+    },
+    {
+      "epoch": 39.05,
+      "grad_norm": 1.1130294799804688,
+      "learning_rate": 2e-05,
+      "loss": 0.03360223,
+      "step": 19525
+    },
+    {
+      "epoch": 39.052,
+      "grad_norm": 1.3653548955917358,
+      "learning_rate": 2e-05,
+      "loss": 0.0470277,
+      "step": 19526
+    },
+    {
+      "epoch": 39.054,
+      "grad_norm": 1.0190237760543823,
+      "learning_rate": 2e-05,
+      "loss": 0.03658837,
+      "step": 19527
+    },
+    {
+      "epoch": 39.056,
+      "grad_norm": 1.2680890560150146,
+      "learning_rate": 2e-05,
+      "loss": 0.04371505,
+      "step": 19528
+    },
+    {
+      "epoch": 39.058,
+      "grad_norm": 1.6253620386123657,
+      "learning_rate": 2e-05,
+      "loss": 0.05293868,
+      "step": 19529
+    },
+    {
+      "epoch": 39.06,
+      "grad_norm": 1.2560430765151978,
+      "learning_rate": 2e-05,
+      "loss": 0.03810944,
+      "step": 19530
+    },
+    {
+      "epoch": 39.062,
+      "grad_norm": 1.2208555936813354,
+      "learning_rate": 2e-05,
+      "loss": 0.04190925,
+      "step": 19531
+    },
+    {
+      "epoch": 39.064,
+      "grad_norm": 1.098758578300476,
+      "learning_rate": 2e-05,
+      "loss": 0.03760337,
+      "step": 19532
+    },
+    {
+      "epoch": 39.066,
+      "grad_norm": 1.0823376178741455,
+      "learning_rate": 2e-05,
+      "loss": 0.040763,
+      "step": 19533
+    },
+    {
+      "epoch": 39.068,
+      "grad_norm": 1.1920530796051025,
+      "learning_rate": 2e-05,
+      "loss": 0.03933562,
+      "step": 19534
+    },
+    {
+      "epoch": 39.07,
+      "grad_norm": 1.3417820930480957,
+      "learning_rate": 2e-05,
+      "loss": 0.07266225,
+      "step": 19535
+    },
+    {
+      "epoch": 39.072,
+      "grad_norm": 1.156474232673645,
+      "learning_rate": 2e-05,
+      "loss": 0.03301832,
+      "step": 19536
+    },
+    {
+      "epoch": 39.074,
+      "grad_norm": 1.7914104461669922,
+      "learning_rate": 2e-05,
+      "loss": 0.05243705,
+      "step": 19537
+    },
+    {
+      "epoch": 39.076,
+      "grad_norm": 1.2239590883255005,
+      "learning_rate": 2e-05,
+      "loss": 0.05815011,
+      "step": 19538
+    },
+    {
+      "epoch": 39.078,
+      "grad_norm": 1.9638248682022095,
+      "learning_rate": 2e-05,
+      "loss": 0.058245,
+      "step": 19539
+    },
+    {
+      "epoch": 39.08,
+      "grad_norm": 1.095078468322754,
+      "learning_rate": 2e-05,
+      "loss": 0.041117,
+      "step": 19540
+    },
+    {
+      "epoch": 39.082,
+      "grad_norm": 1.0802770853042603,
+      "learning_rate": 2e-05,
+      "loss": 0.04389591,
+      "step": 19541
+    },
+    {
+      "epoch": 39.084,
+      "grad_norm": 2.124570369720459,
+      "learning_rate": 2e-05,
+      "loss": 0.05145385,
+      "step": 19542
+    },
+    {
+      "epoch": 39.086,
+      "grad_norm": 1.1902073621749878,
+      "learning_rate": 2e-05,
+      "loss": 0.03708824,
+      "step": 19543
+    },
+    {
+      "epoch": 39.088,
+      "grad_norm": 1.1799651384353638,
+      "learning_rate": 2e-05,
+      "loss": 0.03826425,
+      "step": 19544
+    },
+    {
+      "epoch": 39.09,
+      "grad_norm": 2.205336570739746,
+      "learning_rate": 2e-05,
+      "loss": 0.0574959,
+      "step": 19545
+    },
+    {
+      "epoch": 39.092,
+      "grad_norm": 1.5711607933044434,
+      "learning_rate": 2e-05,
+      "loss": 0.04823216,
+      "step": 19546
+    },
+    {
+      "epoch": 39.094,
+      "grad_norm": 1.6747071743011475,
+      "learning_rate": 2e-05,
+      "loss": 0.063738,
+      "step": 19547
+    },
+    {
+      "epoch": 39.096,
+      "grad_norm": 1.288379430770874,
+      "learning_rate": 2e-05,
+      "loss": 0.05128692,
+      "step": 19548
+    },
+    {
+      "epoch": 39.098,
+      "grad_norm": 1.3630702495574951,
+      "learning_rate": 2e-05,
+      "loss": 0.04941294,
+      "step": 19549
+    },
+    {
+      "epoch": 39.1,
+      "grad_norm": 2.2876434326171875,
+      "learning_rate": 2e-05,
+      "loss": 0.05177325,
+      "step": 19550
+    },
+    {
+      "epoch": 39.102,
+      "grad_norm": 1.884543538093567,
+      "learning_rate": 2e-05,
+      "loss": 0.0452019,
+      "step": 19551
+    },
+    {
+      "epoch": 39.104,
+      "grad_norm": 2.4394567012786865,
+      "learning_rate": 2e-05,
+      "loss": 0.05596255,
+      "step": 19552
+    },
+    {
+      "epoch": 39.106,
+      "grad_norm": 1.3302148580551147,
+      "learning_rate": 2e-05,
+      "loss": 0.04958398,
+      "step": 19553
+    },
+    {
+      "epoch": 39.108,
+      "grad_norm": 0.9878301620483398,
+      "learning_rate": 2e-05,
+      "loss": 0.03418873,
+      "step": 19554
+    },
+    {
+      "epoch": 39.11,
+      "grad_norm": 1.4796167612075806,
+      "learning_rate": 2e-05,
+      "loss": 0.06076609,
+      "step": 19555
+    },
+    {
+      "epoch": 39.112,
+      "grad_norm": 0.9968535304069519,
+      "learning_rate": 2e-05,
+      "loss": 0.0371155,
+      "step": 19556
+    },
+    {
+      "epoch": 39.114,
+      "grad_norm": 1.0637849569320679,
+      "learning_rate": 2e-05,
+      "loss": 0.03895561,
+      "step": 19557
+    },
+    {
+      "epoch": 39.116,
+      "grad_norm": 1.259835958480835,
+      "learning_rate": 2e-05,
+      "loss": 0.05485725,
+      "step": 19558
+    },
+    {
+      "epoch": 39.118,
+      "grad_norm": 1.407067894935608,
+      "learning_rate": 2e-05,
+      "loss": 0.03801053,
+      "step": 19559
+    },
+    {
+      "epoch": 39.12,
+      "grad_norm": 1.506779432296753,
+      "learning_rate": 2e-05,
+      "loss": 0.0610331,
+      "step": 19560
+    },
+    {
+      "epoch": 39.122,
+      "grad_norm": 1.299699306488037,
+      "learning_rate": 2e-05,
+      "loss": 0.06505786,
+      "step": 19561
+    },
+    {
+      "epoch": 39.124,
+      "grad_norm": 1.213064193725586,
+      "learning_rate": 2e-05,
+      "loss": 0.04449017,
+      "step": 19562
+    },
+    {
+      "epoch": 39.126,
+      "grad_norm": 1.0915803909301758,
+      "learning_rate": 2e-05,
+      "loss": 0.04768021,
+      "step": 19563
+    },
+    {
+      "epoch": 39.128,
+      "grad_norm": 1.2898131608963013,
+      "learning_rate": 2e-05,
+      "loss": 0.06006561,
+      "step": 19564
+    },
+    {
+      "epoch": 39.13,
+      "grad_norm": 1.134291410446167,
+      "learning_rate": 2e-05,
+      "loss": 0.04017518,
+      "step": 19565
+    },
+    {
+      "epoch": 39.132,
+      "grad_norm": 1.839423656463623,
+      "learning_rate": 2e-05,
+      "loss": 0.05884816,
+      "step": 19566
+    },
+    {
+      "epoch": 39.134,
+      "grad_norm": 3.196049451828003,
+      "learning_rate": 2e-05,
+      "loss": 0.0544239,
+      "step": 19567
+    },
+    {
+      "epoch": 39.136,
+      "grad_norm": 1.3496533632278442,
+      "learning_rate": 2e-05,
+      "loss": 0.06023031,
+      "step": 19568
+    },
+    {
+      "epoch": 39.138,
+      "grad_norm": 1.06864595413208,
+      "learning_rate": 2e-05,
+      "loss": 0.04296235,
+      "step": 19569
+    },
+    {
+      "epoch": 39.14,
+      "grad_norm": 0.9733324646949768,
+      "learning_rate": 2e-05,
+      "loss": 0.03551897,
+      "step": 19570
+    },
+    {
+      "epoch": 39.142,
+      "grad_norm": 1.19362211227417,
+      "learning_rate": 2e-05,
+      "loss": 0.05036366,
+      "step": 19571
+    },
+    {
+      "epoch": 39.144,
+      "grad_norm": 1.0025893449783325,
+      "learning_rate": 2e-05,
+      "loss": 0.0287686,
+      "step": 19572
+    },
+    {
+      "epoch": 39.146,
+      "grad_norm": 2.9667563438415527,
+      "learning_rate": 2e-05,
+      "loss": 0.0478669,
+      "step": 19573
+    },
+    {
+      "epoch": 39.148,
+      "grad_norm": 1.4983551502227783,
+      "learning_rate": 2e-05,
+      "loss": 0.05278377,
+      "step": 19574
+    },
+    {
+      "epoch": 39.15,
+      "grad_norm": 1.3283973932266235,
+      "learning_rate": 2e-05,
+      "loss": 0.05365999,
+      "step": 19575
+    },
+    {
+      "epoch": 39.152,
+      "grad_norm": 1.4052914381027222,
+      "learning_rate": 2e-05,
+      "loss": 0.04093612,
+      "step": 19576
+    },
+    {
+      "epoch": 39.154,
+      "grad_norm": 1.2109755277633667,
+      "learning_rate": 2e-05,
+      "loss": 0.04671524,
+      "step": 19577
+    },
+    {
+      "epoch": 39.156,
+      "grad_norm": 1.4497380256652832,
+      "learning_rate": 2e-05,
+      "loss": 0.05801364,
+      "step": 19578
+    },
+    {
+      "epoch": 39.158,
+      "grad_norm": 1.0449851751327515,
+      "learning_rate": 2e-05,
+      "loss": 0.03876752,
+      "step": 19579
+    },
+    {
+      "epoch": 39.16,
+      "grad_norm": 0.9883303642272949,
+      "learning_rate": 2e-05,
+      "loss": 0.02868701,
+      "step": 19580
+    },
+    {
+      "epoch": 39.162,
+      "grad_norm": 1.0830588340759277,
+      "learning_rate": 2e-05,
+      "loss": 0.03940789,
+      "step": 19581
+    },
+    {
+      "epoch": 39.164,
+      "grad_norm": 1.4366869926452637,
+      "learning_rate": 2e-05,
+      "loss": 0.07369938,
+      "step": 19582
+    },
+    {
+      "epoch": 39.166,
+      "grad_norm": 1.415928840637207,
+      "learning_rate": 2e-05,
+      "loss": 0.04259369,
+      "step": 19583
+    },
+    {
+      "epoch": 39.168,
+      "grad_norm": 1.3500007390975952,
+      "learning_rate": 2e-05,
+      "loss": 0.05315146,
+      "step": 19584
+    },
+    {
+      "epoch": 39.17,
+      "grad_norm": 1.1341204643249512,
+      "learning_rate": 2e-05,
+      "loss": 0.03991456,
+      "step": 19585
+    },
+    {
+      "epoch": 39.172,
+      "grad_norm": 1.41619074344635,
+      "learning_rate": 2e-05,
+      "loss": 0.05198916,
+      "step": 19586
+    },
+    {
+      "epoch": 39.174,
+      "grad_norm": 1.457855224609375,
+      "learning_rate": 2e-05,
+      "loss": 0.06273693,
+      "step": 19587
+    },
+    {
+      "epoch": 39.176,
+      "grad_norm": 1.1896886825561523,
+      "learning_rate": 2e-05,
+      "loss": 0.05241989,
+      "step": 19588
+    },
+    {
+      "epoch": 39.178,
+      "grad_norm": 1.280698299407959,
+      "learning_rate": 2e-05,
+      "loss": 0.04846141,
+      "step": 19589
+    },
+    {
+      "epoch": 39.18,
+      "grad_norm": 1.0273659229278564,
+      "learning_rate": 2e-05,
+      "loss": 0.03391914,
+      "step": 19590
+    },
+    {
+      "epoch": 39.182,
+      "grad_norm": 1.2925435304641724,
+      "learning_rate": 2e-05,
+      "loss": 0.05286665,
+      "step": 19591
+    },
+    {
+      "epoch": 39.184,
+      "grad_norm": 0.9915168881416321,
+      "learning_rate": 2e-05,
+      "loss": 0.04317682,
+      "step": 19592
+    },
+    {
+      "epoch": 39.186,
+      "grad_norm": 1.0538667440414429,
+      "learning_rate": 2e-05,
+      "loss": 0.02936137,
+      "step": 19593
+    },
+    {
+      "epoch": 39.188,
+      "grad_norm": 1.22129225730896,
+      "learning_rate": 2e-05,
+      "loss": 0.0674229,
+      "step": 19594
+    },
+    {
+      "epoch": 39.19,
+      "grad_norm": 1.3039966821670532,
+      "learning_rate": 2e-05,
+      "loss": 0.04973481,
+      "step": 19595
+    },
+    {
+      "epoch": 39.192,
+      "grad_norm": 1.0738667249679565,
+      "learning_rate": 2e-05,
+      "loss": 0.03898337,
+      "step": 19596
+    },
+    {
+      "epoch": 39.194,
+      "grad_norm": 1.290103793144226,
+      "learning_rate": 2e-05,
+      "loss": 0.04972884,
+      "step": 19597
+    },
+    {
+      "epoch": 39.196,
+      "grad_norm": 1.3575490713119507,
+      "learning_rate": 2e-05,
+      "loss": 0.0455936,
+      "step": 19598
+    },
+    {
+      "epoch": 39.198,
+      "grad_norm": 1.1250419616699219,
+      "learning_rate": 2e-05,
+      "loss": 0.03418372,
+      "step": 19599
+    },
+    {
+      "epoch": 39.2,
+      "grad_norm": 1.4324772357940674,
+      "learning_rate": 2e-05,
+      "loss": 0.04676975,
+      "step": 19600
+    },
+    {
+      "epoch": 39.202,
+      "grad_norm": 1.14113187789917,
+      "learning_rate": 2e-05,
+      "loss": 0.036985,
+      "step": 19601
+    },
+    {
+      "epoch": 39.204,
+      "grad_norm": 1.313949704170227,
+      "learning_rate": 2e-05,
+      "loss": 0.05859777,
+      "step": 19602
+    },
+    {
+      "epoch": 39.206,
+      "grad_norm": 1.142185926437378,
+      "learning_rate": 2e-05,
+      "loss": 0.04276698,
+      "step": 19603
+    },
+    {
+      "epoch": 39.208,
+      "grad_norm": 1.3142091035842896,
+      "learning_rate": 2e-05,
+      "loss": 0.05003233,
+      "step": 19604
+    },
+    {
+      "epoch": 39.21,
+      "grad_norm": 1.3124581575393677,
+      "learning_rate": 2e-05,
+      "loss": 0.06136658,
+      "step": 19605
+    },
+    {
+      "epoch": 39.212,
+      "grad_norm": 1.2220021486282349,
+      "learning_rate": 2e-05,
+      "loss": 0.04311784,
+      "step": 19606
+    },
+    {
+      "epoch": 39.214,
+      "grad_norm": 1.3002612590789795,
+      "learning_rate": 2e-05,
+      "loss": 0.0366911,
+      "step": 19607
+    },
+    {
+      "epoch": 39.216,
+      "grad_norm": 1.4152746200561523,
+      "learning_rate": 2e-05,
+      "loss": 0.0430445,
+      "step": 19608
+    },
+    {
+      "epoch": 39.218,
+      "grad_norm": 1.1827176809310913,
+      "learning_rate": 2e-05,
+      "loss": 0.04736789,
+      "step": 19609
+    },
+    {
+      "epoch": 39.22,
+      "grad_norm": 1.1137902736663818,
+      "learning_rate": 2e-05,
+      "loss": 0.04015935,
+      "step": 19610
+    },
+    {
+      "epoch": 39.222,
+      "grad_norm": 1.7880975008010864,
+      "learning_rate": 2e-05,
+      "loss": 0.04855542,
+      "step": 19611
+    },
+    {
+      "epoch": 39.224,
+      "grad_norm": 1.182569146156311,
+      "learning_rate": 2e-05,
+      "loss": 0.05796768,
+      "step": 19612
+    },
+    {
+      "epoch": 39.226,
+      "grad_norm": 2.2326362133026123,
+      "learning_rate": 2e-05,
+      "loss": 0.04780351,
+      "step": 19613
+    },
+    {
+      "epoch": 39.228,
+      "grad_norm": 1.2634199857711792,
+      "learning_rate": 2e-05,
+      "loss": 0.04764343,
+      "step": 19614
+    },
+    {
+      "epoch": 39.23,
+      "grad_norm": 1.3403682708740234,
+      "learning_rate": 2e-05,
+      "loss": 0.04632024,
+      "step": 19615
+    },
+    {
+      "epoch": 39.232,
+      "grad_norm": 1.009716510772705,
+      "learning_rate": 2e-05,
+      "loss": 0.03490862,
+      "step": 19616
+    },
+    {
+      "epoch": 39.234,
+      "grad_norm": 1.03059983253479,
+      "learning_rate": 2e-05,
+      "loss": 0.03645796,
+      "step": 19617
+    },
+    {
+      "epoch": 39.236,
+      "grad_norm": 1.0769610404968262,
+      "learning_rate": 2e-05,
+      "loss": 0.03788844,
+      "step": 19618
+    },
+    {
+      "epoch": 39.238,
+      "grad_norm": 1.1544963121414185,
+      "learning_rate": 2e-05,
+      "loss": 0.04667365,
+      "step": 19619
+    },
+    {
+      "epoch": 39.24,
+      "grad_norm": 1.1700681447982788,
+      "learning_rate": 2e-05,
+      "loss": 0.04401879,
+      "step": 19620
+    },
+    {
+      "epoch": 39.242,
+      "grad_norm": 1.196091890335083,
+      "learning_rate": 2e-05,
+      "loss": 0.05693919,
+      "step": 19621
+    },
+    {
+      "epoch": 39.244,
+      "grad_norm": 2.257040023803711,
+      "learning_rate": 2e-05,
+      "loss": 0.03902316,
+      "step": 19622
+    },
+    {
+      "epoch": 39.246,
+      "grad_norm": 1.00798499584198,
+      "learning_rate": 2e-05,
+      "loss": 0.03313605,
+      "step": 19623
+    },
+    {
+      "epoch": 39.248,
+      "grad_norm": 1.4052647352218628,
+      "learning_rate": 2e-05,
+      "loss": 0.05165982,
+      "step": 19624
+    },
+    {
+      "epoch": 39.25,
+      "grad_norm": 1.2141923904418945,
+      "learning_rate": 2e-05,
+      "loss": 0.03951096,
+      "step": 19625
+    },
+    {
+      "epoch": 39.252,
+      "grad_norm": 1.3865104913711548,
+      "learning_rate": 2e-05,
+      "loss": 0.04825272,
+      "step": 19626
+    },
+    {
+      "epoch": 39.254,
+      "grad_norm": 1.102805733680725,
+      "learning_rate": 2e-05,
+      "loss": 0.03778962,
+      "step": 19627
+    },
+    {
+      "epoch": 39.256,
+      "grad_norm": 0.9773703813552856,
+      "learning_rate": 2e-05,
+      "loss": 0.03974072,
+      "step": 19628
+    },
+    {
+      "epoch": 39.258,
+      "grad_norm": 1.361572027206421,
+      "learning_rate": 2e-05,
+      "loss": 0.03857236,
+      "step": 19629
+    },
+    {
+      "epoch": 39.26,
+      "grad_norm": 2.139104127883911,
+      "learning_rate": 2e-05,
+      "loss": 0.05093241,
+      "step": 19630
+    },
+    {
+      "epoch": 39.262,
+      "grad_norm": 1.442545771598816,
+      "learning_rate": 2e-05,
+      "loss": 0.04413161,
+      "step": 19631
+    },
+    {
+      "epoch": 39.264,
+      "grad_norm": 1.0915734767913818,
+      "learning_rate": 2e-05,
+      "loss": 0.03375326,
+      "step": 19632
+    },
+    {
+      "epoch": 39.266,
+      "grad_norm": 1.3909891843795776,
+      "learning_rate": 2e-05,
+      "loss": 0.0558477,
+      "step": 19633
+    },
+    {
+      "epoch": 39.268,
+      "grad_norm": 1.2892683744430542,
+      "learning_rate": 2e-05,
+      "loss": 0.05482936,
+      "step": 19634
+    },
+    {
+      "epoch": 39.27,
+      "grad_norm": 1.4405442476272583,
+      "learning_rate": 2e-05,
+      "loss": 0.05377769,
+      "step": 19635
+    },
+    {
+      "epoch": 39.272,
+      "grad_norm": 1.9378975629806519,
+      "learning_rate": 2e-05,
+      "loss": 0.04612908,
+      "step": 19636
+    },
+    {
+      "epoch": 39.274,
+      "grad_norm": 0.9840026497840881,
+      "learning_rate": 2e-05,
+      "loss": 0.04712234,
+      "step": 19637
+    },
+    {
+      "epoch": 39.276,
+      "grad_norm": 1.2211089134216309,
+      "learning_rate": 2e-05,
+      "loss": 0.05115188,
+      "step": 19638
+    },
+    {
+      "epoch": 39.278,
+      "grad_norm": 2.96844220161438,
+      "learning_rate": 2e-05,
+      "loss": 0.04160873,
+      "step": 19639
+    },
+    {
+      "epoch": 39.28,
+      "grad_norm": 2.2236135005950928,
+      "learning_rate": 2e-05,
+      "loss": 0.03255095,
+      "step": 19640
+    },
+    {
+      "epoch": 39.282,
+      "grad_norm": 1.4627702236175537,
+      "learning_rate": 2e-05,
+      "loss": 0.06116011,
+      "step": 19641
+    },
+    {
+      "epoch": 39.284,
+      "grad_norm": 1.137810468673706,
+      "learning_rate": 2e-05,
+      "loss": 0.04151687,
+      "step": 19642
+    },
+    {
+      "epoch": 39.286,
+      "grad_norm": 1.4149549007415771,
+      "learning_rate": 2e-05,
+      "loss": 0.0420138,
+      "step": 19643
+    },
+    {
+      "epoch": 39.288,
+      "grad_norm": 1.1547538042068481,
+      "learning_rate": 2e-05,
+      "loss": 0.04780672,
+      "step": 19644
+    },
+    {
+      "epoch": 39.29,
+      "grad_norm": 1.460732340812683,
+      "learning_rate": 2e-05,
+      "loss": 0.04886123,
+      "step": 19645
+    },
+    {
+      "epoch": 39.292,
+      "grad_norm": 1.4473869800567627,
+      "learning_rate": 2e-05,
+      "loss": 0.04399436,
+      "step": 19646
+    },
+    {
+      "epoch": 39.294,
+      "grad_norm": 1.1946806907653809,
+      "learning_rate": 2e-05,
+      "loss": 0.04288406,
+      "step": 19647
+    },
+    {
+      "epoch": 39.296,
+      "grad_norm": 1.7457919120788574,
+      "learning_rate": 2e-05,
+      "loss": 0.05616899,
+      "step": 19648
+    },
+    {
+      "epoch": 39.298,
+      "grad_norm": 1.1243897676467896,
+      "learning_rate": 2e-05,
+      "loss": 0.04388375,
+      "step": 19649
+    },
+    {
+      "epoch": 39.3,
+      "grad_norm": 1.2903263568878174,
+      "learning_rate": 2e-05,
+      "loss": 0.0539849,
+      "step": 19650
+    },
+    {
+      "epoch": 39.302,
+      "grad_norm": 1.4681276082992554,
+      "learning_rate": 2e-05,
+      "loss": 0.06013412,
+      "step": 19651
+    },
+    {
+      "epoch": 39.304,
+      "grad_norm": 1.7446203231811523,
+      "learning_rate": 2e-05,
+      "loss": 0.07473246,
+      "step": 19652
+    },
+    {
+      "epoch": 39.306,
+      "grad_norm": 1.5309257507324219,
+      "learning_rate": 2e-05,
+      "loss": 0.06777739,
+      "step": 19653
+    },
+    {
+      "epoch": 39.308,
+      "grad_norm": 1.5686246156692505,
+      "learning_rate": 2e-05,
+      "loss": 0.06154291,
+      "step": 19654
+    },
+    {
+      "epoch": 39.31,
+      "grad_norm": 1.139773964881897,
+      "learning_rate": 2e-05,
+      "loss": 0.03822503,
+      "step": 19655
+    },
+    {
+      "epoch": 39.312,
+      "grad_norm": 2.308309555053711,
+      "learning_rate": 2e-05,
+      "loss": 0.04015625,
+      "step": 19656
+    },
+    {
+      "epoch": 39.314,
+      "grad_norm": 1.5338242053985596,
+      "learning_rate": 2e-05,
+      "loss": 0.05909691,
+      "step": 19657
+    },
+    {
+      "epoch": 39.316,
+      "grad_norm": 1.2191451787948608,
+      "learning_rate": 2e-05,
+      "loss": 0.04259738,
+      "step": 19658
+    },
+    {
+      "epoch": 39.318,
+      "grad_norm": 1.2544686794281006,
+      "learning_rate": 2e-05,
+      "loss": 0.06194799,
+      "step": 19659
+    },
+    {
+      "epoch": 39.32,
+      "grad_norm": 1.1385127305984497,
+      "learning_rate": 2e-05,
+      "loss": 0.04202759,
+      "step": 19660
+    },
+    {
+      "epoch": 39.322,
+      "grad_norm": 1.508982539176941,
+      "learning_rate": 2e-05,
+      "loss": 0.0428558,
+      "step": 19661
+    },
+    {
+      "epoch": 39.324,
+      "grad_norm": 1.4978950023651123,
+      "learning_rate": 2e-05,
+      "loss": 0.06243214,
+      "step": 19662
+    },
+    {
+      "epoch": 39.326,
+      "grad_norm": 2.656769037246704,
+      "learning_rate": 2e-05,
+      "loss": 0.06248422,
+      "step": 19663
+    },
+    {
+      "epoch": 39.328,
+      "grad_norm": 1.2013676166534424,
+      "learning_rate": 2e-05,
+      "loss": 0.04629922,
+      "step": 19664
+    },
+    {
+      "epoch": 39.33,
+      "grad_norm": 1.4212125539779663,
+      "learning_rate": 2e-05,
+      "loss": 0.0516174,
+      "step": 19665
+    },
+    {
+      "epoch": 39.332,
+      "grad_norm": 1.4654594659805298,
+      "learning_rate": 2e-05,
+      "loss": 0.04654077,
+      "step": 19666
+    },
+    {
+      "epoch": 39.334,
+      "grad_norm": 1.1820266246795654,
+      "learning_rate": 2e-05,
+      "loss": 0.04918681,
+      "step": 19667
+    },
+    {
+      "epoch": 39.336,
+      "grad_norm": 1.7852951288223267,
+      "learning_rate": 2e-05,
+      "loss": 0.05676861,
+      "step": 19668
+    },
+    {
+      "epoch": 39.338,
+      "grad_norm": 1.3146127462387085,
+      "learning_rate": 2e-05,
+      "loss": 0.06642872,
+      "step": 19669
+    },
+    {
+      "epoch": 39.34,
+      "grad_norm": 1.1040120124816895,
+      "learning_rate": 2e-05,
+      "loss": 0.03930391,
+      "step": 19670
+    },
+    {
+      "epoch": 39.342,
+      "grad_norm": 1.4802639484405518,
+      "learning_rate": 2e-05,
+      "loss": 0.05159596,
+      "step": 19671
+    },
+    {
+      "epoch": 39.344,
+      "grad_norm": 1.3350907564163208,
+      "learning_rate": 2e-05,
+      "loss": 0.05917858,
+      "step": 19672
+    },
+    {
+      "epoch": 39.346,
+      "grad_norm": 1.3922500610351562,
+      "learning_rate": 2e-05,
+      "loss": 0.04254092,
+      "step": 19673
+    },
+    {
+      "epoch": 39.348,
+      "grad_norm": 1.271043062210083,
+      "learning_rate": 2e-05,
+      "loss": 0.05474164,
+      "step": 19674
+    },
+    {
+      "epoch": 39.35,
+      "grad_norm": 1.4507440328598022,
+      "learning_rate": 2e-05,
+      "loss": 0.05754908,
+      "step": 19675
+    },
+    {
+      "epoch": 39.352,
+      "grad_norm": 1.2747180461883545,
+      "learning_rate": 2e-05,
+      "loss": 0.06754243,
+      "step": 19676
+    },
+    {
+      "epoch": 39.354,
+      "grad_norm": 1.7789802551269531,
+      "learning_rate": 2e-05,
+      "loss": 0.0436502,
+      "step": 19677
+    },
+    {
+      "epoch": 39.356,
+      "grad_norm": 2.4285247325897217,
+      "learning_rate": 2e-05,
+      "loss": 0.05950401,
+      "step": 19678
+    },
+    {
+      "epoch": 39.358,
+      "grad_norm": 1.8559372425079346,
+      "learning_rate": 2e-05,
+      "loss": 0.04614679,
+      "step": 19679
+    },
+    {
+      "epoch": 39.36,
+      "grad_norm": 1.1284407377243042,
+      "learning_rate": 2e-05,
+      "loss": 0.05482705,
+      "step": 19680
+    },
+    {
+      "epoch": 39.362,
+      "grad_norm": 1.2401177883148193,
+      "learning_rate": 2e-05,
+      "loss": 0.04206383,
+      "step": 19681
+    },
+    {
+      "epoch": 39.364,
+      "grad_norm": 1.6358333826065063,
+      "learning_rate": 2e-05,
+      "loss": 0.04896562,
+      "step": 19682
+    },
+    {
+      "epoch": 39.366,
+      "grad_norm": 1.0610060691833496,
+      "learning_rate": 2e-05,
+      "loss": 0.03826354,
+      "step": 19683
+    },
+    {
+      "epoch": 39.368,
+      "grad_norm": 1.349070429801941,
+      "learning_rate": 2e-05,
+      "loss": 0.05081733,
+      "step": 19684
+    },
+    {
+      "epoch": 39.37,
+      "grad_norm": 1.0122675895690918,
+      "learning_rate": 2e-05,
+      "loss": 0.03560245,
+      "step": 19685
+    },
+    {
+      "epoch": 39.372,
+      "grad_norm": 1.1908326148986816,
+      "learning_rate": 2e-05,
+      "loss": 0.04643635,
+      "step": 19686
+    },
+    {
+      "epoch": 39.374,
+      "grad_norm": 1.500412106513977,
+      "learning_rate": 2e-05,
+      "loss": 0.04018075,
+      "step": 19687
+    },
+    {
+      "epoch": 39.376,
+      "grad_norm": 1.1961153745651245,
+      "learning_rate": 2e-05,
+      "loss": 0.03434346,
+      "step": 19688
+    },
+    {
+      "epoch": 39.378,
+      "grad_norm": 1.0913282632827759,
+      "learning_rate": 2e-05,
+      "loss": 0.03738025,
+      "step": 19689
+    },
+    {
+      "epoch": 39.38,
+      "grad_norm": 1.1147562265396118,
+      "learning_rate": 2e-05,
+      "loss": 0.04534988,
+      "step": 19690
+    },
+    {
+      "epoch": 39.382,
+      "grad_norm": 1.3109427690505981,
+      "learning_rate": 2e-05,
+      "loss": 0.04121095,
+      "step": 19691
+    },
+    {
+      "epoch": 39.384,
+      "grad_norm": 1.143779993057251,
+      "learning_rate": 2e-05,
+      "loss": 0.03789821,
+      "step": 19692
+    },
+    {
+      "epoch": 39.386,
+      "grad_norm": 1.3124995231628418,
+      "learning_rate": 2e-05,
+      "loss": 0.04957093,
+      "step": 19693
+    },
+    {
+      "epoch": 39.388,
+      "grad_norm": 1.287955403327942,
+      "learning_rate": 2e-05,
+      "loss": 0.05975793,
+      "step": 19694
+    },
+    {
+      "epoch": 39.39,
+      "grad_norm": 2.0113747119903564,
+      "learning_rate": 2e-05,
+      "loss": 0.0466843,
+      "step": 19695
+    },
+    {
+      "epoch": 39.392,
+      "grad_norm": 1.1586730480194092,
+      "learning_rate": 2e-05,
+      "loss": 0.04480777,
+      "step": 19696
+    },
+    {
+      "epoch": 39.394,
+      "grad_norm": 1.1337339878082275,
+      "learning_rate": 2e-05,
+      "loss": 0.03835721,
+      "step": 19697
+    },
+    {
+      "epoch": 39.396,
+      "grad_norm": 1.8198338747024536,
+      "learning_rate": 2e-05,
+      "loss": 0.04342664,
+      "step": 19698
+    },
+    {
+      "epoch": 39.398,
+      "grad_norm": 1.068575382232666,
+      "learning_rate": 2e-05,
+      "loss": 0.02720687,
+      "step": 19699
+    },
+    {
+      "epoch": 39.4,
+      "grad_norm": 2.4282875061035156,
+      "learning_rate": 2e-05,
+      "loss": 0.05048774,
+      "step": 19700
+    },
+    {
+      "epoch": 39.402,
+      "grad_norm": 1.351075291633606,
+      "learning_rate": 2e-05,
+      "loss": 0.05489894,
+      "step": 19701
+    },
+    {
+      "epoch": 39.404,
+      "grad_norm": 1.1262435913085938,
+      "learning_rate": 2e-05,
+      "loss": 0.0376655,
+      "step": 19702
+    },
+    {
+      "epoch": 39.406,
+      "grad_norm": 1.3584601879119873,
+      "learning_rate": 2e-05,
+      "loss": 0.05599149,
+      "step": 19703
+    },
+    {
+      "epoch": 39.408,
+      "grad_norm": 1.2703129053115845,
+      "learning_rate": 2e-05,
+      "loss": 0.06736299,
+      "step": 19704
+    },
+    {
+      "epoch": 39.41,
+      "grad_norm": 1.1085890531539917,
+      "learning_rate": 2e-05,
+      "loss": 0.04166878,
+      "step": 19705
+    },
+    {
+      "epoch": 39.412,
+      "grad_norm": 1.3138976097106934,
+      "learning_rate": 2e-05,
+      "loss": 0.05483936,
+      "step": 19706
+    },
+    {
+      "epoch": 39.414,
+      "grad_norm": 1.3568849563598633,
+      "learning_rate": 2e-05,
+      "loss": 0.0636967,
+      "step": 19707
+    },
+    {
+      "epoch": 39.416,
+      "grad_norm": 1.2971667051315308,
+      "learning_rate": 2e-05,
+      "loss": 0.05253425,
+      "step": 19708
+    },
+    {
+      "epoch": 39.418,
+      "grad_norm": 1.1735317707061768,
+      "learning_rate": 2e-05,
+      "loss": 0.03824739,
+      "step": 19709
+    },
+    {
+      "epoch": 39.42,
+      "grad_norm": 1.3102675676345825,
+      "learning_rate": 2e-05,
+      "loss": 0.05568937,
+      "step": 19710
+    },
+    {
+      "epoch": 39.422,
+      "grad_norm": 1.6656438112258911,
+      "learning_rate": 2e-05,
+      "loss": 0.04870186,
+      "step": 19711
+    },
+    {
+      "epoch": 39.424,
+      "grad_norm": 1.870893120765686,
+      "learning_rate": 2e-05,
+      "loss": 0.06180416,
+      "step": 19712
+    },
+    {
+      "epoch": 39.426,
+      "grad_norm": 1.1797996759414673,
+      "learning_rate": 2e-05,
+      "loss": 0.04374686,
+      "step": 19713
+    },
+    {
+      "epoch": 39.428,
+      "grad_norm": 1.4457714557647705,
+      "learning_rate": 2e-05,
+      "loss": 0.04901951,
+      "step": 19714
+    },
+    {
+      "epoch": 39.43,
+      "grad_norm": 1.1880093812942505,
+      "learning_rate": 2e-05,
+      "loss": 0.05779697,
+      "step": 19715
+    },
+    {
+      "epoch": 39.432,
+      "grad_norm": 1.140159249305725,
+      "learning_rate": 2e-05,
+      "loss": 0.04950743,
+      "step": 19716
+    },
+    {
+      "epoch": 39.434,
+      "grad_norm": 1.0973641872406006,
+      "learning_rate": 2e-05,
+      "loss": 0.03795099,
+      "step": 19717
+    },
+    {
+      "epoch": 39.436,
+      "grad_norm": 1.176837682723999,
+      "learning_rate": 2e-05,
+      "loss": 0.04808137,
+      "step": 19718
+    },
+    {
+      "epoch": 39.438,
+      "grad_norm": 1.2320091724395752,
+      "learning_rate": 2e-05,
+      "loss": 0.04875737,
+      "step": 19719
+    },
+    {
+      "epoch": 39.44,
+      "grad_norm": 1.197391390800476,
+      "learning_rate": 2e-05,
+      "loss": 0.03575107,
+      "step": 19720
+    },
+    {
+      "epoch": 39.442,
+      "grad_norm": 1.4280883073806763,
+      "learning_rate": 2e-05,
+      "loss": 0.05063617,
+      "step": 19721
+    },
+    {
+      "epoch": 39.444,
+      "grad_norm": 1.6574267148971558,
+      "learning_rate": 2e-05,
+      "loss": 0.04451539,
+      "step": 19722
+    },
+    {
+      "epoch": 39.446,
+      "grad_norm": 1.1137486696243286,
+      "learning_rate": 2e-05,
+      "loss": 0.03630831,
+      "step": 19723
+    },
+    {
+      "epoch": 39.448,
+      "grad_norm": 1.1743730306625366,
+      "learning_rate": 2e-05,
+      "loss": 0.04013621,
+      "step": 19724
+    },
+    {
+      "epoch": 39.45,
+      "grad_norm": 1.239919900894165,
+      "learning_rate": 2e-05,
+      "loss": 0.05531546,
+      "step": 19725
+    },
+    {
+      "epoch": 39.452,
+      "grad_norm": 1.6629247665405273,
+      "learning_rate": 2e-05,
+      "loss": 0.06353351,
+      "step": 19726
+    },
+    {
+      "epoch": 39.454,
+      "grad_norm": 1.5018807649612427,
+      "learning_rate": 2e-05,
+      "loss": 0.07370369,
+      "step": 19727
+    },
+    {
+      "epoch": 39.456,
+      "grad_norm": 1.5102373361587524,
+      "learning_rate": 2e-05,
+      "loss": 0.0483415,
+      "step": 19728
+    },
+    {
+      "epoch": 39.458,
+      "grad_norm": 1.106083631515503,
+      "learning_rate": 2e-05,
+      "loss": 0.04570149,
+      "step": 19729
+    },
+    {
+      "epoch": 39.46,
+      "grad_norm": 1.376544713973999,
+      "learning_rate": 2e-05,
+      "loss": 0.05271851,
+      "step": 19730
+    },
+    {
+      "epoch": 39.462,
+      "grad_norm": 1.1500600576400757,
+      "learning_rate": 2e-05,
+      "loss": 0.05022399,
+      "step": 19731
+    },
+    {
+      "epoch": 39.464,
+      "grad_norm": 1.136629343032837,
+      "learning_rate": 2e-05,
+      "loss": 0.04801833,
+      "step": 19732
+    },
+    {
+      "epoch": 39.466,
+      "grad_norm": 1.6498074531555176,
+      "learning_rate": 2e-05,
+      "loss": 0.06893224,
+      "step": 19733
+    },
+    {
+      "epoch": 39.468,
+      "grad_norm": 1.306951642036438,
+      "learning_rate": 2e-05,
+      "loss": 0.06067065,
+      "step": 19734
+    },
+    {
+      "epoch": 39.47,
+      "grad_norm": 1.1913270950317383,
+      "learning_rate": 2e-05,
+      "loss": 0.03944825,
+      "step": 19735
+    },
+    {
+      "epoch": 39.472,
+      "grad_norm": 1.2636955976486206,
+      "learning_rate": 2e-05,
+      "loss": 0.04280094,
+      "step": 19736
+    },
+    {
+      "epoch": 39.474,
+      "grad_norm": 1.0089428424835205,
+      "learning_rate": 2e-05,
+      "loss": 0.03867158,
+      "step": 19737
+    },
+    {
+      "epoch": 39.476,
+      "grad_norm": 1.6654586791992188,
+      "learning_rate": 2e-05,
+      "loss": 0.06643804,
+      "step": 19738
+    },
+    {
+      "epoch": 39.478,
+      "grad_norm": 2.2843704223632812,
+      "learning_rate": 2e-05,
+      "loss": 0.05996677,
+      "step": 19739
+    },
+    {
+      "epoch": 39.48,
+      "grad_norm": 1.0888022184371948,
+      "learning_rate": 2e-05,
+      "loss": 0.0362496,
+      "step": 19740
+    },
+    {
+      "epoch": 39.482,
+      "grad_norm": 1.196286916732788,
+      "learning_rate": 2e-05,
+      "loss": 0.04796268,
+      "step": 19741
+    },
+    {
+      "epoch": 39.484,
+      "grad_norm": 1.6246408224105835,
+      "learning_rate": 2e-05,
+      "loss": 0.06164134,
+      "step": 19742
+    },
+    {
+      "epoch": 39.486,
+      "grad_norm": 1.4118708372116089,
+      "learning_rate": 2e-05,
+      "loss": 0.06805681,
+      "step": 19743
+    },
+    {
+      "epoch": 39.488,
+      "grad_norm": 1.5130550861358643,
+      "learning_rate": 2e-05,
+      "loss": 0.0542859,
+      "step": 19744
+    },
+    {
+      "epoch": 39.49,
+      "grad_norm": 0.8534611463546753,
+      "learning_rate": 2e-05,
+      "loss": 0.02287557,
+      "step": 19745
+    },
+    {
+      "epoch": 39.492,
+      "grad_norm": 1.087668538093567,
+      "learning_rate": 2e-05,
+      "loss": 0.03380512,
+      "step": 19746
+    },
+    {
+      "epoch": 39.494,
+      "grad_norm": 1.2685211896896362,
+      "learning_rate": 2e-05,
+      "loss": 0.05318061,
+      "step": 19747
+    },
+    {
+      "epoch": 39.496,
+      "grad_norm": 1.2841920852661133,
+      "learning_rate": 2e-05,
+      "loss": 0.05055482,
+      "step": 19748
+    },
+    {
+      "epoch": 39.498,
+      "grad_norm": 1.1762704849243164,
+      "learning_rate": 2e-05,
+      "loss": 0.04554278,
+      "step": 19749
+    },
+    {
+      "epoch": 39.5,
+      "grad_norm": 3.7946743965148926,
+      "learning_rate": 2e-05,
+      "loss": 0.03693722,
+      "step": 19750
+    },
+    {
+      "epoch": 39.502,
+      "grad_norm": 1.1349517107009888,
+      "learning_rate": 2e-05,
+      "loss": 0.02860119,
+      "step": 19751
+    },
+    {
+      "epoch": 39.504,
+      "grad_norm": 2.01347017288208,
+      "learning_rate": 2e-05,
+      "loss": 0.04260724,
+      "step": 19752
+    },
+    {
+      "epoch": 39.506,
+      "grad_norm": 1.267555594444275,
+      "learning_rate": 2e-05,
+      "loss": 0.05181779,
+      "step": 19753
+    },
+    {
+      "epoch": 39.508,
+      "grad_norm": 1.1315144300460815,
+      "learning_rate": 2e-05,
+      "loss": 0.04263924,
+      "step": 19754
+    },
+    {
+      "epoch": 39.51,
+      "grad_norm": 2.0047526359558105,
+      "learning_rate": 2e-05,
+      "loss": 0.04533667,
+      "step": 19755
+    },
+    {
+      "epoch": 39.512,
+      "grad_norm": 1.342085361480713,
+      "learning_rate": 2e-05,
+      "loss": 0.04838677,
+      "step": 19756
+    },
+    {
+      "epoch": 39.514,
+      "grad_norm": 1.331516146659851,
+      "learning_rate": 2e-05,
+      "loss": 0.06132433,
+      "step": 19757
+    },
+    {
+      "epoch": 39.516,
+      "grad_norm": 1.1576191186904907,
+      "learning_rate": 2e-05,
+      "loss": 0.04208809,
+      "step": 19758
+    },
+    {
+      "epoch": 39.518,
+      "grad_norm": 0.9372735023498535,
+      "learning_rate": 2e-05,
+      "loss": 0.03229483,
+      "step": 19759
+    },
+    {
+      "epoch": 39.52,
+      "grad_norm": 2.456406831741333,
+      "learning_rate": 2e-05,
+      "loss": 0.05844424,
+      "step": 19760
+    },
+    {
+      "epoch": 39.522,
+      "grad_norm": 1.2554352283477783,
+      "learning_rate": 2e-05,
+      "loss": 0.05017581,
+      "step": 19761
+    },
+    {
+      "epoch": 39.524,
+      "grad_norm": 1.6045249700546265,
+      "learning_rate": 2e-05,
+      "loss": 0.06052469,
+      "step": 19762
+    },
+    {
+      "epoch": 39.526,
+      "grad_norm": 1.3831950426101685,
+      "learning_rate": 2e-05,
+      "loss": 0.05542463,
+      "step": 19763
+    },
+    {
+      "epoch": 39.528,
+      "grad_norm": 1.6362686157226562,
+      "learning_rate": 2e-05,
+      "loss": 0.05046823,
+      "step": 19764
+    },
+    {
+      "epoch": 39.53,
+      "grad_norm": 1.8844678401947021,
+      "learning_rate": 2e-05,
+      "loss": 0.05892744,
+      "step": 19765
+    },
+    {
+      "epoch": 39.532,
+      "grad_norm": 1.026892066001892,
+      "learning_rate": 2e-05,
+      "loss": 0.04069612,
+      "step": 19766
+    },
+    {
+      "epoch": 39.534,
+      "grad_norm": 1.6162285804748535,
+      "learning_rate": 2e-05,
+      "loss": 0.04750034,
+      "step": 19767
+    },
+    {
+      "epoch": 39.536,
+      "grad_norm": 1.3901042938232422,
+      "learning_rate": 2e-05,
+      "loss": 0.05620121,
+      "step": 19768
+    },
+    {
+      "epoch": 39.538,
+      "grad_norm": 1.1447288990020752,
+      "learning_rate": 2e-05,
+      "loss": 0.04215189,
+      "step": 19769
+    },
+    {
+      "epoch": 39.54,
+      "grad_norm": 2.04343843460083,
+      "learning_rate": 2e-05,
+      "loss": 0.04056927,
+      "step": 19770
+    },
+    {
+      "epoch": 39.542,
+      "grad_norm": 1.376480221748352,
+      "learning_rate": 2e-05,
+      "loss": 0.04138695,
+      "step": 19771
+    },
+    {
+      "epoch": 39.544,
+      "grad_norm": 1.5023736953735352,
+      "learning_rate": 2e-05,
+      "loss": 0.04907219,
+      "step": 19772
+    },
+    {
+      "epoch": 39.546,
+      "grad_norm": 2.4118905067443848,
+      "learning_rate": 2e-05,
+      "loss": 0.05148922,
+      "step": 19773
+    },
+    {
+      "epoch": 39.548,
+      "grad_norm": 3.517075300216675,
+      "learning_rate": 2e-05,
+      "loss": 0.05087829,
+      "step": 19774
+    },
+    {
+      "epoch": 39.55,
+      "grad_norm": 1.0423048734664917,
+      "learning_rate": 2e-05,
+      "loss": 0.03624145,
+      "step": 19775
+    },
+    {
+      "epoch": 39.552,
+      "grad_norm": 1.3357534408569336,
+      "learning_rate": 2e-05,
+      "loss": 0.04405185,
+      "step": 19776
+    },
+    {
+      "epoch": 39.554,
+      "grad_norm": 1.2442307472229004,
+      "learning_rate": 2e-05,
+      "loss": 0.04776893,
+      "step": 19777
+    },
+    {
+      "epoch": 39.556,
+      "grad_norm": 1.0064873695373535,
+      "learning_rate": 2e-05,
+      "loss": 0.03839774,
+      "step": 19778
+    },
+    {
+      "epoch": 39.558,
+      "grad_norm": 1.5154045820236206,
+      "learning_rate": 2e-05,
+      "loss": 0.04444873,
+      "step": 19779
+    },
+    {
+      "epoch": 39.56,
+      "grad_norm": 1.2269160747528076,
+      "learning_rate": 2e-05,
+      "loss": 0.04485428,
+      "step": 19780
+    },
+    {
+      "epoch": 39.562,
+      "grad_norm": 1.1330827474594116,
+      "learning_rate": 2e-05,
+      "loss": 0.04535697,
+      "step": 19781
+    },
+    {
+      "epoch": 39.564,
+      "grad_norm": 1.6141606569290161,
+      "learning_rate": 2e-05,
+      "loss": 0.04330592,
+      "step": 19782
+    },
+    {
+      "epoch": 39.566,
+      "grad_norm": 1.2627594470977783,
+      "learning_rate": 2e-05,
+      "loss": 0.03273316,
+      "step": 19783
+    },
+    {
+      "epoch": 39.568,
+      "grad_norm": 1.0792535543441772,
+      "learning_rate": 2e-05,
+      "loss": 0.0406695,
+      "step": 19784
+    },
+    {
+      "epoch": 39.57,
+      "grad_norm": 1.397845983505249,
+      "learning_rate": 2e-05,
+      "loss": 0.05157507,
+      "step": 19785
+    },
+    {
+      "epoch": 39.572,
+      "grad_norm": 2.2438647747039795,
+      "learning_rate": 2e-05,
+      "loss": 0.067011,
+      "step": 19786
+    },
+    {
+      "epoch": 39.574,
+      "grad_norm": 1.2847806215286255,
+      "learning_rate": 2e-05,
+      "loss": 0.03849096,
+      "step": 19787
+    },
+    {
+      "epoch": 39.576,
+      "grad_norm": 1.204473614692688,
+      "learning_rate": 2e-05,
+      "loss": 0.05438119,
+      "step": 19788
+    },
+    {
+      "epoch": 39.578,
+      "grad_norm": 1.9760804176330566,
+      "learning_rate": 2e-05,
+      "loss": 0.05682136,
+      "step": 19789
+    },
+    {
+      "epoch": 39.58,
+      "grad_norm": 1.181208610534668,
+      "learning_rate": 2e-05,
+      "loss": 0.0492033,
+      "step": 19790
+    },
+    {
+      "epoch": 39.582,
+      "grad_norm": 1.273768424987793,
+      "learning_rate": 2e-05,
+      "loss": 0.04544124,
+      "step": 19791
+    },
+    {
+      "epoch": 39.584,
+      "grad_norm": 1.5114247798919678,
+      "learning_rate": 2e-05,
+      "loss": 0.06623958,
+      "step": 19792
+    },
+    {
+      "epoch": 39.586,
+      "grad_norm": 1.1232013702392578,
+      "learning_rate": 2e-05,
+      "loss": 0.04789654,
+      "step": 19793
+    },
+    {
+      "epoch": 39.588,
+      "grad_norm": 1.1755037307739258,
+      "learning_rate": 2e-05,
+      "loss": 0.050361,
+      "step": 19794
+    },
+    {
+      "epoch": 39.59,
+      "grad_norm": 1.6072686910629272,
+      "learning_rate": 2e-05,
+      "loss": 0.06229908,
+      "step": 19795
+    },
+    {
+      "epoch": 39.592,
+      "grad_norm": 1.416585087776184,
+      "learning_rate": 2e-05,
+      "loss": 0.04778691,
+      "step": 19796
+    },
+    {
+      "epoch": 39.594,
+      "grad_norm": 1.3451114892959595,
+      "learning_rate": 2e-05,
+      "loss": 0.05900617,
+      "step": 19797
+    },
+    {
+      "epoch": 39.596,
+      "grad_norm": 1.1347873210906982,
+      "learning_rate": 2e-05,
+      "loss": 0.04485462,
+      "step": 19798
+    },
+    {
+      "epoch": 39.598,
+      "grad_norm": 1.158407211303711,
+      "learning_rate": 2e-05,
+      "loss": 0.04194099,
+      "step": 19799
+    },
+    {
+      "epoch": 39.6,
+      "grad_norm": 1.222896695137024,
+      "learning_rate": 2e-05,
+      "loss": 0.04154717,
+      "step": 19800
+    },
+    {
+      "epoch": 39.602,
+      "grad_norm": 1.1443710327148438,
+      "learning_rate": 2e-05,
+      "loss": 0.04118493,
+      "step": 19801
+    },
+    {
+      "epoch": 39.604,
+      "grad_norm": 1.2057150602340698,
+      "learning_rate": 2e-05,
+      "loss": 0.05563217,
+      "step": 19802
+    },
+    {
+      "epoch": 39.606,
+      "grad_norm": 1.192069172859192,
+      "learning_rate": 2e-05,
+      "loss": 0.0467116,
+      "step": 19803
+    },
+    {
+      "epoch": 39.608,
+      "grad_norm": 1.410652756690979,
+      "learning_rate": 2e-05,
+      "loss": 0.03325736,
+      "step": 19804
+    },
+    {
+      "epoch": 39.61,
+      "grad_norm": 1.1169686317443848,
+      "learning_rate": 2e-05,
+      "loss": 0.03927244,
+      "step": 19805
+    },
+    {
+      "epoch": 39.612,
+      "grad_norm": 1.2299331426620483,
+      "learning_rate": 2e-05,
+      "loss": 0.04283129,
+      "step": 19806
+    },
+    {
+      "epoch": 39.614,
+      "grad_norm": 1.020370602607727,
+      "learning_rate": 2e-05,
+      "loss": 0.03461327,
+      "step": 19807
+    },
+    {
+      "epoch": 39.616,
+      "grad_norm": 1.2263249158859253,
+      "learning_rate": 2e-05,
+      "loss": 0.05317787,
+      "step": 19808
+    },
+    {
+      "epoch": 39.618,
+      "grad_norm": 2.3587546348571777,
+      "learning_rate": 2e-05,
+      "loss": 0.04611384,
+      "step": 19809
+    },
+    {
+      "epoch": 39.62,
+      "grad_norm": 1.2532998323440552,
+      "learning_rate": 2e-05,
+      "loss": 0.04584979,
+      "step": 19810
+    },
+    {
+      "epoch": 39.622,
+      "grad_norm": 0.9980117082595825,
+      "learning_rate": 2e-05,
+      "loss": 0.03598379,
+      "step": 19811
+    },
+    {
+      "epoch": 39.624,
+      "grad_norm": 1.3296000957489014,
+      "learning_rate": 2e-05,
+      "loss": 0.06023554,
+      "step": 19812
+    },
+    {
+      "epoch": 39.626,
+      "grad_norm": 1.3050363063812256,
+      "learning_rate": 2e-05,
+      "loss": 0.03866673,
+      "step": 19813
+    },
+    {
+      "epoch": 39.628,
+      "grad_norm": 1.0335310697555542,
+      "learning_rate": 2e-05,
+      "loss": 0.03773557,
+      "step": 19814
+    },
+    {
+      "epoch": 39.63,
+      "grad_norm": 1.102943778038025,
+      "learning_rate": 2e-05,
+      "loss": 0.05241087,
+      "step": 19815
+    },
+    {
+      "epoch": 39.632,
+      "grad_norm": 2.9029152393341064,
+      "learning_rate": 2e-05,
+      "loss": 0.06153958,
+      "step": 19816
+    },
+    {
+      "epoch": 39.634,
+      "grad_norm": 1.207022786140442,
+      "learning_rate": 2e-05,
+      "loss": 0.0496684,
+      "step": 19817
+    },
+    {
+      "epoch": 39.636,
+      "grad_norm": 1.147627353668213,
+      "learning_rate": 2e-05,
+      "loss": 0.04096475,
+      "step": 19818
+    },
+    {
+      "epoch": 39.638,
+      "grad_norm": 1.1633808612823486,
+      "learning_rate": 2e-05,
+      "loss": 0.0518549,
+      "step": 19819
+    },
+    {
+      "epoch": 39.64,
+      "grad_norm": 2.9582087993621826,
+      "learning_rate": 2e-05,
+      "loss": 0.04467195,
+      "step": 19820
+    },
+    {
+      "epoch": 39.642,
+      "grad_norm": 1.5623703002929688,
+      "learning_rate": 2e-05,
+      "loss": 0.04819065,
+      "step": 19821
+    },
+    {
+      "epoch": 39.644,
+      "grad_norm": 1.2664448022842407,
+      "learning_rate": 2e-05,
+      "loss": 0.04783477,
+      "step": 19822
+    },
+    {
+      "epoch": 39.646,
+      "grad_norm": 1.313110589981079,
+      "learning_rate": 2e-05,
+      "loss": 0.03875414,
+      "step": 19823
+    },
+    {
+      "epoch": 39.648,
+      "grad_norm": 1.150399923324585,
+      "learning_rate": 2e-05,
+      "loss": 0.05495393,
+      "step": 19824
+    },
+    {
+      "epoch": 39.65,
+      "grad_norm": 1.273444414138794,
+      "learning_rate": 2e-05,
+      "loss": 0.05313744,
+      "step": 19825
+    },
+    {
+      "epoch": 39.652,
+      "grad_norm": 1.288045048713684,
+      "learning_rate": 2e-05,
+      "loss": 0.05415475,
+      "step": 19826
+    },
+    {
+      "epoch": 39.654,
+      "grad_norm": 1.1999372243881226,
+      "learning_rate": 2e-05,
+      "loss": 0.04567619,
+      "step": 19827
+    },
+    {
+      "epoch": 39.656,
+      "grad_norm": 1.0808748006820679,
+      "learning_rate": 2e-05,
+      "loss": 0.04751875,
+      "step": 19828
+    },
+    {
+      "epoch": 39.658,
+      "grad_norm": 1.122815489768982,
+      "learning_rate": 2e-05,
+      "loss": 0.0411781,
+      "step": 19829
+    },
+    {
+      "epoch": 39.66,
+      "grad_norm": 1.6972079277038574,
+      "learning_rate": 2e-05,
+      "loss": 0.05976425,
+      "step": 19830
+    },
+    {
+      "epoch": 39.662,
+      "grad_norm": 1.0641003847122192,
+      "learning_rate": 2e-05,
+      "loss": 0.0371467,
+      "step": 19831
+    },
+    {
+      "epoch": 39.664,
+      "grad_norm": 1.3201067447662354,
+      "learning_rate": 2e-05,
+      "loss": 0.04569467,
+      "step": 19832
+    },
+    {
+      "epoch": 39.666,
+      "grad_norm": 1.5276159048080444,
+      "learning_rate": 2e-05,
+      "loss": 0.04464874,
+      "step": 19833
+    },
+    {
+      "epoch": 39.668,
+      "grad_norm": 1.1895759105682373,
+      "learning_rate": 2e-05,
+      "loss": 0.04681417,
+      "step": 19834
+    },
+    {
+      "epoch": 39.67,
+      "grad_norm": 1.2756211757659912,
+      "learning_rate": 2e-05,
+      "loss": 0.04865978,
+      "step": 19835
+    },
+    {
+      "epoch": 39.672,
+      "grad_norm": 1.315328598022461,
+      "learning_rate": 2e-05,
+      "loss": 0.07096517,
+      "step": 19836
+    },
+    {
+      "epoch": 39.674,
+      "grad_norm": 1.2672423124313354,
+      "learning_rate": 2e-05,
+      "loss": 0.04316231,
+      "step": 19837
+    },
+    {
+      "epoch": 39.676,
+      "grad_norm": 1.077695369720459,
+      "learning_rate": 2e-05,
+      "loss": 0.03674557,
+      "step": 19838
+    },
+    {
+      "epoch": 39.678,
+      "grad_norm": 1.8062963485717773,
+      "learning_rate": 2e-05,
+      "loss": 0.05743162,
+      "step": 19839
+    },
+    {
+      "epoch": 39.68,
+      "grad_norm": 1.1730942726135254,
+      "learning_rate": 2e-05,
+      "loss": 0.03555157,
+      "step": 19840
+    },
+    {
+      "epoch": 39.682,
+      "grad_norm": 1.0135302543640137,
+      "learning_rate": 2e-05,
+      "loss": 0.02771145,
+      "step": 19841
+    },
+    {
+      "epoch": 39.684,
+      "grad_norm": 1.5072895288467407,
+      "learning_rate": 2e-05,
+      "loss": 0.06209271,
+      "step": 19842
+    },
+    {
+      "epoch": 39.686,
+      "grad_norm": 1.412246823310852,
+      "learning_rate": 2e-05,
+      "loss": 0.04875721,
+      "step": 19843
+    },
+    {
+      "epoch": 39.688,
+      "grad_norm": 1.330792784690857,
+      "learning_rate": 2e-05,
+      "loss": 0.05368386,
+      "step": 19844
+    },
+    {
+      "epoch": 39.69,
+      "grad_norm": 1.3758635520935059,
+      "learning_rate": 2e-05,
+      "loss": 0.05050598,
+      "step": 19845
+    },
+    {
+      "epoch": 39.692,
+      "grad_norm": 1.2462283372879028,
+      "learning_rate": 2e-05,
+      "loss": 0.0519301,
+      "step": 19846
+    },
+    {
+      "epoch": 39.694,
+      "grad_norm": 1.4314063787460327,
+      "learning_rate": 2e-05,
+      "loss": 0.04765672,
+      "step": 19847
+    },
+    {
+      "epoch": 39.696,
+      "grad_norm": 1.3259481191635132,
+      "learning_rate": 2e-05,
+      "loss": 0.05249347,
+      "step": 19848
+    },
+    {
+      "epoch": 39.698,
+      "grad_norm": 1.4222193956375122,
+      "learning_rate": 2e-05,
+      "loss": 0.0517953,
+      "step": 19849
+    },
+    {
+      "epoch": 39.7,
+      "grad_norm": 1.3300631046295166,
+      "learning_rate": 2e-05,
+      "loss": 0.05170576,
+      "step": 19850
+    },
+    {
+      "epoch": 39.702,
+      "grad_norm": 1.3391106128692627,
+      "learning_rate": 2e-05,
+      "loss": 0.05839144,
+      "step": 19851
+    },
+    {
+      "epoch": 39.704,
+      "grad_norm": 0.9856824278831482,
+      "learning_rate": 2e-05,
+      "loss": 0.03652864,
+      "step": 19852
+    },
+    {
+      "epoch": 39.706,
+      "grad_norm": 1.2734506130218506,
+      "learning_rate": 2e-05,
+      "loss": 0.05710208,
+      "step": 19853
+    },
+    {
+      "epoch": 39.708,
+      "grad_norm": 1.9725619554519653,
+      "learning_rate": 2e-05,
+      "loss": 0.05481605,
+      "step": 19854
+    },
+    {
+      "epoch": 39.71,
+      "grad_norm": 1.1700278520584106,
+      "learning_rate": 2e-05,
+      "loss": 0.05501316,
+      "step": 19855
+    },
+    {
+      "epoch": 39.712,
+      "grad_norm": 0.8477627635002136,
+      "learning_rate": 2e-05,
+      "loss": 0.02406826,
+      "step": 19856
+    },
+    {
+      "epoch": 39.714,
+      "grad_norm": 1.3238353729248047,
+      "learning_rate": 2e-05,
+      "loss": 0.04450976,
+      "step": 19857
+    },
+    {
+      "epoch": 39.716,
+      "grad_norm": 2.597429037094116,
+      "learning_rate": 2e-05,
+      "loss": 0.04625923,
+      "step": 19858
+    },
+    {
+      "epoch": 39.718,
+      "grad_norm": 1.2103281021118164,
+      "learning_rate": 2e-05,
+      "loss": 0.05398448,
+      "step": 19859
+    },
+    {
+      "epoch": 39.72,
+      "grad_norm": 1.249700665473938,
+      "learning_rate": 2e-05,
+      "loss": 0.05766566,
+      "step": 19860
+    },
+    {
+      "epoch": 39.722,
+      "grad_norm": 1.2194814682006836,
+      "learning_rate": 2e-05,
+      "loss": 0.05054,
+      "step": 19861
+    },
+    {
+      "epoch": 39.724,
+      "grad_norm": 1.3230112791061401,
+      "learning_rate": 2e-05,
+      "loss": 0.04640412,
+      "step": 19862
+    },
+    {
+      "epoch": 39.726,
+      "grad_norm": 1.38727867603302,
+      "learning_rate": 2e-05,
+      "loss": 0.04579895,
+      "step": 19863
+    },
+    {
+      "epoch": 39.728,
+      "grad_norm": 1.1224366426467896,
+      "learning_rate": 2e-05,
+      "loss": 0.03637297,
+      "step": 19864
+    },
+    {
+      "epoch": 39.73,
+      "grad_norm": 1.8750441074371338,
+      "learning_rate": 2e-05,
+      "loss": 0.06596401,
+      "step": 19865
+    },
+    {
+      "epoch": 39.732,
+      "grad_norm": 1.2330036163330078,
+      "learning_rate": 2e-05,
+      "loss": 0.06313409,
+      "step": 19866
+    },
+    {
+      "epoch": 39.734,
+      "grad_norm": 1.1567729711532593,
+      "learning_rate": 2e-05,
+      "loss": 0.03799208,
+      "step": 19867
+    },
+    {
+      "epoch": 39.736,
+      "grad_norm": 1.8047136068344116,
+      "learning_rate": 2e-05,
+      "loss": 0.05676999,
+      "step": 19868
+    },
+    {
+      "epoch": 39.738,
+      "grad_norm": 1.2310786247253418,
+      "learning_rate": 2e-05,
+      "loss": 0.04846741,
+      "step": 19869
+    },
+    {
+      "epoch": 39.74,
+      "grad_norm": 1.0939598083496094,
+      "learning_rate": 2e-05,
+      "loss": 0.05183221,
+      "step": 19870
+    },
+    {
+      "epoch": 39.742,
+      "grad_norm": 1.2205032110214233,
+      "learning_rate": 2e-05,
+      "loss": 0.05188471,
+      "step": 19871
+    },
+    {
+      "epoch": 39.744,
+      "grad_norm": 1.108619213104248,
+      "learning_rate": 2e-05,
+      "loss": 0.05545724,
+      "step": 19872
+    },
+    {
+      "epoch": 39.746,
+      "grad_norm": 1.2437492609024048,
+      "learning_rate": 2e-05,
+      "loss": 0.057475,
+      "step": 19873
+    },
+    {
+      "epoch": 39.748,
+      "grad_norm": 1.030137300491333,
+      "learning_rate": 2e-05,
+      "loss": 0.03131631,
+      "step": 19874
+    },
+    {
+      "epoch": 39.75,
+      "grad_norm": 1.290551781654358,
+      "learning_rate": 2e-05,
+      "loss": 0.05823213,
+      "step": 19875
+    },
+    {
+      "epoch": 39.752,
+      "grad_norm": 1.6363590955734253,
+      "learning_rate": 2e-05,
+      "loss": 0.05555134,
+      "step": 19876
+    },
+    {
+      "epoch": 39.754,
+      "grad_norm": 2.338663101196289,
+      "learning_rate": 2e-05,
+      "loss": 0.05460815,
+      "step": 19877
+    },
+    {
+      "epoch": 39.756,
+      "grad_norm": 1.5691533088684082,
+      "learning_rate": 2e-05,
+      "loss": 0.05019718,
+      "step": 19878
+    },
+    {
+      "epoch": 39.758,
+      "grad_norm": 1.2325677871704102,
+      "learning_rate": 2e-05,
+      "loss": 0.04595042,
+      "step": 19879
+    },
+    {
+      "epoch": 39.76,
+      "grad_norm": 1.2312854528427124,
+      "learning_rate": 2e-05,
+      "loss": 0.03850013,
+      "step": 19880
+    },
+    {
+      "epoch": 39.762,
+      "grad_norm": 1.3569062948226929,
+      "learning_rate": 2e-05,
+      "loss": 0.05348967,
+      "step": 19881
+    },
+    {
+      "epoch": 39.764,
+      "grad_norm": 2.510478973388672,
+      "learning_rate": 2e-05,
+      "loss": 0.05543127,
+      "step": 19882
+    },
+    {
+      "epoch": 39.766,
+      "grad_norm": 1.1178513765335083,
+      "learning_rate": 2e-05,
+      "loss": 0.05213483,
+      "step": 19883
+    },
+    {
+      "epoch": 39.768,
+      "grad_norm": 1.2882338762283325,
+      "learning_rate": 2e-05,
+      "loss": 0.06439535,
+      "step": 19884
+    },
+    {
+      "epoch": 39.77,
+      "grad_norm": 1.2819217443466187,
+      "learning_rate": 2e-05,
+      "loss": 0.06040879,
+      "step": 19885
+    },
+    {
+      "epoch": 39.772,
+      "grad_norm": 1.1934309005737305,
+      "learning_rate": 2e-05,
+      "loss": 0.04919674,
+      "step": 19886
+    },
+    {
+      "epoch": 39.774,
+      "grad_norm": 1.0528241395950317,
+      "learning_rate": 2e-05,
+      "loss": 0.03462794,
+      "step": 19887
+    },
+    {
+      "epoch": 39.776,
+      "grad_norm": 1.1237131357192993,
+      "learning_rate": 2e-05,
+      "loss": 0.03184299,
+      "step": 19888
+    },
+    {
+      "epoch": 39.778,
+      "grad_norm": 1.4120211601257324,
+      "learning_rate": 2e-05,
+      "loss": 0.05772817,
+      "step": 19889
+    },
+    {
+      "epoch": 39.78,
+      "grad_norm": 1.1867109537124634,
+      "learning_rate": 2e-05,
+      "loss": 0.04693531,
+      "step": 19890
+    },
+    {
+      "epoch": 39.782,
+      "grad_norm": 1.1656479835510254,
+      "learning_rate": 2e-05,
+      "loss": 0.03705209,
+      "step": 19891
+    },
+    {
+      "epoch": 39.784,
+      "grad_norm": 1.0704561471939087,
+      "learning_rate": 2e-05,
+      "loss": 0.04141489,
+      "step": 19892
+    },
+    {
+      "epoch": 39.786,
+      "grad_norm": 1.3049782514572144,
+      "learning_rate": 2e-05,
+      "loss": 0.06903137,
+      "step": 19893
+    },
+    {
+      "epoch": 39.788,
+      "grad_norm": 1.3069772720336914,
+      "learning_rate": 2e-05,
+      "loss": 0.0366349,
+      "step": 19894
+    },
+    {
+      "epoch": 39.79,
+      "grad_norm": 1.3293408155441284,
+      "learning_rate": 2e-05,
+      "loss": 0.05534213,
+      "step": 19895
+    },
+    {
+      "epoch": 39.792,
+      "grad_norm": 1.7149981260299683,
+      "learning_rate": 2e-05,
+      "loss": 0.06210932,
+      "step": 19896
+    },
+    {
+      "epoch": 39.794,
+      "grad_norm": 1.0320147275924683,
+      "learning_rate": 2e-05,
+      "loss": 0.04126735,
+      "step": 19897
+    },
+    {
+      "epoch": 39.796,
+      "grad_norm": 1.9870035648345947,
+      "learning_rate": 2e-05,
+      "loss": 0.05784594,
+      "step": 19898
+    },
+    {
+      "epoch": 39.798,
+      "grad_norm": 0.9793746471405029,
+      "learning_rate": 2e-05,
+      "loss": 0.03283355,
+      "step": 19899
+    },
+    {
+      "epoch": 39.8,
+      "grad_norm": 1.228532075881958,
+      "learning_rate": 2e-05,
+      "loss": 0.04221442,
+      "step": 19900
+    },
+    {
+      "epoch": 39.802,
+      "grad_norm": 1.2652114629745483,
+      "learning_rate": 2e-05,
+      "loss": 0.05898855,
+      "step": 19901
+    },
+    {
+      "epoch": 39.804,
+      "grad_norm": 1.1174923181533813,
+      "learning_rate": 2e-05,
+      "loss": 0.04276735,
+      "step": 19902
+    },
+    {
+      "epoch": 39.806,
+      "grad_norm": 1.0496095418930054,
+      "learning_rate": 2e-05,
+      "loss": 0.04212025,
+      "step": 19903
+    },
+    {
+      "epoch": 39.808,
+      "grad_norm": 1.0706851482391357,
+      "learning_rate": 2e-05,
+      "loss": 0.03774155,
+      "step": 19904
+    },
+    {
+      "epoch": 39.81,
+      "grad_norm": 1.1920790672302246,
+      "learning_rate": 2e-05,
+      "loss": 0.04768289,
+      "step": 19905
+    },
+    {
+      "epoch": 39.812,
+      "grad_norm": 1.502715826034546,
+      "learning_rate": 2e-05,
+      "loss": 0.04195279,
+      "step": 19906
+    },
+    {
+      "epoch": 39.814,
+      "grad_norm": 1.9300915002822876,
+      "learning_rate": 2e-05,
+      "loss": 0.04984208,
+      "step": 19907
+    },
+    {
+      "epoch": 39.816,
+      "grad_norm": 2.316269636154175,
+      "learning_rate": 2e-05,
+      "loss": 0.04897273,
+      "step": 19908
+    },
+    {
+      "epoch": 39.818,
+      "grad_norm": 1.250138759613037,
+      "learning_rate": 2e-05,
+      "loss": 0.04602588,
+      "step": 19909
+    },
+    {
+      "epoch": 39.82,
+      "grad_norm": 1.653391718864441,
+      "learning_rate": 2e-05,
+      "loss": 0.06034857,
+      "step": 19910
+    },
+    {
+      "epoch": 39.822,
+      "grad_norm": 1.1049057245254517,
+      "learning_rate": 2e-05,
+      "loss": 0.03826465,
+      "step": 19911
+    },
+    {
+      "epoch": 39.824,
+      "grad_norm": 1.1933602094650269,
+      "learning_rate": 2e-05,
+      "loss": 0.03310779,
+      "step": 19912
+    },
+    {
+      "epoch": 39.826,
+      "grad_norm": 3.675931930541992,
+      "learning_rate": 2e-05,
+      "loss": 0.05223902,
+      "step": 19913
+    },
+    {
+      "epoch": 39.828,
+      "grad_norm": 1.1655473709106445,
+      "learning_rate": 2e-05,
+      "loss": 0.0628321,
+      "step": 19914
+    },
+    {
+      "epoch": 39.83,
+      "grad_norm": 1.7604655027389526,
+      "learning_rate": 2e-05,
+      "loss": 0.06078061,
+      "step": 19915
+    },
+    {
+      "epoch": 39.832,
+      "grad_norm": 1.100632667541504,
+      "learning_rate": 2e-05,
+      "loss": 0.04568919,
+      "step": 19916
+    },
+    {
+      "epoch": 39.834,
+      "grad_norm": 1.2774721384048462,
+      "learning_rate": 2e-05,
+      "loss": 0.05435498,
+      "step": 19917
+    },
+    {
+      "epoch": 39.836,
+      "grad_norm": 1.217751145362854,
+      "learning_rate": 2e-05,
+      "loss": 0.05859322,
+      "step": 19918
+    },
+    {
+      "epoch": 39.838,
+      "grad_norm": 1.317592978477478,
+      "learning_rate": 2e-05,
+      "loss": 0.05665311,
+      "step": 19919
+    },
+    {
+      "epoch": 39.84,
+      "grad_norm": 0.949860692024231,
+      "learning_rate": 2e-05,
+      "loss": 0.02791356,
+      "step": 19920
+    },
+    {
+      "epoch": 39.842,
+      "grad_norm": 1.1154917478561401,
+      "learning_rate": 2e-05,
+      "loss": 0.05110167,
+      "step": 19921
+    },
+    {
+      "epoch": 39.844,
+      "grad_norm": 1.161690354347229,
+      "learning_rate": 2e-05,
+      "loss": 0.0497525,
+      "step": 19922
+    },
+    {
+      "epoch": 39.846,
+      "grad_norm": 1.190811038017273,
+      "learning_rate": 2e-05,
+      "loss": 0.0368493,
+      "step": 19923
+    },
+    {
+      "epoch": 39.848,
+      "grad_norm": 1.6861010789871216,
+      "learning_rate": 2e-05,
+      "loss": 0.05476474,
+      "step": 19924
+    },
+    {
+      "epoch": 39.85,
+      "grad_norm": 1.9949997663497925,
+      "learning_rate": 2e-05,
+      "loss": 0.0414875,
+      "step": 19925
+    },
+    {
+      "epoch": 39.852,
+      "grad_norm": 1.2739412784576416,
+      "learning_rate": 2e-05,
+      "loss": 0.05026732,
+      "step": 19926
+    },
+    {
+      "epoch": 39.854,
+      "grad_norm": 1.3758777379989624,
+      "learning_rate": 2e-05,
+      "loss": 0.06855834,
+      "step": 19927
+    },
+    {
+      "epoch": 39.856,
+      "grad_norm": 1.577846884727478,
+      "learning_rate": 2e-05,
+      "loss": 0.05498583,
+      "step": 19928
+    },
+    {
+      "epoch": 39.858,
+      "grad_norm": 1.269954800605774,
+      "learning_rate": 2e-05,
+      "loss": 0.0476332,
+      "step": 19929
+    },
+    {
+      "epoch": 39.86,
+      "grad_norm": 0.9598180055618286,
+      "learning_rate": 2e-05,
+      "loss": 0.04460917,
+      "step": 19930
+    },
+    {
+      "epoch": 39.862,
+      "grad_norm": 1.1671171188354492,
+      "learning_rate": 2e-05,
+      "loss": 0.04713274,
+      "step": 19931
+    },
+    {
+      "epoch": 39.864,
+      "grad_norm": 1.8716930150985718,
+      "learning_rate": 2e-05,
+      "loss": 0.0466135,
+      "step": 19932
+    },
+    {
+      "epoch": 39.866,
+      "grad_norm": 1.3691669702529907,
+      "learning_rate": 2e-05,
+      "loss": 0.04337586,
+      "step": 19933
+    },
+    {
+      "epoch": 39.868,
+      "grad_norm": 1.1902244091033936,
+      "learning_rate": 2e-05,
+      "loss": 0.04896775,
+      "step": 19934
+    },
+    {
+      "epoch": 39.87,
+      "grad_norm": 1.0358586311340332,
+      "learning_rate": 2e-05,
+      "loss": 0.04044461,
+      "step": 19935
+    },
+    {
+      "epoch": 39.872,
+      "grad_norm": 1.4702568054199219,
+      "learning_rate": 2e-05,
+      "loss": 0.06271239,
+      "step": 19936
+    },
+    {
+      "epoch": 39.874,
+      "grad_norm": 1.0855350494384766,
+      "learning_rate": 2e-05,
+      "loss": 0.03931961,
+      "step": 19937
+    },
+    {
+      "epoch": 39.876,
+      "grad_norm": 1.1319669485092163,
+      "learning_rate": 2e-05,
+      "loss": 0.03855097,
+      "step": 19938
+    },
+    {
+      "epoch": 39.878,
+      "grad_norm": 1.354026436805725,
+      "learning_rate": 2e-05,
+      "loss": 0.06375487,
+      "step": 19939
+    },
+    {
+      "epoch": 39.88,
+      "grad_norm": 0.9435744285583496,
+      "learning_rate": 2e-05,
+      "loss": 0.03325384,
+      "step": 19940
+    },
+    {
+      "epoch": 39.882,
+      "grad_norm": 1.2593286037445068,
+      "learning_rate": 2e-05,
+      "loss": 0.05153989,
+      "step": 19941
+    },
+    {
+      "epoch": 39.884,
+      "grad_norm": 0.9804494976997375,
+      "learning_rate": 2e-05,
+      "loss": 0.03355004,
+      "step": 19942
+    },
+    {
+      "epoch": 39.886,
+      "grad_norm": 1.18596351146698,
+      "learning_rate": 2e-05,
+      "loss": 0.0458352,
+      "step": 19943
+    },
+    {
+      "epoch": 39.888,
+      "grad_norm": 1.5126147270202637,
+      "learning_rate": 2e-05,
+      "loss": 0.05285957,
+      "step": 19944
+    },
+    {
+      "epoch": 39.89,
+      "grad_norm": 1.1133012771606445,
+      "learning_rate": 2e-05,
+      "loss": 0.03969733,
+      "step": 19945
+    },
+    {
+      "epoch": 39.892,
+      "grad_norm": 2.6804747581481934,
+      "learning_rate": 2e-05,
+      "loss": 0.06336376,
+      "step": 19946
+    },
+    {
+      "epoch": 39.894,
+      "grad_norm": 3.003247022628784,
+      "learning_rate": 2e-05,
+      "loss": 0.04598739,
+      "step": 19947
+    },
+    {
+      "epoch": 39.896,
+      "grad_norm": 1.1411057710647583,
+      "learning_rate": 2e-05,
+      "loss": 0.03566032,
+      "step": 19948
+    },
+    {
+      "epoch": 39.898,
+      "grad_norm": 1.0855693817138672,
+      "learning_rate": 2e-05,
+      "loss": 0.03875684,
+      "step": 19949
+    },
+    {
+      "epoch": 39.9,
+      "grad_norm": 1.9709842205047607,
+      "learning_rate": 2e-05,
+      "loss": 0.05917383,
+      "step": 19950
+    },
+    {
+      "epoch": 39.902,
+      "grad_norm": 1.1566156148910522,
+      "learning_rate": 2e-05,
+      "loss": 0.05174723,
+      "step": 19951
+    },
+    {
+      "epoch": 39.904,
+      "grad_norm": 1.0482864379882812,
+      "learning_rate": 2e-05,
+      "loss": 0.04446269,
+      "step": 19952
+    },
+    {
+      "epoch": 39.906,
+      "grad_norm": 1.2975873947143555,
+      "learning_rate": 2e-05,
+      "loss": 0.04888238,
+      "step": 19953
+    },
+    {
+      "epoch": 39.908,
+      "grad_norm": 1.3561941385269165,
+      "learning_rate": 2e-05,
+      "loss": 0.05117349,
+      "step": 19954
+    },
+    {
+      "epoch": 39.91,
+      "grad_norm": 0.9680836200714111,
+      "learning_rate": 2e-05,
+      "loss": 0.03723207,
+      "step": 19955
+    },
+    {
+      "epoch": 39.912,
+      "grad_norm": 1.2413675785064697,
+      "learning_rate": 2e-05,
+      "loss": 0.05808984,
+      "step": 19956
+    },
+    {
+      "epoch": 39.914,
+      "grad_norm": 1.634905219078064,
+      "learning_rate": 2e-05,
+      "loss": 0.05397768,
+      "step": 19957
+    },
+    {
+      "epoch": 39.916,
+      "grad_norm": 1.1176763772964478,
+      "learning_rate": 2e-05,
+      "loss": 0.04666106,
+      "step": 19958
+    },
+    {
+      "epoch": 39.918,
+      "grad_norm": 1.37368643283844,
+      "learning_rate": 2e-05,
+      "loss": 0.04420487,
+      "step": 19959
+    },
+    {
+      "epoch": 39.92,
+      "grad_norm": 1.1115273237228394,
+      "learning_rate": 2e-05,
+      "loss": 0.04129547,
+      "step": 19960
+    },
+    {
+      "epoch": 39.922,
+      "grad_norm": 1.0844300985336304,
+      "learning_rate": 2e-05,
+      "loss": 0.02857055,
+      "step": 19961
+    },
+    {
+      "epoch": 39.924,
+      "grad_norm": 2.0654237270355225,
+      "learning_rate": 2e-05,
+      "loss": 0.05185968,
+      "step": 19962
+    },
+    {
+      "epoch": 39.926,
+      "grad_norm": 1.4213045835494995,
+      "learning_rate": 2e-05,
+      "loss": 0.03576629,
+      "step": 19963
+    },
+    {
+      "epoch": 39.928,
+      "grad_norm": 1.399935007095337,
+      "learning_rate": 2e-05,
+      "loss": 0.05327077,
+      "step": 19964
+    },
+    {
+      "epoch": 39.93,
+      "grad_norm": 1.1581571102142334,
+      "learning_rate": 2e-05,
+      "loss": 0.04553617,
+      "step": 19965
+    },
+    {
+      "epoch": 39.932,
+      "grad_norm": 2.8895864486694336,
+      "learning_rate": 2e-05,
+      "loss": 0.0493044,
+      "step": 19966
+    },
+    {
+      "epoch": 39.934,
+      "grad_norm": 1.4619674682617188,
+      "learning_rate": 2e-05,
+      "loss": 0.04714632,
+      "step": 19967
+    },
+    {
+      "epoch": 39.936,
+      "grad_norm": 1.8203679323196411,
+      "learning_rate": 2e-05,
+      "loss": 0.06075837,
+      "step": 19968
+    },
+    {
+      "epoch": 39.938,
+      "grad_norm": 1.141274333000183,
+      "learning_rate": 2e-05,
+      "loss": 0.04157832,
+      "step": 19969
+    },
+    {
+      "epoch": 39.94,
+      "grad_norm": 1.180153727531433,
+      "learning_rate": 2e-05,
+      "loss": 0.05030315,
+      "step": 19970
+    },
+    {
+      "epoch": 39.942,
+      "grad_norm": 1.2318646907806396,
+      "learning_rate": 2e-05,
+      "loss": 0.04882877,
+      "step": 19971
+    },
+    {
+      "epoch": 39.944,
+      "grad_norm": 1.3025565147399902,
+      "learning_rate": 2e-05,
+      "loss": 0.0572556,
+      "step": 19972
+    },
+    {
+      "epoch": 39.946,
+      "grad_norm": 1.211354374885559,
+      "learning_rate": 2e-05,
+      "loss": 0.04019732,
+      "step": 19973
+    },
+    {
+      "epoch": 39.948,
+      "grad_norm": 1.2427458763122559,
+      "learning_rate": 2e-05,
+      "loss": 0.05008765,
+      "step": 19974
+    },
+    {
+      "epoch": 39.95,
+      "grad_norm": 1.2193975448608398,
+      "learning_rate": 2e-05,
+      "loss": 0.05430697,
+      "step": 19975
+    },
+    {
+      "epoch": 39.952,
+      "grad_norm": 1.3001223802566528,
+      "learning_rate": 2e-05,
+      "loss": 0.05283924,
+      "step": 19976
+    },
+    {
+      "epoch": 39.954,
+      "grad_norm": 1.1963660717010498,
+      "learning_rate": 2e-05,
+      "loss": 0.04262936,
+      "step": 19977
+    },
+    {
+      "epoch": 39.956,
+      "grad_norm": 1.0662003755569458,
+      "learning_rate": 2e-05,
+      "loss": 0.02899798,
+      "step": 19978
+    },
+    {
+      "epoch": 39.958,
+      "grad_norm": 1.1126896142959595,
+      "learning_rate": 2e-05,
+      "loss": 0.04382278,
+      "step": 19979
+    },
+    {
+      "epoch": 39.96,
+      "grad_norm": 1.4259101152420044,
+      "learning_rate": 2e-05,
+      "loss": 0.04453399,
+      "step": 19980
+    },
+    {
+      "epoch": 39.962,
+      "grad_norm": 1.2776509523391724,
+      "learning_rate": 2e-05,
+      "loss": 0.05190249,
+      "step": 19981
+    },
+    {
+      "epoch": 39.964,
+      "grad_norm": 1.6215944290161133,
+      "learning_rate": 2e-05,
+      "loss": 0.04128952,
+      "step": 19982
+    },
+    {
+      "epoch": 39.966,
+      "grad_norm": 1.6409664154052734,
+      "learning_rate": 2e-05,
+      "loss": 0.06073495,
+      "step": 19983
+    },
+    {
+      "epoch": 39.968,
+      "grad_norm": 1.4451102018356323,
+      "learning_rate": 2e-05,
+      "loss": 0.04271209,
+      "step": 19984
+    },
+    {
+      "epoch": 39.97,
+      "grad_norm": 1.8261207342147827,
+      "learning_rate": 2e-05,
+      "loss": 0.04644296,
+      "step": 19985
+    },
+    {
+      "epoch": 39.972,
+      "grad_norm": 1.125677466392517,
+      "learning_rate": 2e-05,
+      "loss": 0.0407895,
+      "step": 19986
+    },
+    {
+      "epoch": 39.974,
+      "grad_norm": 1.1863775253295898,
+      "learning_rate": 2e-05,
+      "loss": 0.04969215,
+      "step": 19987
+    },
+    {
+      "epoch": 39.976,
+      "grad_norm": 1.3171180486679077,
+      "learning_rate": 2e-05,
+      "loss": 0.04542191,
+      "step": 19988
+    },
+    {
+      "epoch": 39.978,
+      "grad_norm": 1.259590983390808,
+      "learning_rate": 2e-05,
+      "loss": 0.04899602,
+      "step": 19989
+    },
+    {
+      "epoch": 39.98,
+      "grad_norm": 1.566371202468872,
+      "learning_rate": 2e-05,
+      "loss": 0.05686514,
+      "step": 19990
+    },
+    {
+      "epoch": 39.982,
+      "grad_norm": 1.311056137084961,
+      "learning_rate": 2e-05,
+      "loss": 0.05832597,
+      "step": 19991
+    },
+    {
+      "epoch": 39.984,
+      "grad_norm": 1.8373768329620361,
+      "learning_rate": 2e-05,
+      "loss": 0.06506597,
+      "step": 19992
+    },
+    {
+      "epoch": 39.986,
+      "grad_norm": 1.1901731491088867,
+      "learning_rate": 2e-05,
+      "loss": 0.04623218,
+      "step": 19993
+    },
+    {
+      "epoch": 39.988,
+      "grad_norm": 1.0414644479751587,
+      "learning_rate": 2e-05,
+      "loss": 0.034441,
+      "step": 19994
+    },
+    {
+      "epoch": 39.99,
+      "grad_norm": 1.188096284866333,
+      "learning_rate": 2e-05,
+      "loss": 0.04996618,
+      "step": 19995
+    },
+    {
+      "epoch": 39.992,
+      "grad_norm": 0.9998059868812561,
+      "learning_rate": 2e-05,
+      "loss": 0.03086119,
+      "step": 19996
+    },
+    {
+      "epoch": 39.994,
+      "grad_norm": 1.2495959997177124,
+      "learning_rate": 2e-05,
+      "loss": 0.04241527,
+      "step": 19997
+    },
+    {
+      "epoch": 39.996,
+      "grad_norm": 1.2383801937103271,
+      "learning_rate": 2e-05,
+      "loss": 0.049159,
+      "step": 19998
+    },
+    {
+      "epoch": 39.998,
+      "grad_norm": 3.58790922164917,
+      "learning_rate": 2e-05,
+      "loss": 0.05261368,
+      "step": 19999
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 2.5835986137390137,
+      "learning_rate": 2e-05,
+      "loss": 0.05580256,
+      "step": 20000
+    },
+    {
+      "epoch": 40.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9840319361277445,
+        "Equal_1": 1.0,
+        "Equal_2": 0.9880239520958084,
+        "Equal_3": 0.9940119760479041,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.998003992015968,
+        "Parallel_1": 0.9959919839679359,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.992,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.992,
+        "Perpendicular_3": 0.9018036072144289,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 1.0,
+        "PointLiesOnCircle_3": 0.99,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9860279441117764
+      },
+      "eval_runtime": 319.7794,
+      "eval_samples_per_second": 32.835,
+      "eval_steps_per_second": 0.657,
+      "step": 20000
+    },
+    {
+      "epoch": 40.002,
+      "grad_norm": 1.5618497133255005,
+      "learning_rate": 2e-05,
+      "loss": 0.05388392,
+      "step": 20001
+    },
+    {
+      "epoch": 40.004,
+      "grad_norm": 1.2499672174453735,
+      "learning_rate": 2e-05,
+      "loss": 0.04120894,
+      "step": 20002
+    },
+    {
+      "epoch": 40.006,
+      "grad_norm": 1.2109254598617554,
+      "learning_rate": 2e-05,
+      "loss": 0.03839901,
+      "step": 20003
+    },
+    {
+      "epoch": 40.008,
+      "grad_norm": 1.3738662004470825,
+      "learning_rate": 2e-05,
+      "loss": 0.05607092,
+      "step": 20004
+    },
+    {
+      "epoch": 40.01,
+      "grad_norm": 1.273634672164917,
+      "learning_rate": 2e-05,
+      "loss": 0.05029005,
+      "step": 20005
+    },
+    {
+      "epoch": 40.012,
+      "grad_norm": 1.2355257272720337,
+      "learning_rate": 2e-05,
+      "loss": 0.04626942,
+      "step": 20006
+    },
+    {
+      "epoch": 40.014,
+      "grad_norm": 1.1767611503601074,
+      "learning_rate": 2e-05,
+      "loss": 0.04972693,
+      "step": 20007
+    },
+    {
+      "epoch": 40.016,
+      "grad_norm": 1.1207237243652344,
+      "learning_rate": 2e-05,
+      "loss": 0.03946103,
+      "step": 20008
+    },
+    {
+      "epoch": 40.018,
+      "grad_norm": 2.209026336669922,
+      "learning_rate": 2e-05,
+      "loss": 0.05307207,
+      "step": 20009
+    },
+    {
+      "epoch": 40.02,
+      "grad_norm": 1.8616058826446533,
+      "learning_rate": 2e-05,
+      "loss": 0.05713968,
+      "step": 20010
+    },
+    {
+      "epoch": 40.022,
+      "grad_norm": 1.272833228111267,
+      "learning_rate": 2e-05,
+      "loss": 0.05387183,
+      "step": 20011
+    },
+    {
+      "epoch": 40.024,
+      "grad_norm": 1.3753788471221924,
+      "learning_rate": 2e-05,
+      "loss": 0.04635529,
+      "step": 20012
+    },
+    {
+      "epoch": 40.026,
+      "grad_norm": 1.1166011095046997,
+      "learning_rate": 2e-05,
+      "loss": 0.03984597,
+      "step": 20013
+    },
+    {
+      "epoch": 40.028,
+      "grad_norm": 1.2804628610610962,
+      "learning_rate": 2e-05,
+      "loss": 0.04099889,
+      "step": 20014
+    },
+    {
+      "epoch": 40.03,
+      "grad_norm": 1.6008166074752808,
+      "learning_rate": 2e-05,
+      "loss": 0.04193804,
+      "step": 20015
+    },
+    {
+      "epoch": 40.032,
+      "grad_norm": 1.178210735321045,
+      "learning_rate": 2e-05,
+      "loss": 0.05431996,
+      "step": 20016
+    },
+    {
+      "epoch": 40.034,
+      "grad_norm": 1.2867320775985718,
+      "learning_rate": 2e-05,
+      "loss": 0.04413334,
+      "step": 20017
+    },
+    {
+      "epoch": 40.036,
+      "grad_norm": 1.4618786573410034,
+      "learning_rate": 2e-05,
+      "loss": 0.04305667,
+      "step": 20018
+    },
+    {
+      "epoch": 40.038,
+      "grad_norm": 1.2682616710662842,
+      "learning_rate": 2e-05,
+      "loss": 0.04806349,
+      "step": 20019
+    },
+    {
+      "epoch": 40.04,
+      "grad_norm": 1.1537147760391235,
+      "learning_rate": 2e-05,
+      "loss": 0.0417571,
+      "step": 20020
+    },
+    {
+      "epoch": 40.042,
+      "grad_norm": 0.8992504477500916,
+      "learning_rate": 2e-05,
+      "loss": 0.03431882,
+      "step": 20021
+    },
+    {
+      "epoch": 40.044,
+      "grad_norm": 2.1597468852996826,
+      "learning_rate": 2e-05,
+      "loss": 0.05409172,
+      "step": 20022
+    },
+    {
+      "epoch": 40.046,
+      "grad_norm": 1.5585830211639404,
+      "learning_rate": 2e-05,
+      "loss": 0.0792776,
+      "step": 20023
+    },
+    {
+      "epoch": 40.048,
+      "grad_norm": 1.5551427602767944,
+      "learning_rate": 2e-05,
+      "loss": 0.05392529,
+      "step": 20024
+    },
+    {
+      "epoch": 40.05,
+      "grad_norm": 1.1377781629562378,
+      "learning_rate": 2e-05,
+      "loss": 0.04901553,
+      "step": 20025
+    },
+    {
+      "epoch": 40.052,
+      "grad_norm": 1.1749012470245361,
+      "learning_rate": 2e-05,
+      "loss": 0.04288197,
+      "step": 20026
+    },
+    {
+      "epoch": 40.054,
+      "grad_norm": 1.2844057083129883,
+      "learning_rate": 2e-05,
+      "loss": 0.05152037,
+      "step": 20027
+    },
+    {
+      "epoch": 40.056,
+      "grad_norm": 0.8824015855789185,
+      "learning_rate": 2e-05,
+      "loss": 0.02397311,
+      "step": 20028
+    },
+    {
+      "epoch": 40.058,
+      "grad_norm": 1.6838172674179077,
+      "learning_rate": 2e-05,
+      "loss": 0.05278652,
+      "step": 20029
+    },
+    {
+      "epoch": 40.06,
+      "grad_norm": 1.0910221338272095,
+      "learning_rate": 2e-05,
+      "loss": 0.03923189,
+      "step": 20030
+    },
+    {
+      "epoch": 40.062,
+      "grad_norm": 1.179529070854187,
+      "learning_rate": 2e-05,
+      "loss": 0.04087804,
+      "step": 20031
+    },
+    {
+      "epoch": 40.064,
+      "grad_norm": 1.462498664855957,
+      "learning_rate": 2e-05,
+      "loss": 0.05278556,
+      "step": 20032
+    },
+    {
+      "epoch": 40.066,
+      "grad_norm": 1.1131311655044556,
+      "learning_rate": 2e-05,
+      "loss": 0.04452292,
+      "step": 20033
+    },
+    {
+      "epoch": 40.068,
+      "grad_norm": 1.4193816184997559,
+      "learning_rate": 2e-05,
+      "loss": 0.05834501,
+      "step": 20034
+    },
+    {
+      "epoch": 40.07,
+      "grad_norm": 1.2489386796951294,
+      "learning_rate": 2e-05,
+      "loss": 0.04075392,
+      "step": 20035
+    },
+    {
+      "epoch": 40.072,
+      "grad_norm": 1.0575319528579712,
+      "learning_rate": 2e-05,
+      "loss": 0.0387546,
+      "step": 20036
+    },
+    {
+      "epoch": 40.074,
+      "grad_norm": 1.1875722408294678,
+      "learning_rate": 2e-05,
+      "loss": 0.04389498,
+      "step": 20037
+    },
+    {
+      "epoch": 40.076,
+      "grad_norm": 1.1869674921035767,
+      "learning_rate": 2e-05,
+      "loss": 0.02951438,
+      "step": 20038
+    },
+    {
+      "epoch": 40.078,
+      "grad_norm": 2.2647807598114014,
+      "learning_rate": 2e-05,
+      "loss": 0.06623721,
+      "step": 20039
+    },
+    {
+      "epoch": 40.08,
+      "grad_norm": 1.8764697313308716,
+      "learning_rate": 2e-05,
+      "loss": 0.04736568,
+      "step": 20040
+    },
+    {
+      "epoch": 40.082,
+      "grad_norm": 1.6881517171859741,
+      "learning_rate": 2e-05,
+      "loss": 0.05365419,
+      "step": 20041
+    },
+    {
+      "epoch": 40.084,
+      "grad_norm": 1.8008352518081665,
+      "learning_rate": 2e-05,
+      "loss": 0.06396797,
+      "step": 20042
+    },
+    {
+      "epoch": 40.086,
+      "grad_norm": 1.4374672174453735,
+      "learning_rate": 2e-05,
+      "loss": 0.06762437,
+      "step": 20043
+    },
+    {
+      "epoch": 40.088,
+      "grad_norm": 1.4115700721740723,
+      "learning_rate": 2e-05,
+      "loss": 0.04840955,
+      "step": 20044
+    },
+    {
+      "epoch": 40.09,
+      "grad_norm": 1.0511454343795776,
+      "learning_rate": 2e-05,
+      "loss": 0.03845333,
+      "step": 20045
+    },
+    {
+      "epoch": 40.092,
+      "grad_norm": 1.0274492502212524,
+      "learning_rate": 2e-05,
+      "loss": 0.03665432,
+      "step": 20046
+    },
+    {
+      "epoch": 40.094,
+      "grad_norm": 2.0367865562438965,
+      "learning_rate": 2e-05,
+      "loss": 0.04981969,
+      "step": 20047
+    },
+    {
+      "epoch": 40.096,
+      "grad_norm": 1.2091890573501587,
+      "learning_rate": 2e-05,
+      "loss": 0.0455159,
+      "step": 20048
+    },
+    {
+      "epoch": 40.098,
+      "grad_norm": 1.1141090393066406,
+      "learning_rate": 2e-05,
+      "loss": 0.04782473,
+      "step": 20049
+    },
+    {
+      "epoch": 40.1,
+      "grad_norm": 1.4072805643081665,
+      "learning_rate": 2e-05,
+      "loss": 0.05027459,
+      "step": 20050
+    },
+    {
+      "epoch": 40.102,
+      "grad_norm": 1.0128663778305054,
+      "learning_rate": 2e-05,
+      "loss": 0.03786948,
+      "step": 20051
+    },
+    {
+      "epoch": 40.104,
+      "grad_norm": 1.241088628768921,
+      "learning_rate": 2e-05,
+      "loss": 0.04034789,
+      "step": 20052
+    },
+    {
+      "epoch": 40.106,
+      "grad_norm": 2.3295323848724365,
+      "learning_rate": 2e-05,
+      "loss": 0.04255938,
+      "step": 20053
+    },
+    {
+      "epoch": 40.108,
+      "grad_norm": 1.2355490922927856,
+      "learning_rate": 2e-05,
+      "loss": 0.04925885,
+      "step": 20054
+    },
+    {
+      "epoch": 40.11,
+      "grad_norm": 1.342072606086731,
+      "learning_rate": 2e-05,
+      "loss": 0.05003198,
+      "step": 20055
+    },
+    {
+      "epoch": 40.112,
+      "grad_norm": 1.1754595041275024,
+      "learning_rate": 2e-05,
+      "loss": 0.04749743,
+      "step": 20056
+    },
+    {
+      "epoch": 40.114,
+      "grad_norm": 1.235671043395996,
+      "learning_rate": 2e-05,
+      "loss": 0.05393402,
+      "step": 20057
+    },
+    {
+      "epoch": 40.116,
+      "grad_norm": 1.2485733032226562,
+      "learning_rate": 2e-05,
+      "loss": 0.0627479,
+      "step": 20058
+    },
+    {
+      "epoch": 40.118,
+      "grad_norm": 1.5328861474990845,
+      "learning_rate": 2e-05,
+      "loss": 0.05113653,
+      "step": 20059
+    },
+    {
+      "epoch": 40.12,
+      "grad_norm": 1.499184489250183,
+      "learning_rate": 2e-05,
+      "loss": 0.05487531,
+      "step": 20060
+    },
+    {
+      "epoch": 40.122,
+      "grad_norm": 2.340745449066162,
+      "learning_rate": 2e-05,
+      "loss": 0.05198453,
+      "step": 20061
+    },
+    {
+      "epoch": 40.124,
+      "grad_norm": 1.106819748878479,
+      "learning_rate": 2e-05,
+      "loss": 0.0406749,
+      "step": 20062
+    },
+    {
+      "epoch": 40.126,
+      "grad_norm": 0.9359315037727356,
+      "learning_rate": 2e-05,
+      "loss": 0.02896912,
+      "step": 20063
+    },
+    {
+      "epoch": 40.128,
+      "grad_norm": 1.3963433504104614,
+      "learning_rate": 2e-05,
+      "loss": 0.05835143,
+      "step": 20064
+    },
+    {
+      "epoch": 40.13,
+      "grad_norm": 1.1196558475494385,
+      "learning_rate": 2e-05,
+      "loss": 0.04942945,
+      "step": 20065
+    },
+    {
+      "epoch": 40.132,
+      "grad_norm": 1.1688649654388428,
+      "learning_rate": 2e-05,
+      "loss": 0.05483418,
+      "step": 20066
+    },
+    {
+      "epoch": 40.134,
+      "grad_norm": 1.5023531913757324,
+      "learning_rate": 2e-05,
+      "loss": 0.05200912,
+      "step": 20067
+    },
+    {
+      "epoch": 40.136,
+      "grad_norm": 1.1763683557510376,
+      "learning_rate": 2e-05,
+      "loss": 0.04295605,
+      "step": 20068
+    },
+    {
+      "epoch": 40.138,
+      "grad_norm": 1.2277802228927612,
+      "learning_rate": 2e-05,
+      "loss": 0.04733496,
+      "step": 20069
+    },
+    {
+      "epoch": 40.14,
+      "grad_norm": 1.6241992712020874,
+      "learning_rate": 2e-05,
+      "loss": 0.06077222,
+      "step": 20070
+    },
+    {
+      "epoch": 40.142,
+      "grad_norm": 2.3605220317840576,
+      "learning_rate": 2e-05,
+      "loss": 0.05581881,
+      "step": 20071
+    },
+    {
+      "epoch": 40.144,
+      "grad_norm": 1.1370664834976196,
+      "learning_rate": 2e-05,
+      "loss": 0.04241094,
+      "step": 20072
+    },
+    {
+      "epoch": 40.146,
+      "grad_norm": 1.6637158393859863,
+      "learning_rate": 2e-05,
+      "loss": 0.05778241,
+      "step": 20073
+    },
+    {
+      "epoch": 40.148,
+      "grad_norm": 1.4280550479888916,
+      "learning_rate": 2e-05,
+      "loss": 0.05228476,
+      "step": 20074
+    },
+    {
+      "epoch": 40.15,
+      "grad_norm": 1.5583044290542603,
+      "learning_rate": 2e-05,
+      "loss": 0.05583321,
+      "step": 20075
+    },
+    {
+      "epoch": 40.152,
+      "grad_norm": 1.119242548942566,
+      "learning_rate": 2e-05,
+      "loss": 0.04524466,
+      "step": 20076
+    },
+    {
+      "epoch": 40.154,
+      "grad_norm": 1.4672080278396606,
+      "learning_rate": 2e-05,
+      "loss": 0.06838915,
+      "step": 20077
+    },
+    {
+      "epoch": 40.156,
+      "grad_norm": 2.8078713417053223,
+      "learning_rate": 2e-05,
+      "loss": 0.06085558,
+      "step": 20078
+    },
+    {
+      "epoch": 40.158,
+      "grad_norm": 7.261538028717041,
+      "learning_rate": 2e-05,
+      "loss": 0.04814359,
+      "step": 20079
+    },
+    {
+      "epoch": 40.16,
+      "grad_norm": 1.254329800605774,
+      "learning_rate": 2e-05,
+      "loss": 0.04063053,
+      "step": 20080
+    },
+    {
+      "epoch": 40.162,
+      "grad_norm": 1.3946176767349243,
+      "learning_rate": 2e-05,
+      "loss": 0.04426568,
+      "step": 20081
+    },
+    {
+      "epoch": 40.164,
+      "grad_norm": 1.089093565940857,
+      "learning_rate": 2e-05,
+      "loss": 0.04132479,
+      "step": 20082
+    },
+    {
+      "epoch": 40.166,
+      "grad_norm": 1.2767221927642822,
+      "learning_rate": 2e-05,
+      "loss": 0.05414563,
+      "step": 20083
+    },
+    {
+      "epoch": 40.168,
+      "grad_norm": 1.287773609161377,
+      "learning_rate": 2e-05,
+      "loss": 0.04473702,
+      "step": 20084
+    },
+    {
+      "epoch": 40.17,
+      "grad_norm": 0.8289897441864014,
+      "learning_rate": 2e-05,
+      "loss": 0.03129639,
+      "step": 20085
+    },
+    {
+      "epoch": 40.172,
+      "grad_norm": 1.047951579093933,
+      "learning_rate": 2e-05,
+      "loss": 0.03783755,
+      "step": 20086
+    },
+    {
+      "epoch": 40.174,
+      "grad_norm": 0.9991368651390076,
+      "learning_rate": 2e-05,
+      "loss": 0.03878902,
+      "step": 20087
+    },
+    {
+      "epoch": 40.176,
+      "grad_norm": 1.249922275543213,
+      "learning_rate": 2e-05,
+      "loss": 0.04872551,
+      "step": 20088
+    },
+    {
+      "epoch": 40.178,
+      "grad_norm": 1.3479992151260376,
+      "learning_rate": 2e-05,
+      "loss": 0.07139233,
+      "step": 20089
+    },
+    {
+      "epoch": 40.18,
+      "grad_norm": 2.5222294330596924,
+      "learning_rate": 2e-05,
+      "loss": 0.05283742,
+      "step": 20090
+    },
+    {
+      "epoch": 40.182,
+      "grad_norm": 1.0991319417953491,
+      "learning_rate": 2e-05,
+      "loss": 0.05567234,
+      "step": 20091
+    },
+    {
+      "epoch": 40.184,
+      "grad_norm": 1.6871635913848877,
+      "learning_rate": 2e-05,
+      "loss": 0.08690213,
+      "step": 20092
+    },
+    {
+      "epoch": 40.186,
+      "grad_norm": 1.2134432792663574,
+      "learning_rate": 2e-05,
+      "loss": 0.04679528,
+      "step": 20093
+    },
+    {
+      "epoch": 40.188,
+      "grad_norm": 1.456554889678955,
+      "learning_rate": 2e-05,
+      "loss": 0.05808542,
+      "step": 20094
+    },
+    {
+      "epoch": 40.19,
+      "grad_norm": 1.1095123291015625,
+      "learning_rate": 2e-05,
+      "loss": 0.04110887,
+      "step": 20095
+    },
+    {
+      "epoch": 40.192,
+      "grad_norm": 0.9948191046714783,
+      "learning_rate": 2e-05,
+      "loss": 0.0418423,
+      "step": 20096
+    },
+    {
+      "epoch": 40.194,
+      "grad_norm": 1.4417386054992676,
+      "learning_rate": 2e-05,
+      "loss": 0.06203289,
+      "step": 20097
+    },
+    {
+      "epoch": 40.196,
+      "grad_norm": 2.3071508407592773,
+      "learning_rate": 2e-05,
+      "loss": 0.04821993,
+      "step": 20098
+    },
+    {
+      "epoch": 40.198,
+      "grad_norm": 1.352673888206482,
+      "learning_rate": 2e-05,
+      "loss": 0.03600384,
+      "step": 20099
+    },
+    {
+      "epoch": 40.2,
+      "grad_norm": 1.105892300605774,
+      "learning_rate": 2e-05,
+      "loss": 0.03900053,
+      "step": 20100
+    },
+    {
+      "epoch": 40.202,
+      "grad_norm": 1.2061972618103027,
+      "learning_rate": 2e-05,
+      "loss": 0.05580106,
+      "step": 20101
+    },
+    {
+      "epoch": 40.204,
+      "grad_norm": 1.2827521562576294,
+      "learning_rate": 2e-05,
+      "loss": 0.04791928,
+      "step": 20102
+    },
+    {
+      "epoch": 40.206,
+      "grad_norm": 1.2111438512802124,
+      "learning_rate": 2e-05,
+      "loss": 0.04924633,
+      "step": 20103
+    },
+    {
+      "epoch": 40.208,
+      "grad_norm": 1.4191051721572876,
+      "learning_rate": 2e-05,
+      "loss": 0.04869965,
+      "step": 20104
+    },
+    {
+      "epoch": 40.21,
+      "grad_norm": 1.1339068412780762,
+      "learning_rate": 2e-05,
+      "loss": 0.05062132,
+      "step": 20105
+    },
+    {
+      "epoch": 40.212,
+      "grad_norm": 1.2514359951019287,
+      "learning_rate": 2e-05,
+      "loss": 0.04554267,
+      "step": 20106
+    },
+    {
+      "epoch": 40.214,
+      "grad_norm": 1.1411757469177246,
+      "learning_rate": 2e-05,
+      "loss": 0.04453161,
+      "step": 20107
+    },
+    {
+      "epoch": 40.216,
+      "grad_norm": 1.28267240524292,
+      "learning_rate": 2e-05,
+      "loss": 0.03762868,
+      "step": 20108
+    },
+    {
+      "epoch": 40.218,
+      "grad_norm": 1.2932469844818115,
+      "learning_rate": 2e-05,
+      "loss": 0.05375424,
+      "step": 20109
+    },
+    {
+      "epoch": 40.22,
+      "grad_norm": 1.5914359092712402,
+      "learning_rate": 2e-05,
+      "loss": 0.05069669,
+      "step": 20110
+    },
+    {
+      "epoch": 40.222,
+      "grad_norm": 1.1139858961105347,
+      "learning_rate": 2e-05,
+      "loss": 0.03766536,
+      "step": 20111
+    },
+    {
+      "epoch": 40.224,
+      "grad_norm": 1.3693923950195312,
+      "learning_rate": 2e-05,
+      "loss": 0.04846753,
+      "step": 20112
+    },
+    {
+      "epoch": 40.226,
+      "grad_norm": 1.1039000749588013,
+      "learning_rate": 2e-05,
+      "loss": 0.035409,
+      "step": 20113
+    },
+    {
+      "epoch": 40.228,
+      "grad_norm": 1.1846332550048828,
+      "learning_rate": 2e-05,
+      "loss": 0.04201246,
+      "step": 20114
+    },
+    {
+      "epoch": 40.23,
+      "grad_norm": 2.8767950534820557,
+      "learning_rate": 2e-05,
+      "loss": 0.06352147,
+      "step": 20115
+    },
+    {
+      "epoch": 40.232,
+      "grad_norm": 1.286635160446167,
+      "learning_rate": 2e-05,
+      "loss": 0.05608144,
+      "step": 20116
+    },
+    {
+      "epoch": 40.234,
+      "grad_norm": 1.0336881875991821,
+      "learning_rate": 2e-05,
+      "loss": 0.03256797,
+      "step": 20117
+    },
+    {
+      "epoch": 40.236,
+      "grad_norm": 8.691652297973633,
+      "learning_rate": 2e-05,
+      "loss": 0.0322794,
+      "step": 20118
+    },
+    {
+      "epoch": 40.238,
+      "grad_norm": 2.27980375289917,
+      "learning_rate": 2e-05,
+      "loss": 0.03699232,
+      "step": 20119
+    },
+    {
+      "epoch": 40.24,
+      "grad_norm": 1.9396880865097046,
+      "learning_rate": 2e-05,
+      "loss": 0.06942445,
+      "step": 20120
+    },
+    {
+      "epoch": 40.242,
+      "grad_norm": 1.0819038152694702,
+      "learning_rate": 2e-05,
+      "loss": 0.04012003,
+      "step": 20121
+    },
+    {
+      "epoch": 40.244,
+      "grad_norm": 1.3157330751419067,
+      "learning_rate": 2e-05,
+      "loss": 0.05315004,
+      "step": 20122
+    },
+    {
+      "epoch": 40.246,
+      "grad_norm": 1.216844081878662,
+      "learning_rate": 2e-05,
+      "loss": 0.05938418,
+      "step": 20123
+    },
+    {
+      "epoch": 40.248,
+      "grad_norm": 1.3893853425979614,
+      "learning_rate": 2e-05,
+      "loss": 0.04685809,
+      "step": 20124
+    },
+    {
+      "epoch": 40.25,
+      "grad_norm": 1.0636508464813232,
+      "learning_rate": 2e-05,
+      "loss": 0.03838153,
+      "step": 20125
+    },
+    {
+      "epoch": 40.252,
+      "grad_norm": 1.0984792709350586,
+      "learning_rate": 2e-05,
+      "loss": 0.05215897,
+      "step": 20126
+    },
+    {
+      "epoch": 40.254,
+      "grad_norm": 1.0352123975753784,
+      "learning_rate": 2e-05,
+      "loss": 0.04288472,
+      "step": 20127
+    },
+    {
+      "epoch": 40.256,
+      "grad_norm": 1.0150216817855835,
+      "learning_rate": 2e-05,
+      "loss": 0.03558073,
+      "step": 20128
+    },
+    {
+      "epoch": 40.258,
+      "grad_norm": 1.1401340961456299,
+      "learning_rate": 2e-05,
+      "loss": 0.03874261,
+      "step": 20129
+    },
+    {
+      "epoch": 40.26,
+      "grad_norm": 1.2564568519592285,
+      "learning_rate": 2e-05,
+      "loss": 0.05399378,
+      "step": 20130
+    },
+    {
+      "epoch": 40.262,
+      "grad_norm": 1.2055622339248657,
+      "learning_rate": 2e-05,
+      "loss": 0.05442786,
+      "step": 20131
+    },
+    {
+      "epoch": 40.264,
+      "grad_norm": 1.2611266374588013,
+      "learning_rate": 2e-05,
+      "loss": 0.04908681,
+      "step": 20132
+    },
+    {
+      "epoch": 40.266,
+      "grad_norm": 1.8219330310821533,
+      "learning_rate": 2e-05,
+      "loss": 0.05871871,
+      "step": 20133
+    },
+    {
+      "epoch": 40.268,
+      "grad_norm": 1.118198275566101,
+      "learning_rate": 2e-05,
+      "loss": 0.04319108,
+      "step": 20134
+    },
+    {
+      "epoch": 40.27,
+      "grad_norm": 1.0127724409103394,
+      "learning_rate": 2e-05,
+      "loss": 0.03908618,
+      "step": 20135
+    },
+    {
+      "epoch": 40.272,
+      "grad_norm": 1.1128853559494019,
+      "learning_rate": 2e-05,
+      "loss": 0.04767665,
+      "step": 20136
+    },
+    {
+      "epoch": 40.274,
+      "grad_norm": 1.198813557624817,
+      "learning_rate": 2e-05,
+      "loss": 0.04259463,
+      "step": 20137
+    },
+    {
+      "epoch": 40.276,
+      "grad_norm": 1.0490918159484863,
+      "learning_rate": 2e-05,
+      "loss": 0.02958579,
+      "step": 20138
+    },
+    {
+      "epoch": 40.278,
+      "grad_norm": 1.3297768831253052,
+      "learning_rate": 2e-05,
+      "loss": 0.05636571,
+      "step": 20139
+    },
+    {
+      "epoch": 40.28,
+      "grad_norm": 1.2086269855499268,
+      "learning_rate": 2e-05,
+      "loss": 0.04821461,
+      "step": 20140
+    },
+    {
+      "epoch": 40.282,
+      "grad_norm": 2.00396728515625,
+      "learning_rate": 2e-05,
+      "loss": 0.03132957,
+      "step": 20141
+    },
+    {
+      "epoch": 40.284,
+      "grad_norm": 1.4229713678359985,
+      "learning_rate": 2e-05,
+      "loss": 0.04370468,
+      "step": 20142
+    },
+    {
+      "epoch": 40.286,
+      "grad_norm": 1.6150267124176025,
+      "learning_rate": 2e-05,
+      "loss": 0.05704456,
+      "step": 20143
+    },
+    {
+      "epoch": 40.288,
+      "grad_norm": 0.9842348098754883,
+      "learning_rate": 2e-05,
+      "loss": 0.02401066,
+      "step": 20144
+    },
+    {
+      "epoch": 40.29,
+      "grad_norm": 1.246802806854248,
+      "learning_rate": 2e-05,
+      "loss": 0.03580172,
+      "step": 20145
+    },
+    {
+      "epoch": 40.292,
+      "grad_norm": 1.4223207235336304,
+      "learning_rate": 2e-05,
+      "loss": 0.04059513,
+      "step": 20146
+    },
+    {
+      "epoch": 40.294,
+      "grad_norm": 1.252277135848999,
+      "learning_rate": 2e-05,
+      "loss": 0.05555695,
+      "step": 20147
+    },
+    {
+      "epoch": 40.296,
+      "grad_norm": 1.1984466314315796,
+      "learning_rate": 2e-05,
+      "loss": 0.04715933,
+      "step": 20148
+    },
+    {
+      "epoch": 40.298,
+      "grad_norm": 1.1603848934173584,
+      "learning_rate": 2e-05,
+      "loss": 0.04525081,
+      "step": 20149
+    },
+    {
+      "epoch": 40.3,
+      "grad_norm": 1.5191564559936523,
+      "learning_rate": 2e-05,
+      "loss": 0.05356734,
+      "step": 20150
+    },
+    {
+      "epoch": 40.302,
+      "grad_norm": 2.3878793716430664,
+      "learning_rate": 2e-05,
+      "loss": 0.04827441,
+      "step": 20151
+    },
+    {
+      "epoch": 40.304,
+      "grad_norm": 1.3220070600509644,
+      "learning_rate": 2e-05,
+      "loss": 0.05512287,
+      "step": 20152
+    },
+    {
+      "epoch": 40.306,
+      "grad_norm": 1.7690134048461914,
+      "learning_rate": 2e-05,
+      "loss": 0.07210412,
+      "step": 20153
+    },
+    {
+      "epoch": 40.308,
+      "grad_norm": 3.268732786178589,
+      "learning_rate": 2e-05,
+      "loss": 0.05046576,
+      "step": 20154
+    },
+    {
+      "epoch": 40.31,
+      "grad_norm": 1.3176329135894775,
+      "learning_rate": 2e-05,
+      "loss": 0.04881255,
+      "step": 20155
+    },
+    {
+      "epoch": 40.312,
+      "grad_norm": 1.2548283338546753,
+      "learning_rate": 2e-05,
+      "loss": 0.04261008,
+      "step": 20156
+    },
+    {
+      "epoch": 40.314,
+      "grad_norm": 1.4517444372177124,
+      "learning_rate": 2e-05,
+      "loss": 0.04030664,
+      "step": 20157
+    },
+    {
+      "epoch": 40.316,
+      "grad_norm": 1.297654628753662,
+      "learning_rate": 2e-05,
+      "loss": 0.04864812,
+      "step": 20158
+    },
+    {
+      "epoch": 40.318,
+      "grad_norm": 1.3605537414550781,
+      "learning_rate": 2e-05,
+      "loss": 0.05398957,
+      "step": 20159
+    },
+    {
+      "epoch": 40.32,
+      "grad_norm": 1.0863500833511353,
+      "learning_rate": 2e-05,
+      "loss": 0.04536635,
+      "step": 20160
+    },
+    {
+      "epoch": 40.322,
+      "grad_norm": 1.3427700996398926,
+      "learning_rate": 2e-05,
+      "loss": 0.04969367,
+      "step": 20161
+    },
+    {
+      "epoch": 40.324,
+      "grad_norm": 1.6353856325149536,
+      "learning_rate": 2e-05,
+      "loss": 0.04320132,
+      "step": 20162
+    },
+    {
+      "epoch": 40.326,
+      "grad_norm": 1.2196747064590454,
+      "learning_rate": 2e-05,
+      "loss": 0.03800016,
+      "step": 20163
+    },
+    {
+      "epoch": 40.328,
+      "grad_norm": 1.7924127578735352,
+      "learning_rate": 2e-05,
+      "loss": 0.07237592,
+      "step": 20164
+    },
+    {
+      "epoch": 40.33,
+      "grad_norm": 1.7914211750030518,
+      "learning_rate": 2e-05,
+      "loss": 0.05794446,
+      "step": 20165
+    },
+    {
+      "epoch": 40.332,
+      "grad_norm": 1.2573678493499756,
+      "learning_rate": 2e-05,
+      "loss": 0.04247341,
+      "step": 20166
+    },
+    {
+      "epoch": 40.334,
+      "grad_norm": 1.1113066673278809,
+      "learning_rate": 2e-05,
+      "loss": 0.05276919,
+      "step": 20167
+    },
+    {
+      "epoch": 40.336,
+      "grad_norm": 1.5416284799575806,
+      "learning_rate": 2e-05,
+      "loss": 0.05891626,
+      "step": 20168
+    },
+    {
+      "epoch": 40.338,
+      "grad_norm": 1.1009063720703125,
+      "learning_rate": 2e-05,
+      "loss": 0.0362423,
+      "step": 20169
+    },
+    {
+      "epoch": 40.34,
+      "grad_norm": 1.2229732275009155,
+      "learning_rate": 2e-05,
+      "loss": 0.04161092,
+      "step": 20170
+    },
+    {
+      "epoch": 40.342,
+      "grad_norm": 1.4811763763427734,
+      "learning_rate": 2e-05,
+      "loss": 0.05466798,
+      "step": 20171
+    },
+    {
+      "epoch": 40.344,
+      "grad_norm": 1.1758745908737183,
+      "learning_rate": 2e-05,
+      "loss": 0.04076952,
+      "step": 20172
+    },
+    {
+      "epoch": 40.346,
+      "grad_norm": 1.35615873336792,
+      "learning_rate": 2e-05,
+      "loss": 0.0481509,
+      "step": 20173
+    },
+    {
+      "epoch": 40.348,
+      "grad_norm": 1.1503114700317383,
+      "learning_rate": 2e-05,
+      "loss": 0.03523769,
+      "step": 20174
+    },
+    {
+      "epoch": 40.35,
+      "grad_norm": 1.1574171781539917,
+      "learning_rate": 2e-05,
+      "loss": 0.03359126,
+      "step": 20175
+    },
+    {
+      "epoch": 40.352,
+      "grad_norm": 1.2408151626586914,
+      "learning_rate": 2e-05,
+      "loss": 0.05215861,
+      "step": 20176
+    },
+    {
+      "epoch": 40.354,
+      "grad_norm": 1.2449312210083008,
+      "learning_rate": 2e-05,
+      "loss": 0.05467416,
+      "step": 20177
+    },
+    {
+      "epoch": 40.356,
+      "grad_norm": 0.9955729246139526,
+      "learning_rate": 2e-05,
+      "loss": 0.03369155,
+      "step": 20178
+    },
+    {
+      "epoch": 40.358,
+      "grad_norm": 1.5078949928283691,
+      "learning_rate": 2e-05,
+      "loss": 0.0526335,
+      "step": 20179
+    },
+    {
+      "epoch": 40.36,
+      "grad_norm": 1.6468613147735596,
+      "learning_rate": 2e-05,
+      "loss": 0.0292394,
+      "step": 20180
+    },
+    {
+      "epoch": 40.362,
+      "grad_norm": 1.2313929796218872,
+      "learning_rate": 2e-05,
+      "loss": 0.05515527,
+      "step": 20181
+    },
+    {
+      "epoch": 40.364,
+      "grad_norm": 1.1853318214416504,
+      "learning_rate": 2e-05,
+      "loss": 0.04114341,
+      "step": 20182
+    },
+    {
+      "epoch": 40.366,
+      "grad_norm": 2.5944883823394775,
+      "learning_rate": 2e-05,
+      "loss": 0.04679396,
+      "step": 20183
+    },
+    {
+      "epoch": 40.368,
+      "grad_norm": 1.1291593313217163,
+      "learning_rate": 2e-05,
+      "loss": 0.03800917,
+      "step": 20184
+    },
+    {
+      "epoch": 40.37,
+      "grad_norm": 1.2536524534225464,
+      "learning_rate": 2e-05,
+      "loss": 0.03788491,
+      "step": 20185
+    },
+    {
+      "epoch": 40.372,
+      "grad_norm": 1.110937237739563,
+      "learning_rate": 2e-05,
+      "loss": 0.03799309,
+      "step": 20186
+    },
+    {
+      "epoch": 40.374,
+      "grad_norm": 1.246289610862732,
+      "learning_rate": 2e-05,
+      "loss": 0.05560606,
+      "step": 20187
+    },
+    {
+      "epoch": 40.376,
+      "grad_norm": 1.2249597311019897,
+      "learning_rate": 2e-05,
+      "loss": 0.05621621,
+      "step": 20188
+    },
+    {
+      "epoch": 40.378,
+      "grad_norm": 1.4832850694656372,
+      "learning_rate": 2e-05,
+      "loss": 0.05166864,
+      "step": 20189
+    },
+    {
+      "epoch": 40.38,
+      "grad_norm": 1.8920128345489502,
+      "learning_rate": 2e-05,
+      "loss": 0.06390877,
+      "step": 20190
+    },
+    {
+      "epoch": 40.382,
+      "grad_norm": 1.3039014339447021,
+      "learning_rate": 2e-05,
+      "loss": 0.04282786,
+      "step": 20191
+    },
+    {
+      "epoch": 40.384,
+      "grad_norm": 1.2546817064285278,
+      "learning_rate": 2e-05,
+      "loss": 0.05530909,
+      "step": 20192
+    },
+    {
+      "epoch": 40.386,
+      "grad_norm": 1.0137908458709717,
+      "learning_rate": 2e-05,
+      "loss": 0.03802532,
+      "step": 20193
+    },
+    {
+      "epoch": 40.388,
+      "grad_norm": 3.2582902908325195,
+      "learning_rate": 2e-05,
+      "loss": 0.04777801,
+      "step": 20194
+    },
+    {
+      "epoch": 40.39,
+      "grad_norm": 1.6945880651474,
+      "learning_rate": 2e-05,
+      "loss": 0.06353603,
+      "step": 20195
+    },
+    {
+      "epoch": 40.392,
+      "grad_norm": 1.861193299293518,
+      "learning_rate": 2e-05,
+      "loss": 0.0449325,
+      "step": 20196
+    },
+    {
+      "epoch": 40.394,
+      "grad_norm": 1.4629795551300049,
+      "learning_rate": 2e-05,
+      "loss": 0.06012824,
+      "step": 20197
+    },
+    {
+      "epoch": 40.396,
+      "grad_norm": 1.067794919013977,
+      "learning_rate": 2e-05,
+      "loss": 0.04181834,
+      "step": 20198
+    },
+    {
+      "epoch": 40.398,
+      "grad_norm": 1.1766245365142822,
+      "learning_rate": 2e-05,
+      "loss": 0.04275657,
+      "step": 20199
+    },
+    {
+      "epoch": 40.4,
+      "grad_norm": 3.7547171115875244,
+      "learning_rate": 2e-05,
+      "loss": 0.05930864,
+      "step": 20200
+    },
+    {
+      "epoch": 40.402,
+      "grad_norm": 1.0909967422485352,
+      "learning_rate": 2e-05,
+      "loss": 0.04474407,
+      "step": 20201
+    },
+    {
+      "epoch": 40.404,
+      "grad_norm": 1.1844147443771362,
+      "learning_rate": 2e-05,
+      "loss": 0.04869113,
+      "step": 20202
+    },
+    {
+      "epoch": 40.406,
+      "grad_norm": 0.9126821160316467,
+      "learning_rate": 2e-05,
+      "loss": 0.03085716,
+      "step": 20203
+    },
+    {
+      "epoch": 40.408,
+      "grad_norm": 1.0406267642974854,
+      "learning_rate": 2e-05,
+      "loss": 0.03293871,
+      "step": 20204
+    },
+    {
+      "epoch": 40.41,
+      "grad_norm": 0.9543370008468628,
+      "learning_rate": 2e-05,
+      "loss": 0.03391955,
+      "step": 20205
+    },
+    {
+      "epoch": 40.412,
+      "grad_norm": 0.8653976917266846,
+      "learning_rate": 2e-05,
+      "loss": 0.02381743,
+      "step": 20206
+    },
+    {
+      "epoch": 40.414,
+      "grad_norm": 1.078957200050354,
+      "learning_rate": 2e-05,
+      "loss": 0.04199693,
+      "step": 20207
+    },
+    {
+      "epoch": 40.416,
+      "grad_norm": 4.299551963806152,
+      "learning_rate": 2e-05,
+      "loss": 0.05353047,
+      "step": 20208
+    },
+    {
+      "epoch": 40.418,
+      "grad_norm": 1.4246419668197632,
+      "learning_rate": 2e-05,
+      "loss": 0.04532846,
+      "step": 20209
+    },
+    {
+      "epoch": 40.42,
+      "grad_norm": 1.2102235555648804,
+      "learning_rate": 2e-05,
+      "loss": 0.05503858,
+      "step": 20210
+    },
+    {
+      "epoch": 40.422,
+      "grad_norm": 1.1240041255950928,
+      "learning_rate": 2e-05,
+      "loss": 0.03953891,
+      "step": 20211
+    },
+    {
+      "epoch": 40.424,
+      "grad_norm": 1.3952480554580688,
+      "learning_rate": 2e-05,
+      "loss": 0.05917449,
+      "step": 20212
+    },
+    {
+      "epoch": 40.426,
+      "grad_norm": 1.2208077907562256,
+      "learning_rate": 2e-05,
+      "loss": 0.03653198,
+      "step": 20213
+    },
+    {
+      "epoch": 40.428,
+      "grad_norm": 1.1388483047485352,
+      "learning_rate": 2e-05,
+      "loss": 0.04844316,
+      "step": 20214
+    },
+    {
+      "epoch": 40.43,
+      "grad_norm": 1.294251561164856,
+      "learning_rate": 2e-05,
+      "loss": 0.04915586,
+      "step": 20215
+    },
+    {
+      "epoch": 40.432,
+      "grad_norm": 2.817385196685791,
+      "learning_rate": 2e-05,
+      "loss": 0.0448599,
+      "step": 20216
+    },
+    {
+      "epoch": 40.434,
+      "grad_norm": 1.1939760446548462,
+      "learning_rate": 2e-05,
+      "loss": 0.04789099,
+      "step": 20217
+    },
+    {
+      "epoch": 40.436,
+      "grad_norm": 1.2851756811141968,
+      "learning_rate": 2e-05,
+      "loss": 0.05403595,
+      "step": 20218
+    },
+    {
+      "epoch": 40.438,
+      "grad_norm": 1.0391663312911987,
+      "learning_rate": 2e-05,
+      "loss": 0.04843492,
+      "step": 20219
+    },
+    {
+      "epoch": 40.44,
+      "grad_norm": 1.0590914487838745,
+      "learning_rate": 2e-05,
+      "loss": 0.04758727,
+      "step": 20220
+    },
+    {
+      "epoch": 40.442,
+      "grad_norm": 1.6666922569274902,
+      "learning_rate": 2e-05,
+      "loss": 0.05023216,
+      "step": 20221
+    },
+    {
+      "epoch": 40.444,
+      "grad_norm": 1.1484979391098022,
+      "learning_rate": 2e-05,
+      "loss": 0.0386183,
+      "step": 20222
+    },
+    {
+      "epoch": 40.446,
+      "grad_norm": 2.4382147789001465,
+      "learning_rate": 2e-05,
+      "loss": 0.06324873,
+      "step": 20223
+    },
+    {
+      "epoch": 40.448,
+      "grad_norm": 1.1172566413879395,
+      "learning_rate": 2e-05,
+      "loss": 0.04193846,
+      "step": 20224
+    },
+    {
+      "epoch": 40.45,
+      "grad_norm": 4.376603603363037,
+      "learning_rate": 2e-05,
+      "loss": 0.05779196,
+      "step": 20225
+    },
+    {
+      "epoch": 40.452,
+      "grad_norm": 1.0015493631362915,
+      "learning_rate": 2e-05,
+      "loss": 0.02681638,
+      "step": 20226
+    },
+    {
+      "epoch": 40.454,
+      "grad_norm": 1.31887948513031,
+      "learning_rate": 2e-05,
+      "loss": 0.05519604,
+      "step": 20227
+    },
+    {
+      "epoch": 40.456,
+      "grad_norm": 1.7096487283706665,
+      "learning_rate": 2e-05,
+      "loss": 0.07397369,
+      "step": 20228
+    },
+    {
+      "epoch": 40.458,
+      "grad_norm": 1.3221633434295654,
+      "learning_rate": 2e-05,
+      "loss": 0.05340026,
+      "step": 20229
+    },
+    {
+      "epoch": 40.46,
+      "grad_norm": 1.0366623401641846,
+      "learning_rate": 2e-05,
+      "loss": 0.03435313,
+      "step": 20230
+    },
+    {
+      "epoch": 40.462,
+      "grad_norm": 1.385132908821106,
+      "learning_rate": 2e-05,
+      "loss": 0.06896439,
+      "step": 20231
+    },
+    {
+      "epoch": 40.464,
+      "grad_norm": 2.499610662460327,
+      "learning_rate": 2e-05,
+      "loss": 0.06424156,
+      "step": 20232
+    },
+    {
+      "epoch": 40.466,
+      "grad_norm": 1.1677004098892212,
+      "learning_rate": 2e-05,
+      "loss": 0.03992238,
+      "step": 20233
+    },
+    {
+      "epoch": 40.468,
+      "grad_norm": 3.2381439208984375,
+      "learning_rate": 2e-05,
+      "loss": 0.05073539,
+      "step": 20234
+    },
+    {
+      "epoch": 40.47,
+      "grad_norm": 1.6042966842651367,
+      "learning_rate": 2e-05,
+      "loss": 0.03290394,
+      "step": 20235
+    },
+    {
+      "epoch": 40.472,
+      "grad_norm": 1.2547248601913452,
+      "learning_rate": 2e-05,
+      "loss": 0.04911025,
+      "step": 20236
+    },
+    {
+      "epoch": 40.474,
+      "grad_norm": 1.3358653783798218,
+      "learning_rate": 2e-05,
+      "loss": 0.0705348,
+      "step": 20237
+    },
+    {
+      "epoch": 40.476,
+      "grad_norm": 1.3888798952102661,
+      "learning_rate": 2e-05,
+      "loss": 0.04967166,
+      "step": 20238
+    },
+    {
+      "epoch": 40.478,
+      "grad_norm": 1.2543953657150269,
+      "learning_rate": 2e-05,
+      "loss": 0.04666407,
+      "step": 20239
+    },
+    {
+      "epoch": 40.48,
+      "grad_norm": 1.0593624114990234,
+      "learning_rate": 2e-05,
+      "loss": 0.03600551,
+      "step": 20240
+    },
+    {
+      "epoch": 40.482,
+      "grad_norm": 1.375593900680542,
+      "learning_rate": 2e-05,
+      "loss": 0.06600478,
+      "step": 20241
+    },
+    {
+      "epoch": 40.484,
+      "grad_norm": 1.3863525390625,
+      "learning_rate": 2e-05,
+      "loss": 0.03970123,
+      "step": 20242
+    },
+    {
+      "epoch": 40.486,
+      "grad_norm": 1.4013750553131104,
+      "learning_rate": 2e-05,
+      "loss": 0.05529185,
+      "step": 20243
+    },
+    {
+      "epoch": 40.488,
+      "grad_norm": 1.3088881969451904,
+      "learning_rate": 2e-05,
+      "loss": 0.05219998,
+      "step": 20244
+    },
+    {
+      "epoch": 40.49,
+      "grad_norm": 1.4817276000976562,
+      "learning_rate": 2e-05,
+      "loss": 0.05724048,
+      "step": 20245
+    },
+    {
+      "epoch": 40.492,
+      "grad_norm": 1.4716103076934814,
+      "learning_rate": 2e-05,
+      "loss": 0.05594869,
+      "step": 20246
+    },
+    {
+      "epoch": 40.494,
+      "grad_norm": 1.3419957160949707,
+      "learning_rate": 2e-05,
+      "loss": 0.06096876,
+      "step": 20247
+    },
+    {
+      "epoch": 40.496,
+      "grad_norm": 0.9824522137641907,
+      "learning_rate": 2e-05,
+      "loss": 0.03611157,
+      "step": 20248
+    },
+    {
+      "epoch": 40.498,
+      "grad_norm": 1.4187341928482056,
+      "learning_rate": 2e-05,
+      "loss": 0.05075159,
+      "step": 20249
+    },
+    {
+      "epoch": 40.5,
+      "grad_norm": 1.4452311992645264,
+      "learning_rate": 2e-05,
+      "loss": 0.04892913,
+      "step": 20250
+    },
+    {
+      "epoch": 40.502,
+      "grad_norm": 1.3794788122177124,
+      "learning_rate": 2e-05,
+      "loss": 0.04053912,
+      "step": 20251
+    },
+    {
+      "epoch": 40.504,
+      "grad_norm": 1.2607942819595337,
+      "learning_rate": 2e-05,
+      "loss": 0.04043246,
+      "step": 20252
+    },
+    {
+      "epoch": 40.506,
+      "grad_norm": 1.2887449264526367,
+      "learning_rate": 2e-05,
+      "loss": 0.04719897,
+      "step": 20253
+    },
+    {
+      "epoch": 40.508,
+      "grad_norm": 1.3014177083969116,
+      "learning_rate": 2e-05,
+      "loss": 0.05486921,
+      "step": 20254
+    },
+    {
+      "epoch": 40.51,
+      "grad_norm": 1.0283253192901611,
+      "learning_rate": 2e-05,
+      "loss": 0.03235307,
+      "step": 20255
+    },
+    {
+      "epoch": 40.512,
+      "grad_norm": 2.2812840938568115,
+      "learning_rate": 2e-05,
+      "loss": 0.05784634,
+      "step": 20256
+    },
+    {
+      "epoch": 40.514,
+      "grad_norm": 1.0626236200332642,
+      "learning_rate": 2e-05,
+      "loss": 0.0345856,
+      "step": 20257
+    },
+    {
+      "epoch": 40.516,
+      "grad_norm": 1.3558917045593262,
+      "learning_rate": 2e-05,
+      "loss": 0.05744244,
+      "step": 20258
+    },
+    {
+      "epoch": 40.518,
+      "grad_norm": 1.1966677904129028,
+      "learning_rate": 2e-05,
+      "loss": 0.04859721,
+      "step": 20259
+    },
+    {
+      "epoch": 40.52,
+      "grad_norm": 2.496509552001953,
+      "learning_rate": 2e-05,
+      "loss": 0.05492661,
+      "step": 20260
+    },
+    {
+      "epoch": 40.522,
+      "grad_norm": 1.3958922624588013,
+      "learning_rate": 2e-05,
+      "loss": 0.04244683,
+      "step": 20261
+    },
+    {
+      "epoch": 40.524,
+      "grad_norm": 1.9222729206085205,
+      "learning_rate": 2e-05,
+      "loss": 0.04568821,
+      "step": 20262
+    },
+    {
+      "epoch": 40.526,
+      "grad_norm": 1.3235441446304321,
+      "learning_rate": 2e-05,
+      "loss": 0.04356699,
+      "step": 20263
+    },
+    {
+      "epoch": 40.528,
+      "grad_norm": 1.4533641338348389,
+      "learning_rate": 2e-05,
+      "loss": 0.052623,
+      "step": 20264
+    },
+    {
+      "epoch": 40.53,
+      "grad_norm": 1.1498498916625977,
+      "learning_rate": 2e-05,
+      "loss": 0.04153707,
+      "step": 20265
+    },
+    {
+      "epoch": 40.532,
+      "grad_norm": 1.6196757555007935,
+      "learning_rate": 2e-05,
+      "loss": 0.05171479,
+      "step": 20266
+    },
+    {
+      "epoch": 40.534,
+      "grad_norm": 1.4504308700561523,
+      "learning_rate": 2e-05,
+      "loss": 0.04189808,
+      "step": 20267
+    },
+    {
+      "epoch": 40.536,
+      "grad_norm": 2.833212375640869,
+      "learning_rate": 2e-05,
+      "loss": 0.05008708,
+      "step": 20268
+    },
+    {
+      "epoch": 40.538,
+      "grad_norm": 1.1308425664901733,
+      "learning_rate": 2e-05,
+      "loss": 0.03203578,
+      "step": 20269
+    },
+    {
+      "epoch": 40.54,
+      "grad_norm": 1.2283337116241455,
+      "learning_rate": 2e-05,
+      "loss": 0.05175584,
+      "step": 20270
+    },
+    {
+      "epoch": 40.542,
+      "grad_norm": 1.1894733905792236,
+      "learning_rate": 2e-05,
+      "loss": 0.04817605,
+      "step": 20271
+    },
+    {
+      "epoch": 40.544,
+      "grad_norm": 1.2060327529907227,
+      "learning_rate": 2e-05,
+      "loss": 0.04768998,
+      "step": 20272
+    },
+    {
+      "epoch": 40.546,
+      "grad_norm": 1.2168883085250854,
+      "learning_rate": 2e-05,
+      "loss": 0.05099332,
+      "step": 20273
+    },
+    {
+      "epoch": 40.548,
+      "grad_norm": 1.1243617534637451,
+      "learning_rate": 2e-05,
+      "loss": 0.03701821,
+      "step": 20274
+    },
+    {
+      "epoch": 40.55,
+      "grad_norm": 1.3077539205551147,
+      "learning_rate": 2e-05,
+      "loss": 0.05925671,
+      "step": 20275
+    },
+    {
+      "epoch": 40.552,
+      "grad_norm": 1.9214268922805786,
+      "learning_rate": 2e-05,
+      "loss": 0.05207685,
+      "step": 20276
+    },
+    {
+      "epoch": 40.554,
+      "grad_norm": 1.2245525121688843,
+      "learning_rate": 2e-05,
+      "loss": 0.03042131,
+      "step": 20277
+    },
+    {
+      "epoch": 40.556,
+      "grad_norm": 2.7672958374023438,
+      "learning_rate": 2e-05,
+      "loss": 0.06857204,
+      "step": 20278
+    },
+    {
+      "epoch": 40.558,
+      "grad_norm": 1.2869014739990234,
+      "learning_rate": 2e-05,
+      "loss": 0.04283034,
+      "step": 20279
+    },
+    {
+      "epoch": 40.56,
+      "grad_norm": 1.1815762519836426,
+      "learning_rate": 2e-05,
+      "loss": 0.04052944,
+      "step": 20280
+    },
+    {
+      "epoch": 40.562,
+      "grad_norm": 2.087244749069214,
+      "learning_rate": 2e-05,
+      "loss": 0.06331733,
+      "step": 20281
+    },
+    {
+      "epoch": 40.564,
+      "grad_norm": 1.3826656341552734,
+      "learning_rate": 2e-05,
+      "loss": 0.04571678,
+      "step": 20282
+    },
+    {
+      "epoch": 40.566,
+      "grad_norm": 3.096768617630005,
+      "learning_rate": 2e-05,
+      "loss": 0.05192983,
+      "step": 20283
+    },
+    {
+      "epoch": 40.568,
+      "grad_norm": 1.186075210571289,
+      "learning_rate": 2e-05,
+      "loss": 0.05161081,
+      "step": 20284
+    },
+    {
+      "epoch": 40.57,
+      "grad_norm": 1.5007069110870361,
+      "learning_rate": 2e-05,
+      "loss": 0.0494061,
+      "step": 20285
+    },
+    {
+      "epoch": 40.572,
+      "grad_norm": 1.4686963558197021,
+      "learning_rate": 2e-05,
+      "loss": 0.04155823,
+      "step": 20286
+    },
+    {
+      "epoch": 40.574,
+      "grad_norm": 1.156620979309082,
+      "learning_rate": 2e-05,
+      "loss": 0.04370104,
+      "step": 20287
+    },
+    {
+      "epoch": 40.576,
+      "grad_norm": 0.9128440022468567,
+      "learning_rate": 2e-05,
+      "loss": 0.03572758,
+      "step": 20288
+    },
+    {
+      "epoch": 40.578,
+      "grad_norm": 2.242276668548584,
+      "learning_rate": 2e-05,
+      "loss": 0.0504857,
+      "step": 20289
+    },
+    {
+      "epoch": 40.58,
+      "grad_norm": 3.176802396774292,
+      "learning_rate": 2e-05,
+      "loss": 0.05922861,
+      "step": 20290
+    },
+    {
+      "epoch": 40.582,
+      "grad_norm": 1.0698144435882568,
+      "learning_rate": 2e-05,
+      "loss": 0.03699166,
+      "step": 20291
+    },
+    {
+      "epoch": 40.584,
+      "grad_norm": 1.1629245281219482,
+      "learning_rate": 2e-05,
+      "loss": 0.04336737,
+      "step": 20292
+    },
+    {
+      "epoch": 40.586,
+      "grad_norm": 1.35090172290802,
+      "learning_rate": 2e-05,
+      "loss": 0.05229773,
+      "step": 20293
+    },
+    {
+      "epoch": 40.588,
+      "grad_norm": 1.5074291229248047,
+      "learning_rate": 2e-05,
+      "loss": 0.03177891,
+      "step": 20294
+    },
+    {
+      "epoch": 40.59,
+      "grad_norm": 1.5969630479812622,
+      "learning_rate": 2e-05,
+      "loss": 0.02993333,
+      "step": 20295
+    },
+    {
+      "epoch": 40.592,
+      "grad_norm": 1.2182235717773438,
+      "learning_rate": 2e-05,
+      "loss": 0.05838227,
+      "step": 20296
+    },
+    {
+      "epoch": 40.594,
+      "grad_norm": 1.0844266414642334,
+      "learning_rate": 2e-05,
+      "loss": 0.04683068,
+      "step": 20297
+    },
+    {
+      "epoch": 40.596,
+      "grad_norm": 1.113297939300537,
+      "learning_rate": 2e-05,
+      "loss": 0.04225264,
+      "step": 20298
+    },
+    {
+      "epoch": 40.598,
+      "grad_norm": 1.3273909091949463,
+      "learning_rate": 2e-05,
+      "loss": 0.05528468,
+      "step": 20299
+    },
+    {
+      "epoch": 40.6,
+      "grad_norm": 1.0608805418014526,
+      "learning_rate": 2e-05,
+      "loss": 0.04179974,
+      "step": 20300
+    },
+    {
+      "epoch": 40.602,
+      "grad_norm": 2.0102715492248535,
+      "learning_rate": 2e-05,
+      "loss": 0.05286686,
+      "step": 20301
+    },
+    {
+      "epoch": 40.604,
+      "grad_norm": 1.4773969650268555,
+      "learning_rate": 2e-05,
+      "loss": 0.07245906,
+      "step": 20302
+    },
+    {
+      "epoch": 40.606,
+      "grad_norm": 1.4324716329574585,
+      "learning_rate": 2e-05,
+      "loss": 0.0385082,
+      "step": 20303
+    },
+    {
+      "epoch": 40.608,
+      "grad_norm": 1.1375013589859009,
+      "learning_rate": 2e-05,
+      "loss": 0.04455408,
+      "step": 20304
+    },
+    {
+      "epoch": 40.61,
+      "grad_norm": 1.8112099170684814,
+      "learning_rate": 2e-05,
+      "loss": 0.04102049,
+      "step": 20305
+    },
+    {
+      "epoch": 40.612,
+      "grad_norm": 1.1016380786895752,
+      "learning_rate": 2e-05,
+      "loss": 0.04536098,
+      "step": 20306
+    },
+    {
+      "epoch": 40.614,
+      "grad_norm": 1.2239372730255127,
+      "learning_rate": 2e-05,
+      "loss": 0.03934653,
+      "step": 20307
+    },
+    {
+      "epoch": 40.616,
+      "grad_norm": 1.4652773141860962,
+      "learning_rate": 2e-05,
+      "loss": 0.04647703,
+      "step": 20308
+    },
+    {
+      "epoch": 40.618,
+      "grad_norm": 1.3046026229858398,
+      "learning_rate": 2e-05,
+      "loss": 0.04674635,
+      "step": 20309
+    },
+    {
+      "epoch": 40.62,
+      "grad_norm": 1.244250774383545,
+      "learning_rate": 2e-05,
+      "loss": 0.04536857,
+      "step": 20310
+    },
+    {
+      "epoch": 40.622,
+      "grad_norm": 1.1843619346618652,
+      "learning_rate": 2e-05,
+      "loss": 0.05122893,
+      "step": 20311
+    },
+    {
+      "epoch": 40.624,
+      "grad_norm": 1.3475395441055298,
+      "learning_rate": 2e-05,
+      "loss": 0.05014464,
+      "step": 20312
+    },
+    {
+      "epoch": 40.626,
+      "grad_norm": 3.323194742202759,
+      "learning_rate": 2e-05,
+      "loss": 0.05366961,
+      "step": 20313
+    },
+    {
+      "epoch": 40.628,
+      "grad_norm": 1.2654601335525513,
+      "learning_rate": 2e-05,
+      "loss": 0.03832448,
+      "step": 20314
+    },
+    {
+      "epoch": 40.63,
+      "grad_norm": 1.2079969644546509,
+      "learning_rate": 2e-05,
+      "loss": 0.03466445,
+      "step": 20315
+    },
+    {
+      "epoch": 40.632,
+      "grad_norm": 1.2388030290603638,
+      "learning_rate": 2e-05,
+      "loss": 0.05009665,
+      "step": 20316
+    },
+    {
+      "epoch": 40.634,
+      "grad_norm": 1.3457781076431274,
+      "learning_rate": 2e-05,
+      "loss": 0.06065813,
+      "step": 20317
+    },
+    {
+      "epoch": 40.636,
+      "grad_norm": 1.2236225605010986,
+      "learning_rate": 2e-05,
+      "loss": 0.05330027,
+      "step": 20318
+    },
+    {
+      "epoch": 40.638,
+      "grad_norm": 1.1197084188461304,
+      "learning_rate": 2e-05,
+      "loss": 0.04628935,
+      "step": 20319
+    },
+    {
+      "epoch": 40.64,
+      "grad_norm": 2.9297568798065186,
+      "learning_rate": 2e-05,
+      "loss": 0.06352062,
+      "step": 20320
+    },
+    {
+      "epoch": 40.642,
+      "grad_norm": 1.181127667427063,
+      "learning_rate": 2e-05,
+      "loss": 0.04264445,
+      "step": 20321
+    },
+    {
+      "epoch": 40.644,
+      "grad_norm": 3.967763662338257,
+      "learning_rate": 2e-05,
+      "loss": 0.06111223,
+      "step": 20322
+    },
+    {
+      "epoch": 40.646,
+      "grad_norm": 1.2990795373916626,
+      "learning_rate": 2e-05,
+      "loss": 0.03806433,
+      "step": 20323
+    },
+    {
+      "epoch": 40.648,
+      "grad_norm": 1.437103271484375,
+      "learning_rate": 2e-05,
+      "loss": 0.05348209,
+      "step": 20324
+    },
+    {
+      "epoch": 40.65,
+      "grad_norm": 1.2915222644805908,
+      "learning_rate": 2e-05,
+      "loss": 0.04917194,
+      "step": 20325
+    },
+    {
+      "epoch": 40.652,
+      "grad_norm": 1.3285936117172241,
+      "learning_rate": 2e-05,
+      "loss": 0.05632872,
+      "step": 20326
+    },
+    {
+      "epoch": 40.654,
+      "grad_norm": 1.1885918378829956,
+      "learning_rate": 2e-05,
+      "loss": 0.04639854,
+      "step": 20327
+    },
+    {
+      "epoch": 40.656,
+      "grad_norm": 1.2425380945205688,
+      "learning_rate": 2e-05,
+      "loss": 0.04708571,
+      "step": 20328
+    },
+    {
+      "epoch": 40.658,
+      "grad_norm": 1.2747244834899902,
+      "learning_rate": 2e-05,
+      "loss": 0.05863928,
+      "step": 20329
+    },
+    {
+      "epoch": 40.66,
+      "grad_norm": 1.2093666791915894,
+      "learning_rate": 2e-05,
+      "loss": 0.04224218,
+      "step": 20330
+    },
+    {
+      "epoch": 40.662,
+      "grad_norm": 1.0434030294418335,
+      "learning_rate": 2e-05,
+      "loss": 0.0403709,
+      "step": 20331
+    },
+    {
+      "epoch": 40.664,
+      "grad_norm": 1.101930856704712,
+      "learning_rate": 2e-05,
+      "loss": 0.04556001,
+      "step": 20332
+    },
+    {
+      "epoch": 40.666,
+      "grad_norm": 1.3830711841583252,
+      "learning_rate": 2e-05,
+      "loss": 0.0546868,
+      "step": 20333
+    },
+    {
+      "epoch": 40.668,
+      "grad_norm": 1.0567309856414795,
+      "learning_rate": 2e-05,
+      "loss": 0.04143964,
+      "step": 20334
+    },
+    {
+      "epoch": 40.67,
+      "grad_norm": 1.3042546510696411,
+      "learning_rate": 2e-05,
+      "loss": 0.04565054,
+      "step": 20335
+    },
+    {
+      "epoch": 40.672,
+      "grad_norm": 1.3436415195465088,
+      "learning_rate": 2e-05,
+      "loss": 0.06098578,
+      "step": 20336
+    },
+    {
+      "epoch": 40.674,
+      "grad_norm": 1.1577244997024536,
+      "learning_rate": 2e-05,
+      "loss": 0.04822986,
+      "step": 20337
+    },
+    {
+      "epoch": 40.676,
+      "grad_norm": 1.5855461359024048,
+      "learning_rate": 2e-05,
+      "loss": 0.0444434,
+      "step": 20338
+    },
+    {
+      "epoch": 40.678,
+      "grad_norm": 1.7393256425857544,
+      "learning_rate": 2e-05,
+      "loss": 0.0486975,
+      "step": 20339
+    },
+    {
+      "epoch": 40.68,
+      "grad_norm": 3.428238868713379,
+      "learning_rate": 2e-05,
+      "loss": 0.04472708,
+      "step": 20340
+    },
+    {
+      "epoch": 40.682,
+      "grad_norm": 1.6236741542816162,
+      "learning_rate": 2e-05,
+      "loss": 0.05305525,
+      "step": 20341
+    },
+    {
+      "epoch": 40.684,
+      "grad_norm": 1.2626410722732544,
+      "learning_rate": 2e-05,
+      "loss": 0.05415434,
+      "step": 20342
+    },
+    {
+      "epoch": 40.686,
+      "grad_norm": 1.0027401447296143,
+      "learning_rate": 2e-05,
+      "loss": 0.03447507,
+      "step": 20343
+    },
+    {
+      "epoch": 40.688,
+      "grad_norm": 1.2402012348175049,
+      "learning_rate": 2e-05,
+      "loss": 0.04778074,
+      "step": 20344
+    },
+    {
+      "epoch": 40.69,
+      "grad_norm": 1.050713300704956,
+      "learning_rate": 2e-05,
+      "loss": 0.03177363,
+      "step": 20345
+    },
+    {
+      "epoch": 40.692,
+      "grad_norm": 1.2489185333251953,
+      "learning_rate": 2e-05,
+      "loss": 0.04607525,
+      "step": 20346
+    },
+    {
+      "epoch": 40.694,
+      "grad_norm": 1.1416654586791992,
+      "learning_rate": 2e-05,
+      "loss": 0.04840235,
+      "step": 20347
+    },
+    {
+      "epoch": 40.696,
+      "grad_norm": 1.1122640371322632,
+      "learning_rate": 2e-05,
+      "loss": 0.03137276,
+      "step": 20348
+    },
+    {
+      "epoch": 40.698,
+      "grad_norm": 1.4408119916915894,
+      "learning_rate": 2e-05,
+      "loss": 0.05424093,
+      "step": 20349
+    },
+    {
+      "epoch": 40.7,
+      "grad_norm": 0.9952658414840698,
+      "learning_rate": 2e-05,
+      "loss": 0.02818977,
+      "step": 20350
+    },
+    {
+      "epoch": 40.702,
+      "grad_norm": 1.1479781866073608,
+      "learning_rate": 2e-05,
+      "loss": 0.03366838,
+      "step": 20351
+    },
+    {
+      "epoch": 40.704,
+      "grad_norm": 1.1322118043899536,
+      "learning_rate": 2e-05,
+      "loss": 0.04626714,
+      "step": 20352
+    },
+    {
+      "epoch": 40.706,
+      "grad_norm": 1.3414576053619385,
+      "learning_rate": 2e-05,
+      "loss": 0.04931616,
+      "step": 20353
+    },
+    {
+      "epoch": 40.708,
+      "grad_norm": 1.433391809463501,
+      "learning_rate": 2e-05,
+      "loss": 0.04131094,
+      "step": 20354
+    },
+    {
+      "epoch": 40.71,
+      "grad_norm": 1.145910382270813,
+      "learning_rate": 2e-05,
+      "loss": 0.04541127,
+      "step": 20355
+    },
+    {
+      "epoch": 40.712,
+      "grad_norm": 1.17069411277771,
+      "learning_rate": 2e-05,
+      "loss": 0.04991463,
+      "step": 20356
+    },
+    {
+      "epoch": 40.714,
+      "grad_norm": 1.102471113204956,
+      "learning_rate": 2e-05,
+      "loss": 0.04218239,
+      "step": 20357
+    },
+    {
+      "epoch": 40.716,
+      "grad_norm": 1.172473430633545,
+      "learning_rate": 2e-05,
+      "loss": 0.0438892,
+      "step": 20358
+    },
+    {
+      "epoch": 40.718,
+      "grad_norm": 1.8067024946212769,
+      "learning_rate": 2e-05,
+      "loss": 0.05643699,
+      "step": 20359
+    },
+    {
+      "epoch": 40.72,
+      "grad_norm": 1.3518229722976685,
+      "learning_rate": 2e-05,
+      "loss": 0.0555013,
+      "step": 20360
+    },
+    {
+      "epoch": 40.722,
+      "grad_norm": 1.3783165216445923,
+      "learning_rate": 2e-05,
+      "loss": 0.05945724,
+      "step": 20361
+    },
+    {
+      "epoch": 40.724,
+      "grad_norm": 1.1957453489303589,
+      "learning_rate": 2e-05,
+      "loss": 0.05680922,
+      "step": 20362
+    },
+    {
+      "epoch": 40.726,
+      "grad_norm": 1.1252014636993408,
+      "learning_rate": 2e-05,
+      "loss": 0.05653615,
+      "step": 20363
+    },
+    {
+      "epoch": 40.728,
+      "grad_norm": 1.2244309186935425,
+      "learning_rate": 2e-05,
+      "loss": 0.04203859,
+      "step": 20364
+    },
+    {
+      "epoch": 40.73,
+      "grad_norm": 5.841495037078857,
+      "learning_rate": 2e-05,
+      "loss": 0.05102485,
+      "step": 20365
+    },
+    {
+      "epoch": 40.732,
+      "grad_norm": 2.0462138652801514,
+      "learning_rate": 2e-05,
+      "loss": 0.05800704,
+      "step": 20366
+    },
+    {
+      "epoch": 40.734,
+      "grad_norm": 1.3569488525390625,
+      "learning_rate": 2e-05,
+      "loss": 0.04887813,
+      "step": 20367
+    },
+    {
+      "epoch": 40.736,
+      "grad_norm": 0.9353485703468323,
+      "learning_rate": 2e-05,
+      "loss": 0.0317823,
+      "step": 20368
+    },
+    {
+      "epoch": 40.738,
+      "grad_norm": 1.1075727939605713,
+      "learning_rate": 2e-05,
+      "loss": 0.04405333,
+      "step": 20369
+    },
+    {
+      "epoch": 40.74,
+      "grad_norm": 1.120461344718933,
+      "learning_rate": 2e-05,
+      "loss": 0.04158332,
+      "step": 20370
+    },
+    {
+      "epoch": 40.742,
+      "grad_norm": 1.173218846321106,
+      "learning_rate": 2e-05,
+      "loss": 0.03538671,
+      "step": 20371
+    },
+    {
+      "epoch": 40.744,
+      "grad_norm": 1.3119045495986938,
+      "learning_rate": 2e-05,
+      "loss": 0.04965501,
+      "step": 20372
+    },
+    {
+      "epoch": 40.746,
+      "grad_norm": 1.4258747100830078,
+      "learning_rate": 2e-05,
+      "loss": 0.04786938,
+      "step": 20373
+    },
+    {
+      "epoch": 40.748,
+      "grad_norm": 1.2724413871765137,
+      "learning_rate": 2e-05,
+      "loss": 0.04649073,
+      "step": 20374
+    },
+    {
+      "epoch": 40.75,
+      "grad_norm": 1.6921082735061646,
+      "learning_rate": 2e-05,
+      "loss": 0.03883822,
+      "step": 20375
+    },
+    {
+      "epoch": 40.752,
+      "grad_norm": 3.84258770942688,
+      "learning_rate": 2e-05,
+      "loss": 0.05532685,
+      "step": 20376
+    },
+    {
+      "epoch": 40.754,
+      "grad_norm": 1.4394497871398926,
+      "learning_rate": 2e-05,
+      "loss": 0.04756043,
+      "step": 20377
+    },
+    {
+      "epoch": 40.756,
+      "grad_norm": 1.1952511072158813,
+      "learning_rate": 2e-05,
+      "loss": 0.05040652,
+      "step": 20378
+    },
+    {
+      "epoch": 40.758,
+      "grad_norm": 1.139320969581604,
+      "learning_rate": 2e-05,
+      "loss": 0.0432628,
+      "step": 20379
+    },
+    {
+      "epoch": 40.76,
+      "grad_norm": 1.0140758752822876,
+      "learning_rate": 2e-05,
+      "loss": 0.0302791,
+      "step": 20380
+    },
+    {
+      "epoch": 40.762,
+      "grad_norm": 1.783044695854187,
+      "learning_rate": 2e-05,
+      "loss": 0.05844506,
+      "step": 20381
+    },
+    {
+      "epoch": 40.764,
+      "grad_norm": 1.1988928318023682,
+      "learning_rate": 2e-05,
+      "loss": 0.03283492,
+      "step": 20382
+    },
+    {
+      "epoch": 40.766,
+      "grad_norm": 0.9184706807136536,
+      "learning_rate": 2e-05,
+      "loss": 0.03624507,
+      "step": 20383
+    },
+    {
+      "epoch": 40.768,
+      "grad_norm": 1.7441927194595337,
+      "learning_rate": 2e-05,
+      "loss": 0.03876618,
+      "step": 20384
+    },
+    {
+      "epoch": 40.77,
+      "grad_norm": 1.261252522468567,
+      "learning_rate": 2e-05,
+      "loss": 0.05795658,
+      "step": 20385
+    },
+    {
+      "epoch": 40.772,
+      "grad_norm": 1.1532663106918335,
+      "learning_rate": 2e-05,
+      "loss": 0.050796,
+      "step": 20386
+    },
+    {
+      "epoch": 40.774,
+      "grad_norm": 1.3474946022033691,
+      "learning_rate": 2e-05,
+      "loss": 0.0455665,
+      "step": 20387
+    },
+    {
+      "epoch": 40.776,
+      "grad_norm": 1.175278663635254,
+      "learning_rate": 2e-05,
+      "loss": 0.03921409,
+      "step": 20388
+    },
+    {
+      "epoch": 40.778,
+      "grad_norm": 1.2267823219299316,
+      "learning_rate": 2e-05,
+      "loss": 0.05682051,
+      "step": 20389
+    },
+    {
+      "epoch": 40.78,
+      "grad_norm": 3.2952206134796143,
+      "learning_rate": 2e-05,
+      "loss": 0.04363419,
+      "step": 20390
+    },
+    {
+      "epoch": 40.782,
+      "grad_norm": 1.7365856170654297,
+      "learning_rate": 2e-05,
+      "loss": 0.05543724,
+      "step": 20391
+    },
+    {
+      "epoch": 40.784,
+      "grad_norm": 1.2324695587158203,
+      "learning_rate": 2e-05,
+      "loss": 0.05429393,
+      "step": 20392
+    },
+    {
+      "epoch": 40.786,
+      "grad_norm": 1.2314908504486084,
+      "learning_rate": 2e-05,
+      "loss": 0.0417459,
+      "step": 20393
+    },
+    {
+      "epoch": 40.788,
+      "grad_norm": 1.052890658378601,
+      "learning_rate": 2e-05,
+      "loss": 0.04568375,
+      "step": 20394
+    },
+    {
+      "epoch": 40.79,
+      "grad_norm": 1.1527416706085205,
+      "learning_rate": 2e-05,
+      "loss": 0.05163721,
+      "step": 20395
+    },
+    {
+      "epoch": 40.792,
+      "grad_norm": 1.255745768547058,
+      "learning_rate": 2e-05,
+      "loss": 0.04441888,
+      "step": 20396
+    },
+    {
+      "epoch": 40.794,
+      "grad_norm": 2.4157254695892334,
+      "learning_rate": 2e-05,
+      "loss": 0.05280969,
+      "step": 20397
+    },
+    {
+      "epoch": 40.796,
+      "grad_norm": 1.297743797302246,
+      "learning_rate": 2e-05,
+      "loss": 0.03922386,
+      "step": 20398
+    },
+    {
+      "epoch": 40.798,
+      "grad_norm": 1.3191263675689697,
+      "learning_rate": 2e-05,
+      "loss": 0.0459812,
+      "step": 20399
+    },
+    {
+      "epoch": 40.8,
+      "grad_norm": 1.6138203144073486,
+      "learning_rate": 2e-05,
+      "loss": 0.04809796,
+      "step": 20400
+    },
+    {
+      "epoch": 40.802,
+      "grad_norm": 1.0761232376098633,
+      "learning_rate": 2e-05,
+      "loss": 0.04102475,
+      "step": 20401
+    },
+    {
+      "epoch": 40.804,
+      "grad_norm": 1.3820741176605225,
+      "learning_rate": 2e-05,
+      "loss": 0.05293287,
+      "step": 20402
+    },
+    {
+      "epoch": 40.806,
+      "grad_norm": 0.8618505597114563,
+      "learning_rate": 2e-05,
+      "loss": 0.03630167,
+      "step": 20403
+    },
+    {
+      "epoch": 40.808,
+      "grad_norm": 1.2943086624145508,
+      "learning_rate": 2e-05,
+      "loss": 0.04525239,
+      "step": 20404
+    },
+    {
+      "epoch": 40.81,
+      "grad_norm": 1.6318711042404175,
+      "learning_rate": 2e-05,
+      "loss": 0.04626258,
+      "step": 20405
+    },
+    {
+      "epoch": 40.812,
+      "grad_norm": 1.0166434049606323,
+      "learning_rate": 2e-05,
+      "loss": 0.03998704,
+      "step": 20406
+    },
+    {
+      "epoch": 40.814,
+      "grad_norm": 2.893320083618164,
+      "learning_rate": 2e-05,
+      "loss": 0.0571047,
+      "step": 20407
+    },
+    {
+      "epoch": 40.816,
+      "grad_norm": 1.0632132291793823,
+      "learning_rate": 2e-05,
+      "loss": 0.03629584,
+      "step": 20408
+    },
+    {
+      "epoch": 40.818,
+      "grad_norm": 1.168479084968567,
+      "learning_rate": 2e-05,
+      "loss": 0.03368823,
+      "step": 20409
+    },
+    {
+      "epoch": 40.82,
+      "grad_norm": 1.0956116914749146,
+      "learning_rate": 2e-05,
+      "loss": 0.03800235,
+      "step": 20410
+    },
+    {
+      "epoch": 40.822,
+      "grad_norm": 1.0127363204956055,
+      "learning_rate": 2e-05,
+      "loss": 0.03506695,
+      "step": 20411
+    },
+    {
+      "epoch": 40.824,
+      "grad_norm": 1.6961843967437744,
+      "learning_rate": 2e-05,
+      "loss": 0.04757579,
+      "step": 20412
+    },
+    {
+      "epoch": 40.826,
+      "grad_norm": 1.0485764741897583,
+      "learning_rate": 2e-05,
+      "loss": 0.03640955,
+      "step": 20413
+    },
+    {
+      "epoch": 40.828,
+      "grad_norm": 1.4049381017684937,
+      "learning_rate": 2e-05,
+      "loss": 0.05531392,
+      "step": 20414
+    },
+    {
+      "epoch": 40.83,
+      "grad_norm": 1.3320997953414917,
+      "learning_rate": 2e-05,
+      "loss": 0.04148456,
+      "step": 20415
+    },
+    {
+      "epoch": 40.832,
+      "grad_norm": 1.2389609813690186,
+      "learning_rate": 2e-05,
+      "loss": 0.04860774,
+      "step": 20416
+    },
+    {
+      "epoch": 40.834,
+      "grad_norm": 0.9642683863639832,
+      "learning_rate": 2e-05,
+      "loss": 0.04272363,
+      "step": 20417
+    },
+    {
+      "epoch": 40.836,
+      "grad_norm": 1.2460993528366089,
+      "learning_rate": 2e-05,
+      "loss": 0.05987071,
+      "step": 20418
+    },
+    {
+      "epoch": 40.838,
+      "grad_norm": 1.2246595621109009,
+      "learning_rate": 2e-05,
+      "loss": 0.05187958,
+      "step": 20419
+    },
+    {
+      "epoch": 40.84,
+      "grad_norm": 1.6453964710235596,
+      "learning_rate": 2e-05,
+      "loss": 0.04696222,
+      "step": 20420
+    },
+    {
+      "epoch": 40.842,
+      "grad_norm": 1.2105334997177124,
+      "learning_rate": 2e-05,
+      "loss": 0.04552323,
+      "step": 20421
+    },
+    {
+      "epoch": 40.844,
+      "grad_norm": 1.1809000968933105,
+      "learning_rate": 2e-05,
+      "loss": 0.04364845,
+      "step": 20422
+    },
+    {
+      "epoch": 40.846,
+      "grad_norm": 2.2245469093322754,
+      "learning_rate": 2e-05,
+      "loss": 0.04627731,
+      "step": 20423
+    },
+    {
+      "epoch": 40.848,
+      "grad_norm": 1.180808186531067,
+      "learning_rate": 2e-05,
+      "loss": 0.04993726,
+      "step": 20424
+    },
+    {
+      "epoch": 40.85,
+      "grad_norm": 1.7033965587615967,
+      "learning_rate": 2e-05,
+      "loss": 0.06255874,
+      "step": 20425
+    },
+    {
+      "epoch": 40.852,
+      "grad_norm": 1.2189267873764038,
+      "learning_rate": 2e-05,
+      "loss": 0.04405816,
+      "step": 20426
+    },
+    {
+      "epoch": 40.854,
+      "grad_norm": 3.7718513011932373,
+      "learning_rate": 2e-05,
+      "loss": 0.05504301,
+      "step": 20427
+    },
+    {
+      "epoch": 40.856,
+      "grad_norm": 1.1234560012817383,
+      "learning_rate": 2e-05,
+      "loss": 0.04025323,
+      "step": 20428
+    },
+    {
+      "epoch": 40.858,
+      "grad_norm": 1.3701295852661133,
+      "learning_rate": 2e-05,
+      "loss": 0.04871433,
+      "step": 20429
+    },
+    {
+      "epoch": 40.86,
+      "grad_norm": 1.6501892805099487,
+      "learning_rate": 2e-05,
+      "loss": 0.06243546,
+      "step": 20430
+    },
+    {
+      "epoch": 40.862,
+      "grad_norm": 1.19022536277771,
+      "learning_rate": 2e-05,
+      "loss": 0.04894875,
+      "step": 20431
+    },
+    {
+      "epoch": 40.864,
+      "grad_norm": 1.0078957080841064,
+      "learning_rate": 2e-05,
+      "loss": 0.03035384,
+      "step": 20432
+    },
+    {
+      "epoch": 40.866,
+      "grad_norm": 1.052426815032959,
+      "learning_rate": 2e-05,
+      "loss": 0.0399336,
+      "step": 20433
+    },
+    {
+      "epoch": 40.868,
+      "grad_norm": 1.7193706035614014,
+      "learning_rate": 2e-05,
+      "loss": 0.05601721,
+      "step": 20434
+    },
+    {
+      "epoch": 40.87,
+      "grad_norm": 1.1827843189239502,
+      "learning_rate": 2e-05,
+      "loss": 0.0472433,
+      "step": 20435
+    },
+    {
+      "epoch": 40.872,
+      "grad_norm": 2.2266764640808105,
+      "learning_rate": 2e-05,
+      "loss": 0.05473574,
+      "step": 20436
+    },
+    {
+      "epoch": 40.874,
+      "grad_norm": 1.7117822170257568,
+      "learning_rate": 2e-05,
+      "loss": 0.0380583,
+      "step": 20437
+    },
+    {
+      "epoch": 40.876,
+      "grad_norm": 1.234930396080017,
+      "learning_rate": 2e-05,
+      "loss": 0.04837093,
+      "step": 20438
+    },
+    {
+      "epoch": 40.878,
+      "grad_norm": 1.377922773361206,
+      "learning_rate": 2e-05,
+      "loss": 0.04900629,
+      "step": 20439
+    },
+    {
+      "epoch": 40.88,
+      "grad_norm": 1.206617832183838,
+      "learning_rate": 2e-05,
+      "loss": 0.05211646,
+      "step": 20440
+    },
+    {
+      "epoch": 40.882,
+      "grad_norm": 1.1824190616607666,
+      "learning_rate": 2e-05,
+      "loss": 0.03713381,
+      "step": 20441
+    },
+    {
+      "epoch": 40.884,
+      "grad_norm": 0.9823435544967651,
+      "learning_rate": 2e-05,
+      "loss": 0.03660665,
+      "step": 20442
+    },
+    {
+      "epoch": 40.886,
+      "grad_norm": 1.2949517965316772,
+      "learning_rate": 2e-05,
+      "loss": 0.05451689,
+      "step": 20443
+    },
+    {
+      "epoch": 40.888,
+      "grad_norm": 1.2883609533309937,
+      "learning_rate": 2e-05,
+      "loss": 0.05717325,
+      "step": 20444
+    },
+    {
+      "epoch": 40.89,
+      "grad_norm": 0.9961985945701599,
+      "learning_rate": 2e-05,
+      "loss": 0.03337915,
+      "step": 20445
+    },
+    {
+      "epoch": 40.892,
+      "grad_norm": 1.195377230644226,
+      "learning_rate": 2e-05,
+      "loss": 0.0425334,
+      "step": 20446
+    },
+    {
+      "epoch": 40.894,
+      "grad_norm": 1.027193546295166,
+      "learning_rate": 2e-05,
+      "loss": 0.02752158,
+      "step": 20447
+    },
+    {
+      "epoch": 40.896,
+      "grad_norm": 1.1972559690475464,
+      "learning_rate": 2e-05,
+      "loss": 0.04988524,
+      "step": 20448
+    },
+    {
+      "epoch": 40.898,
+      "grad_norm": 1.175160527229309,
+      "learning_rate": 2e-05,
+      "loss": 0.04154052,
+      "step": 20449
+    },
+    {
+      "epoch": 40.9,
+      "grad_norm": 1.1122380495071411,
+      "learning_rate": 2e-05,
+      "loss": 0.03493803,
+      "step": 20450
+    },
+    {
+      "epoch": 40.902,
+      "grad_norm": 1.1690434217453003,
+      "learning_rate": 2e-05,
+      "loss": 0.04058939,
+      "step": 20451
+    },
+    {
+      "epoch": 40.904,
+      "grad_norm": 2.595667839050293,
+      "learning_rate": 2e-05,
+      "loss": 0.08465319,
+      "step": 20452
+    },
+    {
+      "epoch": 40.906,
+      "grad_norm": 1.2236500978469849,
+      "learning_rate": 2e-05,
+      "loss": 0.04924143,
+      "step": 20453
+    },
+    {
+      "epoch": 40.908,
+      "grad_norm": 2.681835174560547,
+      "learning_rate": 2e-05,
+      "loss": 0.06435752,
+      "step": 20454
+    },
+    {
+      "epoch": 40.91,
+      "grad_norm": 1.396338701248169,
+      "learning_rate": 2e-05,
+      "loss": 0.04998089,
+      "step": 20455
+    },
+    {
+      "epoch": 40.912,
+      "grad_norm": 1.3858076333999634,
+      "learning_rate": 2e-05,
+      "loss": 0.05152861,
+      "step": 20456
+    },
+    {
+      "epoch": 40.914,
+      "grad_norm": 1.376036286354065,
+      "learning_rate": 2e-05,
+      "loss": 0.04543985,
+      "step": 20457
+    },
+    {
+      "epoch": 40.916,
+      "grad_norm": 1.57783842086792,
+      "learning_rate": 2e-05,
+      "loss": 0.05842921,
+      "step": 20458
+    },
+    {
+      "epoch": 40.918,
+      "grad_norm": 1.5693801641464233,
+      "learning_rate": 2e-05,
+      "loss": 0.0650731,
+      "step": 20459
+    },
+    {
+      "epoch": 40.92,
+      "grad_norm": 1.4704762697219849,
+      "learning_rate": 2e-05,
+      "loss": 0.05608998,
+      "step": 20460
+    },
+    {
+      "epoch": 40.922,
+      "grad_norm": 1.1098164319992065,
+      "learning_rate": 2e-05,
+      "loss": 0.04608128,
+      "step": 20461
+    },
+    {
+      "epoch": 40.924,
+      "grad_norm": 1.1128933429718018,
+      "learning_rate": 2e-05,
+      "loss": 0.03824999,
+      "step": 20462
+    },
+    {
+      "epoch": 40.926,
+      "grad_norm": 1.244307518005371,
+      "learning_rate": 2e-05,
+      "loss": 0.06200198,
+      "step": 20463
+    },
+    {
+      "epoch": 40.928,
+      "grad_norm": 1.2337552309036255,
+      "learning_rate": 2e-05,
+      "loss": 0.03821502,
+      "step": 20464
+    },
+    {
+      "epoch": 40.93,
+      "grad_norm": 1.2475448846817017,
+      "learning_rate": 2e-05,
+      "loss": 0.05957456,
+      "step": 20465
+    },
+    {
+      "epoch": 40.932,
+      "grad_norm": 2.0048727989196777,
+      "learning_rate": 2e-05,
+      "loss": 0.0534838,
+      "step": 20466
+    },
+    {
+      "epoch": 40.934,
+      "grad_norm": 1.6427809000015259,
+      "learning_rate": 2e-05,
+      "loss": 0.03990686,
+      "step": 20467
+    },
+    {
+      "epoch": 40.936,
+      "grad_norm": 1.199326753616333,
+      "learning_rate": 2e-05,
+      "loss": 0.04839782,
+      "step": 20468
+    },
+    {
+      "epoch": 40.938,
+      "grad_norm": 1.3400622606277466,
+      "learning_rate": 2e-05,
+      "loss": 0.05205151,
+      "step": 20469
+    },
+    {
+      "epoch": 40.94,
+      "grad_norm": 1.2996052503585815,
+      "learning_rate": 2e-05,
+      "loss": 0.06201072,
+      "step": 20470
+    },
+    {
+      "epoch": 40.942,
+      "grad_norm": 1.287192463874817,
+      "learning_rate": 2e-05,
+      "loss": 0.05062536,
+      "step": 20471
+    },
+    {
+      "epoch": 40.944,
+      "grad_norm": 3.311911106109619,
+      "learning_rate": 2e-05,
+      "loss": 0.05273374,
+      "step": 20472
+    },
+    {
+      "epoch": 40.946,
+      "grad_norm": 1.086965799331665,
+      "learning_rate": 2e-05,
+      "loss": 0.02739996,
+      "step": 20473
+    },
+    {
+      "epoch": 40.948,
+      "grad_norm": 1.349342942237854,
+      "learning_rate": 2e-05,
+      "loss": 0.05531573,
+      "step": 20474
+    },
+    {
+      "epoch": 40.95,
+      "grad_norm": 1.4824485778808594,
+      "learning_rate": 2e-05,
+      "loss": 0.05945705,
+      "step": 20475
+    },
+    {
+      "epoch": 40.952,
+      "grad_norm": 1.3184701204299927,
+      "learning_rate": 2e-05,
+      "loss": 0.04314397,
+      "step": 20476
+    },
+    {
+      "epoch": 40.954,
+      "grad_norm": 1.385398030281067,
+      "learning_rate": 2e-05,
+      "loss": 0.06099976,
+      "step": 20477
+    },
+    {
+      "epoch": 40.956,
+      "grad_norm": 1.1655802726745605,
+      "learning_rate": 2e-05,
+      "loss": 0.04490764,
+      "step": 20478
+    },
+    {
+      "epoch": 40.958,
+      "grad_norm": 4.062130928039551,
+      "learning_rate": 2e-05,
+      "loss": 0.05665839,
+      "step": 20479
+    },
+    {
+      "epoch": 40.96,
+      "grad_norm": 1.4232219457626343,
+      "learning_rate": 2e-05,
+      "loss": 0.0396259,
+      "step": 20480
+    },
+    {
+      "epoch": 40.962,
+      "grad_norm": 1.2747482061386108,
+      "learning_rate": 2e-05,
+      "loss": 0.05678995,
+      "step": 20481
+    },
+    {
+      "epoch": 40.964,
+      "grad_norm": 1.5956790447235107,
+      "learning_rate": 2e-05,
+      "loss": 0.04028114,
+      "step": 20482
+    },
+    {
+      "epoch": 40.966,
+      "grad_norm": 1.2978450059890747,
+      "learning_rate": 2e-05,
+      "loss": 0.04526225,
+      "step": 20483
+    },
+    {
+      "epoch": 40.968,
+      "grad_norm": 1.3045350313186646,
+      "learning_rate": 2e-05,
+      "loss": 0.05053577,
+      "step": 20484
+    },
+    {
+      "epoch": 40.97,
+      "grad_norm": 1.1985734701156616,
+      "learning_rate": 2e-05,
+      "loss": 0.04328226,
+      "step": 20485
+    },
+    {
+      "epoch": 40.972,
+      "grad_norm": 1.285112977027893,
+      "learning_rate": 2e-05,
+      "loss": 0.04697558,
+      "step": 20486
+    },
+    {
+      "epoch": 40.974,
+      "grad_norm": 1.4183571338653564,
+      "learning_rate": 2e-05,
+      "loss": 0.05476032,
+      "step": 20487
+    },
+    {
+      "epoch": 40.976,
+      "grad_norm": 1.4861711263656616,
+      "learning_rate": 2e-05,
+      "loss": 0.05896805,
+      "step": 20488
+    },
+    {
+      "epoch": 40.978,
+      "grad_norm": 1.6024667024612427,
+      "learning_rate": 2e-05,
+      "loss": 0.04853802,
+      "step": 20489
+    },
+    {
+      "epoch": 40.98,
+      "grad_norm": 1.2148326635360718,
+      "learning_rate": 2e-05,
+      "loss": 0.05488897,
+      "step": 20490
+    },
+    {
+      "epoch": 40.982,
+      "grad_norm": 1.6598411798477173,
+      "learning_rate": 2e-05,
+      "loss": 0.06064874,
+      "step": 20491
+    },
+    {
+      "epoch": 40.984,
+      "grad_norm": 1.2032941579818726,
+      "learning_rate": 2e-05,
+      "loss": 0.04319712,
+      "step": 20492
+    },
+    {
+      "epoch": 40.986,
+      "grad_norm": 2.1562469005584717,
+      "learning_rate": 2e-05,
+      "loss": 0.04315053,
+      "step": 20493
+    },
+    {
+      "epoch": 40.988,
+      "grad_norm": 1.3768160343170166,
+      "learning_rate": 2e-05,
+      "loss": 0.0549387,
+      "step": 20494
+    },
+    {
+      "epoch": 40.99,
+      "grad_norm": 1.250068187713623,
+      "learning_rate": 2e-05,
+      "loss": 0.04305088,
+      "step": 20495
+    },
+    {
+      "epoch": 40.992,
+      "grad_norm": 1.2710068225860596,
+      "learning_rate": 2e-05,
+      "loss": 0.04697052,
+      "step": 20496
+    },
+    {
+      "epoch": 40.994,
+      "grad_norm": 1.4347896575927734,
+      "learning_rate": 2e-05,
+      "loss": 0.050442,
+      "step": 20497
+    },
+    {
+      "epoch": 40.996,
+      "grad_norm": 0.8269062638282776,
+      "learning_rate": 2e-05,
+      "loss": 0.02071571,
+      "step": 20498
+    },
+    {
+      "epoch": 40.998,
+      "grad_norm": 4.110866069793701,
+      "learning_rate": 2e-05,
+      "loss": 0.03941682,
+      "step": 20499
+    },
+    {
+      "epoch": 41.0,
+      "grad_norm": 1.8041988611221313,
+      "learning_rate": 2e-05,
+      "loss": 0.04850418,
+      "step": 20500
+    },
+    {
+      "epoch": 41.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.998,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9880239520958084,
+        "Equal_1": 0.998,
+        "Equal_2": 0.9880239520958084,
+        "Equal_3": 0.9900199600798403,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9940119760479041,
+        "Parallel_1": 0.9839679358717435,
+        "Parallel_2": 0.9959919839679359,
+        "Parallel_3": 0.986,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.996,
+        "Perpendicular_3": 0.8977955911823647,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.994,
+        "PointLiesOnCircle_3": 0.998,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9820359281437125
+      },
+      "eval_runtime": 321.056,
+      "eval_samples_per_second": 32.705,
+      "eval_steps_per_second": 0.654,
+      "step": 20500
+    },
+    {
+      "epoch": 41.002,
+      "grad_norm": 1.1957684755325317,
+      "learning_rate": 2e-05,
+      "loss": 0.03748061,
+      "step": 20501
+    },
+    {
+      "epoch": 41.004,
+      "grad_norm": 1.1992721557617188,
+      "learning_rate": 2e-05,
+      "loss": 0.05174942,
+      "step": 20502
+    },
+    {
+      "epoch": 41.006,
+      "grad_norm": 1.3277416229248047,
+      "learning_rate": 2e-05,
+      "loss": 0.04400465,
+      "step": 20503
+    },
+    {
+      "epoch": 41.008,
+      "grad_norm": 1.6822006702423096,
+      "learning_rate": 2e-05,
+      "loss": 0.05937324,
+      "step": 20504
+    },
+    {
+      "epoch": 41.01,
+      "grad_norm": 2.2135109901428223,
+      "learning_rate": 2e-05,
+      "loss": 0.06282458,
+      "step": 20505
+    },
+    {
+      "epoch": 41.012,
+      "grad_norm": 1.2385971546173096,
+      "learning_rate": 2e-05,
+      "loss": 0.05500904,
+      "step": 20506
+    },
+    {
+      "epoch": 41.014,
+      "grad_norm": 1.4644334316253662,
+      "learning_rate": 2e-05,
+      "loss": 0.03757091,
+      "step": 20507
+    },
+    {
+      "epoch": 41.016,
+      "grad_norm": 1.2504773139953613,
+      "learning_rate": 2e-05,
+      "loss": 0.04865737,
+      "step": 20508
+    },
+    {
+      "epoch": 41.018,
+      "grad_norm": 1.0606939792633057,
+      "learning_rate": 2e-05,
+      "loss": 0.046809,
+      "step": 20509
+    },
+    {
+      "epoch": 41.02,
+      "grad_norm": 0.9825658798217773,
+      "learning_rate": 2e-05,
+      "loss": 0.03104939,
+      "step": 20510
+    },
+    {
+      "epoch": 41.022,
+      "grad_norm": 0.9851643443107605,
+      "learning_rate": 2e-05,
+      "loss": 0.03083806,
+      "step": 20511
+    },
+    {
+      "epoch": 41.024,
+      "grad_norm": 1.2548247575759888,
+      "learning_rate": 2e-05,
+      "loss": 0.05934072,
+      "step": 20512
+    },
+    {
+      "epoch": 41.026,
+      "grad_norm": 2.8248443603515625,
+      "learning_rate": 2e-05,
+      "loss": 0.05007793,
+      "step": 20513
+    },
+    {
+      "epoch": 41.028,
+      "grad_norm": 1.832990288734436,
+      "learning_rate": 2e-05,
+      "loss": 0.0635509,
+      "step": 20514
+    },
+    {
+      "epoch": 41.03,
+      "grad_norm": 1.1892746686935425,
+      "learning_rate": 2e-05,
+      "loss": 0.04696298,
+      "step": 20515
+    },
+    {
+      "epoch": 41.032,
+      "grad_norm": 1.8308331966400146,
+      "learning_rate": 2e-05,
+      "loss": 0.04524075,
+      "step": 20516
+    },
+    {
+      "epoch": 41.034,
+      "grad_norm": 1.3011291027069092,
+      "learning_rate": 2e-05,
+      "loss": 0.04992253,
+      "step": 20517
+    },
+    {
+      "epoch": 41.036,
+      "grad_norm": 1.1970986127853394,
+      "learning_rate": 2e-05,
+      "loss": 0.05594495,
+      "step": 20518
+    },
+    {
+      "epoch": 41.038,
+      "grad_norm": 1.1719766855239868,
+      "learning_rate": 2e-05,
+      "loss": 0.04220466,
+      "step": 20519
+    },
+    {
+      "epoch": 41.04,
+      "grad_norm": 1.5280590057373047,
+      "learning_rate": 2e-05,
+      "loss": 0.05499358,
+      "step": 20520
+    },
+    {
+      "epoch": 41.042,
+      "grad_norm": 1.2882308959960938,
+      "learning_rate": 2e-05,
+      "loss": 0.04562774,
+      "step": 20521
+    },
+    {
+      "epoch": 41.044,
+      "grad_norm": 2.0022976398468018,
+      "learning_rate": 2e-05,
+      "loss": 0.04248636,
+      "step": 20522
+    },
+    {
+      "epoch": 41.046,
+      "grad_norm": 1.5285600423812866,
+      "learning_rate": 2e-05,
+      "loss": 0.06069652,
+      "step": 20523
+    },
+    {
+      "epoch": 41.048,
+      "grad_norm": 1.181755781173706,
+      "learning_rate": 2e-05,
+      "loss": 0.04825043,
+      "step": 20524
+    },
+    {
+      "epoch": 41.05,
+      "grad_norm": 1.338629126548767,
+      "learning_rate": 2e-05,
+      "loss": 0.06323239,
+      "step": 20525
+    },
+    {
+      "epoch": 41.052,
+      "grad_norm": 1.1997042894363403,
+      "learning_rate": 2e-05,
+      "loss": 0.05054116,
+      "step": 20526
+    },
+    {
+      "epoch": 41.054,
+      "grad_norm": 1.3715033531188965,
+      "learning_rate": 2e-05,
+      "loss": 0.04951176,
+      "step": 20527
+    },
+    {
+      "epoch": 41.056,
+      "grad_norm": 1.3828208446502686,
+      "learning_rate": 2e-05,
+      "loss": 0.04772475,
+      "step": 20528
+    },
+    {
+      "epoch": 41.058,
+      "grad_norm": 1.0656706094741821,
+      "learning_rate": 2e-05,
+      "loss": 0.04354571,
+      "step": 20529
+    },
+    {
+      "epoch": 41.06,
+      "grad_norm": 0.9382243156433105,
+      "learning_rate": 2e-05,
+      "loss": 0.03308392,
+      "step": 20530
+    },
+    {
+      "epoch": 41.062,
+      "grad_norm": 1.2257845401763916,
+      "learning_rate": 2e-05,
+      "loss": 0.05354673,
+      "step": 20531
+    },
+    {
+      "epoch": 41.064,
+      "grad_norm": 1.0932214260101318,
+      "learning_rate": 2e-05,
+      "loss": 0.05004593,
+      "step": 20532
+    },
+    {
+      "epoch": 41.066,
+      "grad_norm": 1.3086014986038208,
+      "learning_rate": 2e-05,
+      "loss": 0.05718559,
+      "step": 20533
+    },
+    {
+      "epoch": 41.068,
+      "grad_norm": 1.2380306720733643,
+      "learning_rate": 2e-05,
+      "loss": 0.05476848,
+      "step": 20534
+    },
+    {
+      "epoch": 41.07,
+      "grad_norm": 1.06734037399292,
+      "learning_rate": 2e-05,
+      "loss": 0.03656657,
+      "step": 20535
+    },
+    {
+      "epoch": 41.072,
+      "grad_norm": 1.3546667098999023,
+      "learning_rate": 2e-05,
+      "loss": 0.05762174,
+      "step": 20536
+    },
+    {
+      "epoch": 41.074,
+      "grad_norm": 1.0288851261138916,
+      "learning_rate": 2e-05,
+      "loss": 0.0424634,
+      "step": 20537
+    },
+    {
+      "epoch": 41.076,
+      "grad_norm": 1.131334900856018,
+      "learning_rate": 2e-05,
+      "loss": 0.05600655,
+      "step": 20538
+    },
+    {
+      "epoch": 41.078,
+      "grad_norm": 1.6864545345306396,
+      "learning_rate": 2e-05,
+      "loss": 0.04475303,
+      "step": 20539
+    },
+    {
+      "epoch": 41.08,
+      "grad_norm": 1.9584969282150269,
+      "learning_rate": 2e-05,
+      "loss": 0.04874922,
+      "step": 20540
+    },
+    {
+      "epoch": 41.082,
+      "grad_norm": 1.2000778913497925,
+      "learning_rate": 2e-05,
+      "loss": 0.04340697,
+      "step": 20541
+    },
+    {
+      "epoch": 41.084,
+      "grad_norm": 1.0453670024871826,
+      "learning_rate": 2e-05,
+      "loss": 0.05378638,
+      "step": 20542
+    },
+    {
+      "epoch": 41.086,
+      "grad_norm": 1.0404855012893677,
+      "learning_rate": 2e-05,
+      "loss": 0.0372684,
+      "step": 20543
+    },
+    {
+      "epoch": 41.088,
+      "grad_norm": 1.7498598098754883,
+      "learning_rate": 2e-05,
+      "loss": 0.04698805,
+      "step": 20544
+    },
+    {
+      "epoch": 41.09,
+      "grad_norm": 1.0921574831008911,
+      "learning_rate": 2e-05,
+      "loss": 0.04534142,
+      "step": 20545
+    },
+    {
+      "epoch": 41.092,
+      "grad_norm": 1.1065009832382202,
+      "learning_rate": 2e-05,
+      "loss": 0.05014817,
+      "step": 20546
+    },
+    {
+      "epoch": 41.094,
+      "grad_norm": 1.1455692052841187,
+      "learning_rate": 2e-05,
+      "loss": 0.06051344,
+      "step": 20547
+    },
+    {
+      "epoch": 41.096,
+      "grad_norm": 1.4879392385482788,
+      "learning_rate": 2e-05,
+      "loss": 0.04928049,
+      "step": 20548
+    },
+    {
+      "epoch": 41.098,
+      "grad_norm": 1.4718266725540161,
+      "learning_rate": 2e-05,
+      "loss": 0.05406002,
+      "step": 20549
+    },
+    {
+      "epoch": 41.1,
+      "grad_norm": 1.2394520044326782,
+      "learning_rate": 2e-05,
+      "loss": 0.04753721,
+      "step": 20550
+    },
+    {
+      "epoch": 41.102,
+      "grad_norm": 1.0697596073150635,
+      "learning_rate": 2e-05,
+      "loss": 0.03903735,
+      "step": 20551
+    },
+    {
+      "epoch": 41.104,
+      "grad_norm": 1.3443745374679565,
+      "learning_rate": 2e-05,
+      "loss": 0.0416825,
+      "step": 20552
+    },
+    {
+      "epoch": 41.106,
+      "grad_norm": 1.0805517435073853,
+      "learning_rate": 2e-05,
+      "loss": 0.02990261,
+      "step": 20553
+    },
+    {
+      "epoch": 41.108,
+      "grad_norm": 1.2133495807647705,
+      "learning_rate": 2e-05,
+      "loss": 0.04597092,
+      "step": 20554
+    },
+    {
+      "epoch": 41.11,
+      "grad_norm": 2.3126041889190674,
+      "learning_rate": 2e-05,
+      "loss": 0.04624292,
+      "step": 20555
+    },
+    {
+      "epoch": 41.112,
+      "grad_norm": 1.1802953481674194,
+      "learning_rate": 2e-05,
+      "loss": 0.05155454,
+      "step": 20556
+    },
+    {
+      "epoch": 41.114,
+      "grad_norm": 1.4480260610580444,
+      "learning_rate": 2e-05,
+      "loss": 0.0588602,
+      "step": 20557
+    },
+    {
+      "epoch": 41.116,
+      "grad_norm": 1.3759326934814453,
+      "learning_rate": 2e-05,
+      "loss": 0.04992274,
+      "step": 20558
+    },
+    {
+      "epoch": 41.118,
+      "grad_norm": 1.1570967435836792,
+      "learning_rate": 2e-05,
+      "loss": 0.04519217,
+      "step": 20559
+    },
+    {
+      "epoch": 41.12,
+      "grad_norm": 1.2483819723129272,
+      "learning_rate": 2e-05,
+      "loss": 0.04176389,
+      "step": 20560
+    },
+    {
+      "epoch": 41.122,
+      "grad_norm": 1.4002913236618042,
+      "learning_rate": 2e-05,
+      "loss": 0.04440395,
+      "step": 20561
+    },
+    {
+      "epoch": 41.124,
+      "grad_norm": 1.4284930229187012,
+      "learning_rate": 2e-05,
+      "loss": 0.05776929,
+      "step": 20562
+    },
+    {
+      "epoch": 41.126,
+      "grad_norm": 1.3323142528533936,
+      "learning_rate": 2e-05,
+      "loss": 0.0383847,
+      "step": 20563
+    },
+    {
+      "epoch": 41.128,
+      "grad_norm": 1.1827408075332642,
+      "learning_rate": 2e-05,
+      "loss": 0.0447496,
+      "step": 20564
+    },
+    {
+      "epoch": 41.13,
+      "grad_norm": 1.2362679243087769,
+      "learning_rate": 2e-05,
+      "loss": 0.04746515,
+      "step": 20565
+    },
+    {
+      "epoch": 41.132,
+      "grad_norm": 1.9688647985458374,
+      "learning_rate": 2e-05,
+      "loss": 0.04489559,
+      "step": 20566
+    },
+    {
+      "epoch": 41.134,
+      "grad_norm": 1.304279088973999,
+      "learning_rate": 2e-05,
+      "loss": 0.04824959,
+      "step": 20567
+    },
+    {
+      "epoch": 41.136,
+      "grad_norm": 1.7030314207077026,
+      "learning_rate": 2e-05,
+      "loss": 0.04330122,
+      "step": 20568
+    },
+    {
+      "epoch": 41.138,
+      "grad_norm": 1.1935453414916992,
+      "learning_rate": 2e-05,
+      "loss": 0.05471293,
+      "step": 20569
+    },
+    {
+      "epoch": 41.14,
+      "grad_norm": 1.0641679763793945,
+      "learning_rate": 2e-05,
+      "loss": 0.04133367,
+      "step": 20570
+    },
+    {
+      "epoch": 41.142,
+      "grad_norm": 0.9642373323440552,
+      "learning_rate": 2e-05,
+      "loss": 0.03811979,
+      "step": 20571
+    },
+    {
+      "epoch": 41.144,
+      "grad_norm": 1.7914299964904785,
+      "learning_rate": 2e-05,
+      "loss": 0.04343174,
+      "step": 20572
+    },
+    {
+      "epoch": 41.146,
+      "grad_norm": 1.4745582342147827,
+      "learning_rate": 2e-05,
+      "loss": 0.03861541,
+      "step": 20573
+    },
+    {
+      "epoch": 41.148,
+      "grad_norm": 0.9670395255088806,
+      "learning_rate": 2e-05,
+      "loss": 0.03732292,
+      "step": 20574
+    },
+    {
+      "epoch": 41.15,
+      "grad_norm": 1.2237446308135986,
+      "learning_rate": 2e-05,
+      "loss": 0.04279939,
+      "step": 20575
+    },
+    {
+      "epoch": 41.152,
+      "grad_norm": 1.173431634902954,
+      "learning_rate": 2e-05,
+      "loss": 0.05236944,
+      "step": 20576
+    },
+    {
+      "epoch": 41.154,
+      "grad_norm": 2.5610194206237793,
+      "learning_rate": 2e-05,
+      "loss": 0.05082991,
+      "step": 20577
+    },
+    {
+      "epoch": 41.156,
+      "grad_norm": 1.1480480432510376,
+      "learning_rate": 2e-05,
+      "loss": 0.03853033,
+      "step": 20578
+    },
+    {
+      "epoch": 41.158,
+      "grad_norm": 1.0263923406600952,
+      "learning_rate": 2e-05,
+      "loss": 0.0398251,
+      "step": 20579
+    },
+    {
+      "epoch": 41.16,
+      "grad_norm": 1.241070032119751,
+      "learning_rate": 2e-05,
+      "loss": 0.04581333,
+      "step": 20580
+    },
+    {
+      "epoch": 41.162,
+      "grad_norm": 0.9623987674713135,
+      "learning_rate": 2e-05,
+      "loss": 0.02442548,
+      "step": 20581
+    },
+    {
+      "epoch": 41.164,
+      "grad_norm": 1.2360503673553467,
+      "learning_rate": 2e-05,
+      "loss": 0.04588886,
+      "step": 20582
+    },
+    {
+      "epoch": 41.166,
+      "grad_norm": 1.4302246570587158,
+      "learning_rate": 2e-05,
+      "loss": 0.0555368,
+      "step": 20583
+    },
+    {
+      "epoch": 41.168,
+      "grad_norm": 1.1271694898605347,
+      "learning_rate": 2e-05,
+      "loss": 0.04541831,
+      "step": 20584
+    },
+    {
+      "epoch": 41.17,
+      "grad_norm": 1.3913869857788086,
+      "learning_rate": 2e-05,
+      "loss": 0.05388426,
+      "step": 20585
+    },
+    {
+      "epoch": 41.172,
+      "grad_norm": 1.4658206701278687,
+      "learning_rate": 2e-05,
+      "loss": 0.05629757,
+      "step": 20586
+    },
+    {
+      "epoch": 41.174,
+      "grad_norm": 1.5316065549850464,
+      "learning_rate": 2e-05,
+      "loss": 0.04545237,
+      "step": 20587
+    },
+    {
+      "epoch": 41.176,
+      "grad_norm": 1.3376739025115967,
+      "learning_rate": 2e-05,
+      "loss": 0.04239886,
+      "step": 20588
+    },
+    {
+      "epoch": 41.178,
+      "grad_norm": 1.168369174003601,
+      "learning_rate": 2e-05,
+      "loss": 0.04730329,
+      "step": 20589
+    },
+    {
+      "epoch": 41.18,
+      "grad_norm": 1.132712483406067,
+      "learning_rate": 2e-05,
+      "loss": 0.0453882,
+      "step": 20590
+    },
+    {
+      "epoch": 41.182,
+      "grad_norm": 1.2484111785888672,
+      "learning_rate": 2e-05,
+      "loss": 0.05077428,
+      "step": 20591
+    },
+    {
+      "epoch": 41.184,
+      "grad_norm": 1.2351433038711548,
+      "learning_rate": 2e-05,
+      "loss": 0.04053054,
+      "step": 20592
+    },
+    {
+      "epoch": 41.186,
+      "grad_norm": 1.127816081047058,
+      "learning_rate": 2e-05,
+      "loss": 0.03557635,
+      "step": 20593
+    },
+    {
+      "epoch": 41.188,
+      "grad_norm": 1.6981974840164185,
+      "learning_rate": 2e-05,
+      "loss": 0.05980735,
+      "step": 20594
+    },
+    {
+      "epoch": 41.19,
+      "grad_norm": 1.2019031047821045,
+      "learning_rate": 2e-05,
+      "loss": 0.04414729,
+      "step": 20595
+    },
+    {
+      "epoch": 41.192,
+      "grad_norm": 2.026304244995117,
+      "learning_rate": 2e-05,
+      "loss": 0.06682765,
+      "step": 20596
+    },
+    {
+      "epoch": 41.194,
+      "grad_norm": 1.1906565427780151,
+      "learning_rate": 2e-05,
+      "loss": 0.04870375,
+      "step": 20597
+    },
+    {
+      "epoch": 41.196,
+      "grad_norm": 1.194569706916809,
+      "learning_rate": 2e-05,
+      "loss": 0.05204153,
+      "step": 20598
+    },
+    {
+      "epoch": 41.198,
+      "grad_norm": 1.4061965942382812,
+      "learning_rate": 2e-05,
+      "loss": 0.05822208,
+      "step": 20599
+    },
+    {
+      "epoch": 41.2,
+      "grad_norm": 1.2633558511734009,
+      "learning_rate": 2e-05,
+      "loss": 0.0512266,
+      "step": 20600
+    },
+    {
+      "epoch": 41.202,
+      "grad_norm": 1.4011318683624268,
+      "learning_rate": 2e-05,
+      "loss": 0.05827274,
+      "step": 20601
+    },
+    {
+      "epoch": 41.204,
+      "grad_norm": 1.0192710161209106,
+      "learning_rate": 2e-05,
+      "loss": 0.03732585,
+      "step": 20602
+    },
+    {
+      "epoch": 41.206,
+      "grad_norm": 1.1007825136184692,
+      "learning_rate": 2e-05,
+      "loss": 0.04062988,
+      "step": 20603
+    },
+    {
+      "epoch": 41.208,
+      "grad_norm": 1.4489682912826538,
+      "learning_rate": 2e-05,
+      "loss": 0.04679845,
+      "step": 20604
+    },
+    {
+      "epoch": 41.21,
+      "grad_norm": 1.3930132389068604,
+      "learning_rate": 2e-05,
+      "loss": 0.04038769,
+      "step": 20605
+    },
+    {
+      "epoch": 41.212,
+      "grad_norm": 1.1776801347732544,
+      "learning_rate": 2e-05,
+      "loss": 0.05422183,
+      "step": 20606
+    },
+    {
+      "epoch": 41.214,
+      "grad_norm": 1.3811092376708984,
+      "learning_rate": 2e-05,
+      "loss": 0.05448051,
+      "step": 20607
+    },
+    {
+      "epoch": 41.216,
+      "grad_norm": 1.6993544101715088,
+      "learning_rate": 2e-05,
+      "loss": 0.05991393,
+      "step": 20608
+    },
+    {
+      "epoch": 41.218,
+      "grad_norm": 1.1804338693618774,
+      "learning_rate": 2e-05,
+      "loss": 0.0513094,
+      "step": 20609
+    },
+    {
+      "epoch": 41.22,
+      "grad_norm": 1.2084537744522095,
+      "learning_rate": 2e-05,
+      "loss": 0.04306343,
+      "step": 20610
+    },
+    {
+      "epoch": 41.222,
+      "grad_norm": 2.1294620037078857,
+      "learning_rate": 2e-05,
+      "loss": 0.05107892,
+      "step": 20611
+    },
+    {
+      "epoch": 41.224,
+      "grad_norm": 1.1537929773330688,
+      "learning_rate": 2e-05,
+      "loss": 0.04525739,
+      "step": 20612
+    },
+    {
+      "epoch": 41.226,
+      "grad_norm": 1.0452954769134521,
+      "learning_rate": 2e-05,
+      "loss": 0.03667298,
+      "step": 20613
+    },
+    {
+      "epoch": 41.228,
+      "grad_norm": 1.2010400295257568,
+      "learning_rate": 2e-05,
+      "loss": 0.03450677,
+      "step": 20614
+    },
+    {
+      "epoch": 41.23,
+      "grad_norm": 1.1608492136001587,
+      "learning_rate": 2e-05,
+      "loss": 0.0470752,
+      "step": 20615
+    },
+    {
+      "epoch": 41.232,
+      "grad_norm": 1.367234706878662,
+      "learning_rate": 2e-05,
+      "loss": 0.04004624,
+      "step": 20616
+    },
+    {
+      "epoch": 41.234,
+      "grad_norm": 3.438629388809204,
+      "learning_rate": 2e-05,
+      "loss": 0.05718534,
+      "step": 20617
+    },
+    {
+      "epoch": 41.236,
+      "grad_norm": 1.245473861694336,
+      "learning_rate": 2e-05,
+      "loss": 0.04939371,
+      "step": 20618
+    },
+    {
+      "epoch": 41.238,
+      "grad_norm": 1.5737378597259521,
+      "learning_rate": 2e-05,
+      "loss": 0.06649335,
+      "step": 20619
+    },
+    {
+      "epoch": 41.24,
+      "grad_norm": 2.474226951599121,
+      "learning_rate": 2e-05,
+      "loss": 0.05714843,
+      "step": 20620
+    },
+    {
+      "epoch": 41.242,
+      "grad_norm": 1.1201341152191162,
+      "learning_rate": 2e-05,
+      "loss": 0.03632017,
+      "step": 20621
+    },
+    {
+      "epoch": 41.244,
+      "grad_norm": 1.3226417303085327,
+      "learning_rate": 2e-05,
+      "loss": 0.05535428,
+      "step": 20622
+    },
+    {
+      "epoch": 41.246,
+      "grad_norm": 1.5820775032043457,
+      "learning_rate": 2e-05,
+      "loss": 0.06497006,
+      "step": 20623
+    },
+    {
+      "epoch": 41.248,
+      "grad_norm": 1.6483792066574097,
+      "learning_rate": 2e-05,
+      "loss": 0.04980063,
+      "step": 20624
+    },
+    {
+      "epoch": 41.25,
+      "grad_norm": 0.8462343215942383,
+      "learning_rate": 2e-05,
+      "loss": 0.0305022,
+      "step": 20625
+    },
+    {
+      "epoch": 41.252,
+      "grad_norm": 1.3294532299041748,
+      "learning_rate": 2e-05,
+      "loss": 0.05518986,
+      "step": 20626
+    },
+    {
+      "epoch": 41.254,
+      "grad_norm": 1.0134047269821167,
+      "learning_rate": 2e-05,
+      "loss": 0.04463737,
+      "step": 20627
+    },
+    {
+      "epoch": 41.256,
+      "grad_norm": 1.3394545316696167,
+      "learning_rate": 2e-05,
+      "loss": 0.06053488,
+      "step": 20628
+    },
+    {
+      "epoch": 41.258,
+      "grad_norm": 1.1298450231552124,
+      "learning_rate": 2e-05,
+      "loss": 0.03974798,
+      "step": 20629
+    },
+    {
+      "epoch": 41.26,
+      "grad_norm": 1.2260725498199463,
+      "learning_rate": 2e-05,
+      "loss": 0.04857574,
+      "step": 20630
+    },
+    {
+      "epoch": 41.262,
+      "grad_norm": 1.2204060554504395,
+      "learning_rate": 2e-05,
+      "loss": 0.04478389,
+      "step": 20631
+    },
+    {
+      "epoch": 41.264,
+      "grad_norm": 2.948246479034424,
+      "learning_rate": 2e-05,
+      "loss": 0.05024678,
+      "step": 20632
+    },
+    {
+      "epoch": 41.266,
+      "grad_norm": 1.3143848180770874,
+      "learning_rate": 2e-05,
+      "loss": 0.04200232,
+      "step": 20633
+    },
+    {
+      "epoch": 41.268,
+      "grad_norm": 1.1079685688018799,
+      "learning_rate": 2e-05,
+      "loss": 0.03887783,
+      "step": 20634
+    },
+    {
+      "epoch": 41.27,
+      "grad_norm": 1.3387101888656616,
+      "learning_rate": 2e-05,
+      "loss": 0.04824906,
+      "step": 20635
+    },
+    {
+      "epoch": 41.272,
+      "grad_norm": 1.2918139696121216,
+      "learning_rate": 2e-05,
+      "loss": 0.05828239,
+      "step": 20636
+    },
+    {
+      "epoch": 41.274,
+      "grad_norm": 1.2324050664901733,
+      "learning_rate": 2e-05,
+      "loss": 0.05050058,
+      "step": 20637
+    },
+    {
+      "epoch": 41.276,
+      "grad_norm": 1.08914315700531,
+      "learning_rate": 2e-05,
+      "loss": 0.03811881,
+      "step": 20638
+    },
+    {
+      "epoch": 41.278,
+      "grad_norm": 1.5355799198150635,
+      "learning_rate": 2e-05,
+      "loss": 0.06125507,
+      "step": 20639
+    },
+    {
+      "epoch": 41.28,
+      "grad_norm": 1.1275166273117065,
+      "learning_rate": 2e-05,
+      "loss": 0.04609001,
+      "step": 20640
+    },
+    {
+      "epoch": 41.282,
+      "grad_norm": 1.394644021987915,
+      "learning_rate": 2e-05,
+      "loss": 0.05548574,
+      "step": 20641
+    },
+    {
+      "epoch": 41.284,
+      "grad_norm": 1.1803781986236572,
+      "learning_rate": 2e-05,
+      "loss": 0.04673609,
+      "step": 20642
+    },
+    {
+      "epoch": 41.286,
+      "grad_norm": 1.088181972503662,
+      "learning_rate": 2e-05,
+      "loss": 0.04835239,
+      "step": 20643
+    },
+    {
+      "epoch": 41.288,
+      "grad_norm": 1.2936146259307861,
+      "learning_rate": 2e-05,
+      "loss": 0.0537074,
+      "step": 20644
+    },
+    {
+      "epoch": 41.29,
+      "grad_norm": 1.0376815795898438,
+      "learning_rate": 2e-05,
+      "loss": 0.04589043,
+      "step": 20645
+    },
+    {
+      "epoch": 41.292,
+      "grad_norm": 1.2658627033233643,
+      "learning_rate": 2e-05,
+      "loss": 0.04795525,
+      "step": 20646
+    },
+    {
+      "epoch": 41.294,
+      "grad_norm": 1.5672342777252197,
+      "learning_rate": 2e-05,
+      "loss": 0.06098137,
+      "step": 20647
+    },
+    {
+      "epoch": 41.296,
+      "grad_norm": 1.1743220090866089,
+      "learning_rate": 2e-05,
+      "loss": 0.03908478,
+      "step": 20648
+    },
+    {
+      "epoch": 41.298,
+      "grad_norm": 1.368518590927124,
+      "learning_rate": 2e-05,
+      "loss": 0.03311605,
+      "step": 20649
+    },
+    {
+      "epoch": 41.3,
+      "grad_norm": 0.9655233025550842,
+      "learning_rate": 2e-05,
+      "loss": 0.03225183,
+      "step": 20650
+    },
+    {
+      "epoch": 41.302,
+      "grad_norm": 2.152411937713623,
+      "learning_rate": 2e-05,
+      "loss": 0.04487934,
+      "step": 20651
+    },
+    {
+      "epoch": 41.304,
+      "grad_norm": 1.6587146520614624,
+      "learning_rate": 2e-05,
+      "loss": 0.05637205,
+      "step": 20652
+    },
+    {
+      "epoch": 41.306,
+      "grad_norm": 1.024804949760437,
+      "learning_rate": 2e-05,
+      "loss": 0.03787657,
+      "step": 20653
+    },
+    {
+      "epoch": 41.308,
+      "grad_norm": 0.8710086941719055,
+      "learning_rate": 2e-05,
+      "loss": 0.02578922,
+      "step": 20654
+    },
+    {
+      "epoch": 41.31,
+      "grad_norm": 1.093598484992981,
+      "learning_rate": 2e-05,
+      "loss": 0.03931287,
+      "step": 20655
+    },
+    {
+      "epoch": 41.312,
+      "grad_norm": 1.3158975839614868,
+      "learning_rate": 2e-05,
+      "loss": 0.0523903,
+      "step": 20656
+    },
+    {
+      "epoch": 41.314,
+      "grad_norm": 1.2842652797698975,
+      "learning_rate": 2e-05,
+      "loss": 0.0551443,
+      "step": 20657
+    },
+    {
+      "epoch": 41.316,
+      "grad_norm": 0.9673425555229187,
+      "learning_rate": 2e-05,
+      "loss": 0.03543061,
+      "step": 20658
+    },
+    {
+      "epoch": 41.318,
+      "grad_norm": 1.4158899784088135,
+      "learning_rate": 2e-05,
+      "loss": 0.05106701,
+      "step": 20659
+    },
+    {
+      "epoch": 41.32,
+      "grad_norm": 1.7559031248092651,
+      "learning_rate": 2e-05,
+      "loss": 0.03872785,
+      "step": 20660
+    },
+    {
+      "epoch": 41.322,
+      "grad_norm": 1.1263608932495117,
+      "learning_rate": 2e-05,
+      "loss": 0.03599942,
+      "step": 20661
+    },
+    {
+      "epoch": 41.324,
+      "grad_norm": 0.9912429451942444,
+      "learning_rate": 2e-05,
+      "loss": 0.02866612,
+      "step": 20662
+    },
+    {
+      "epoch": 41.326,
+      "grad_norm": 1.0485600233078003,
+      "learning_rate": 2e-05,
+      "loss": 0.03809555,
+      "step": 20663
+    },
+    {
+      "epoch": 41.328,
+      "grad_norm": 2.218872308731079,
+      "learning_rate": 2e-05,
+      "loss": 0.05056027,
+      "step": 20664
+    },
+    {
+      "epoch": 41.33,
+      "grad_norm": 1.2448780536651611,
+      "learning_rate": 2e-05,
+      "loss": 0.06671609,
+      "step": 20665
+    },
+    {
+      "epoch": 41.332,
+      "grad_norm": 1.1735291481018066,
+      "learning_rate": 2e-05,
+      "loss": 0.04490344,
+      "step": 20666
+    },
+    {
+      "epoch": 41.334,
+      "grad_norm": 1.6742703914642334,
+      "learning_rate": 2e-05,
+      "loss": 0.06206613,
+      "step": 20667
+    },
+    {
+      "epoch": 41.336,
+      "grad_norm": 1.2419872283935547,
+      "learning_rate": 2e-05,
+      "loss": 0.03979301,
+      "step": 20668
+    },
+    {
+      "epoch": 41.338,
+      "grad_norm": 2.8356029987335205,
+      "learning_rate": 2e-05,
+      "loss": 0.05524603,
+      "step": 20669
+    },
+    {
+      "epoch": 41.34,
+      "grad_norm": 2.10284161567688,
+      "learning_rate": 2e-05,
+      "loss": 0.05853007,
+      "step": 20670
+    },
+    {
+      "epoch": 41.342,
+      "grad_norm": 1.0346118211746216,
+      "learning_rate": 2e-05,
+      "loss": 0.0342777,
+      "step": 20671
+    },
+    {
+      "epoch": 41.344,
+      "grad_norm": 1.1584864854812622,
+      "learning_rate": 2e-05,
+      "loss": 0.04522749,
+      "step": 20672
+    },
+    {
+      "epoch": 41.346,
+      "grad_norm": 1.0336374044418335,
+      "learning_rate": 2e-05,
+      "loss": 0.0341766,
+      "step": 20673
+    },
+    {
+      "epoch": 41.348,
+      "grad_norm": 1.2833257913589478,
+      "learning_rate": 2e-05,
+      "loss": 0.04493989,
+      "step": 20674
+    },
+    {
+      "epoch": 41.35,
+      "grad_norm": 1.196931004524231,
+      "learning_rate": 2e-05,
+      "loss": 0.03288671,
+      "step": 20675
+    },
+    {
+      "epoch": 41.352,
+      "grad_norm": 1.6481664180755615,
+      "learning_rate": 2e-05,
+      "loss": 0.04984744,
+      "step": 20676
+    },
+    {
+      "epoch": 41.354,
+      "grad_norm": 1.3201029300689697,
+      "learning_rate": 2e-05,
+      "loss": 0.05488572,
+      "step": 20677
+    },
+    {
+      "epoch": 41.356,
+      "grad_norm": 1.2787818908691406,
+      "learning_rate": 2e-05,
+      "loss": 0.0575133,
+      "step": 20678
+    },
+    {
+      "epoch": 41.358,
+      "grad_norm": 1.1592620611190796,
+      "learning_rate": 2e-05,
+      "loss": 0.04519965,
+      "step": 20679
+    },
+    {
+      "epoch": 41.36,
+      "grad_norm": 1.7062658071517944,
+      "learning_rate": 2e-05,
+      "loss": 0.04929303,
+      "step": 20680
+    },
+    {
+      "epoch": 41.362,
+      "grad_norm": 1.4524036645889282,
+      "learning_rate": 2e-05,
+      "loss": 0.05601212,
+      "step": 20681
+    },
+    {
+      "epoch": 41.364,
+      "grad_norm": 1.2186477184295654,
+      "learning_rate": 2e-05,
+      "loss": 0.0421271,
+      "step": 20682
+    },
+    {
+      "epoch": 41.366,
+      "grad_norm": 1.7632216215133667,
+      "learning_rate": 2e-05,
+      "loss": 0.05948129,
+      "step": 20683
+    },
+    {
+      "epoch": 41.368,
+      "grad_norm": 1.5571186542510986,
+      "learning_rate": 2e-05,
+      "loss": 0.04970782,
+      "step": 20684
+    },
+    {
+      "epoch": 41.37,
+      "grad_norm": 1.611167311668396,
+      "learning_rate": 2e-05,
+      "loss": 0.04187579,
+      "step": 20685
+    },
+    {
+      "epoch": 41.372,
+      "grad_norm": 1.0352783203125,
+      "learning_rate": 2e-05,
+      "loss": 0.02740442,
+      "step": 20686
+    },
+    {
+      "epoch": 41.374,
+      "grad_norm": 1.3376938104629517,
+      "learning_rate": 2e-05,
+      "loss": 0.06241329,
+      "step": 20687
+    },
+    {
+      "epoch": 41.376,
+      "grad_norm": 1.135635495185852,
+      "learning_rate": 2e-05,
+      "loss": 0.04911846,
+      "step": 20688
+    },
+    {
+      "epoch": 41.378,
+      "grad_norm": 1.2492220401763916,
+      "learning_rate": 2e-05,
+      "loss": 0.05272296,
+      "step": 20689
+    },
+    {
+      "epoch": 41.38,
+      "grad_norm": 1.3152111768722534,
+      "learning_rate": 2e-05,
+      "loss": 0.05239505,
+      "step": 20690
+    },
+    {
+      "epoch": 41.382,
+      "grad_norm": 1.3397350311279297,
+      "learning_rate": 2e-05,
+      "loss": 0.05067999,
+      "step": 20691
+    },
+    {
+      "epoch": 41.384,
+      "grad_norm": 1.1833877563476562,
+      "learning_rate": 2e-05,
+      "loss": 0.04897395,
+      "step": 20692
+    },
+    {
+      "epoch": 41.386,
+      "grad_norm": 1.2837327718734741,
+      "learning_rate": 2e-05,
+      "loss": 0.05838244,
+      "step": 20693
+    },
+    {
+      "epoch": 41.388,
+      "grad_norm": 1.150177240371704,
+      "learning_rate": 2e-05,
+      "loss": 0.04375577,
+      "step": 20694
+    },
+    {
+      "epoch": 41.39,
+      "grad_norm": 1.4795578718185425,
+      "learning_rate": 2e-05,
+      "loss": 0.05091795,
+      "step": 20695
+    },
+    {
+      "epoch": 41.392,
+      "grad_norm": 1.2051365375518799,
+      "learning_rate": 2e-05,
+      "loss": 0.05093031,
+      "step": 20696
+    },
+    {
+      "epoch": 41.394,
+      "grad_norm": 0.9962400197982788,
+      "learning_rate": 2e-05,
+      "loss": 0.04091921,
+      "step": 20697
+    },
+    {
+      "epoch": 41.396,
+      "grad_norm": 2.7451603412628174,
+      "learning_rate": 2e-05,
+      "loss": 0.05224776,
+      "step": 20698
+    },
+    {
+      "epoch": 41.398,
+      "grad_norm": 1.16928231716156,
+      "learning_rate": 2e-05,
+      "loss": 0.05099126,
+      "step": 20699
+    },
+    {
+      "epoch": 41.4,
+      "grad_norm": 1.571977972984314,
+      "learning_rate": 2e-05,
+      "loss": 0.04833336,
+      "step": 20700
+    },
+    {
+      "epoch": 41.402,
+      "grad_norm": 1.3276727199554443,
+      "learning_rate": 2e-05,
+      "loss": 0.04746715,
+      "step": 20701
+    },
+    {
+      "epoch": 41.404,
+      "grad_norm": 1.438899040222168,
+      "learning_rate": 2e-05,
+      "loss": 0.05122091,
+      "step": 20702
+    },
+    {
+      "epoch": 41.406,
+      "grad_norm": 1.1340113878250122,
+      "learning_rate": 2e-05,
+      "loss": 0.04921956,
+      "step": 20703
+    },
+    {
+      "epoch": 41.408,
+      "grad_norm": 1.3475137948989868,
+      "learning_rate": 2e-05,
+      "loss": 0.0532689,
+      "step": 20704
+    },
+    {
+      "epoch": 41.41,
+      "grad_norm": 3.48881196975708,
+      "learning_rate": 2e-05,
+      "loss": 0.05500069,
+      "step": 20705
+    },
+    {
+      "epoch": 41.412,
+      "grad_norm": 1.010732889175415,
+      "learning_rate": 2e-05,
+      "loss": 0.03719215,
+      "step": 20706
+    },
+    {
+      "epoch": 41.414,
+      "grad_norm": 1.2824602127075195,
+      "learning_rate": 2e-05,
+      "loss": 0.05517687,
+      "step": 20707
+    },
+    {
+      "epoch": 41.416,
+      "grad_norm": 1.1974467039108276,
+      "learning_rate": 2e-05,
+      "loss": 0.03475419,
+      "step": 20708
+    },
+    {
+      "epoch": 41.418,
+      "grad_norm": 0.9742454290390015,
+      "learning_rate": 2e-05,
+      "loss": 0.03150909,
+      "step": 20709
+    },
+    {
+      "epoch": 41.42,
+      "grad_norm": 4.395928382873535,
+      "learning_rate": 2e-05,
+      "loss": 0.06255428,
+      "step": 20710
+    },
+    {
+      "epoch": 41.422,
+      "grad_norm": 2.005077362060547,
+      "learning_rate": 2e-05,
+      "loss": 0.05040678,
+      "step": 20711
+    },
+    {
+      "epoch": 41.424,
+      "grad_norm": 1.1284871101379395,
+      "learning_rate": 2e-05,
+      "loss": 0.04268694,
+      "step": 20712
+    },
+    {
+      "epoch": 41.426,
+      "grad_norm": 1.1108037233352661,
+      "learning_rate": 2e-05,
+      "loss": 0.05018146,
+      "step": 20713
+    },
+    {
+      "epoch": 41.428,
+      "grad_norm": 1.2900370359420776,
+      "learning_rate": 2e-05,
+      "loss": 0.05577334,
+      "step": 20714
+    },
+    {
+      "epoch": 41.43,
+      "grad_norm": 1.625792145729065,
+      "learning_rate": 2e-05,
+      "loss": 0.0519428,
+      "step": 20715
+    },
+    {
+      "epoch": 41.432,
+      "grad_norm": 1.2618361711502075,
+      "learning_rate": 2e-05,
+      "loss": 0.06108624,
+      "step": 20716
+    },
+    {
+      "epoch": 41.434,
+      "grad_norm": 2.6890978813171387,
+      "learning_rate": 2e-05,
+      "loss": 0.05818431,
+      "step": 20717
+    },
+    {
+      "epoch": 41.436,
+      "grad_norm": 1.5866303443908691,
+      "learning_rate": 2e-05,
+      "loss": 0.05809556,
+      "step": 20718
+    },
+    {
+      "epoch": 41.438,
+      "grad_norm": 3.1994481086730957,
+      "learning_rate": 2e-05,
+      "loss": 0.03941204,
+      "step": 20719
+    },
+    {
+      "epoch": 41.44,
+      "grad_norm": 0.9686678051948547,
+      "learning_rate": 2e-05,
+      "loss": 0.03777551,
+      "step": 20720
+    },
+    {
+      "epoch": 41.442,
+      "grad_norm": 1.2619547843933105,
+      "learning_rate": 2e-05,
+      "loss": 0.05709205,
+      "step": 20721
+    },
+    {
+      "epoch": 41.444,
+      "grad_norm": 1.1282532215118408,
+      "learning_rate": 2e-05,
+      "loss": 0.03835287,
+      "step": 20722
+    },
+    {
+      "epoch": 41.446,
+      "grad_norm": 1.2010022401809692,
+      "learning_rate": 2e-05,
+      "loss": 0.04767352,
+      "step": 20723
+    },
+    {
+      "epoch": 41.448,
+      "grad_norm": 1.447494387626648,
+      "learning_rate": 2e-05,
+      "loss": 0.05089083,
+      "step": 20724
+    },
+    {
+      "epoch": 41.45,
+      "grad_norm": 1.101004719734192,
+      "learning_rate": 2e-05,
+      "loss": 0.0408543,
+      "step": 20725
+    },
+    {
+      "epoch": 41.452,
+      "grad_norm": 1.2500224113464355,
+      "learning_rate": 2e-05,
+      "loss": 0.03975309,
+      "step": 20726
+    },
+    {
+      "epoch": 41.454,
+      "grad_norm": 1.746552586555481,
+      "learning_rate": 2e-05,
+      "loss": 0.06327257,
+      "step": 20727
+    },
+    {
+      "epoch": 41.456,
+      "grad_norm": 1.121049165725708,
+      "learning_rate": 2e-05,
+      "loss": 0.03917,
+      "step": 20728
+    },
+    {
+      "epoch": 41.458,
+      "grad_norm": 1.5828957557678223,
+      "learning_rate": 2e-05,
+      "loss": 0.04092988,
+      "step": 20729
+    },
+    {
+      "epoch": 41.46,
+      "grad_norm": 1.3952829837799072,
+      "learning_rate": 2e-05,
+      "loss": 0.05349658,
+      "step": 20730
+    },
+    {
+      "epoch": 41.462,
+      "grad_norm": 1.1374624967575073,
+      "learning_rate": 2e-05,
+      "loss": 0.04330311,
+      "step": 20731
+    },
+    {
+      "epoch": 41.464,
+      "grad_norm": 2.1786279678344727,
+      "learning_rate": 2e-05,
+      "loss": 0.04306266,
+      "step": 20732
+    },
+    {
+      "epoch": 41.466,
+      "grad_norm": 1.1938217878341675,
+      "learning_rate": 2e-05,
+      "loss": 0.04537656,
+      "step": 20733
+    },
+    {
+      "epoch": 41.468,
+      "grad_norm": 1.673171877861023,
+      "learning_rate": 2e-05,
+      "loss": 0.04567283,
+      "step": 20734
+    },
+    {
+      "epoch": 41.47,
+      "grad_norm": 1.8601455688476562,
+      "learning_rate": 2e-05,
+      "loss": 0.04573951,
+      "step": 20735
+    },
+    {
+      "epoch": 41.472,
+      "grad_norm": 1.2552533149719238,
+      "learning_rate": 2e-05,
+      "loss": 0.05038974,
+      "step": 20736
+    },
+    {
+      "epoch": 41.474,
+      "grad_norm": 1.2557790279388428,
+      "learning_rate": 2e-05,
+      "loss": 0.04997993,
+      "step": 20737
+    },
+    {
+      "epoch": 41.476,
+      "grad_norm": 1.156934142112732,
+      "learning_rate": 2e-05,
+      "loss": 0.04020359,
+      "step": 20738
+    },
+    {
+      "epoch": 41.478,
+      "grad_norm": 1.0187697410583496,
+      "learning_rate": 2e-05,
+      "loss": 0.0350149,
+      "step": 20739
+    },
+    {
+      "epoch": 41.48,
+      "grad_norm": 1.5980960130691528,
+      "learning_rate": 2e-05,
+      "loss": 0.05507297,
+      "step": 20740
+    },
+    {
+      "epoch": 41.482,
+      "grad_norm": 1.2604917287826538,
+      "learning_rate": 2e-05,
+      "loss": 0.04280219,
+      "step": 20741
+    },
+    {
+      "epoch": 41.484,
+      "grad_norm": 1.2176854610443115,
+      "learning_rate": 2e-05,
+      "loss": 0.04632607,
+      "step": 20742
+    },
+    {
+      "epoch": 41.486,
+      "grad_norm": 1.7439926862716675,
+      "learning_rate": 2e-05,
+      "loss": 0.06102863,
+      "step": 20743
+    },
+    {
+      "epoch": 41.488,
+      "grad_norm": 1.1306809186935425,
+      "learning_rate": 2e-05,
+      "loss": 0.04056504,
+      "step": 20744
+    },
+    {
+      "epoch": 41.49,
+      "grad_norm": 1.6267520189285278,
+      "learning_rate": 2e-05,
+      "loss": 0.07167388,
+      "step": 20745
+    },
+    {
+      "epoch": 41.492,
+      "grad_norm": 1.483768343925476,
+      "learning_rate": 2e-05,
+      "loss": 0.03321824,
+      "step": 20746
+    },
+    {
+      "epoch": 41.494,
+      "grad_norm": 1.3936223983764648,
+      "learning_rate": 2e-05,
+      "loss": 0.06460038,
+      "step": 20747
+    },
+    {
+      "epoch": 41.496,
+      "grad_norm": 1.23568594455719,
+      "learning_rate": 2e-05,
+      "loss": 0.06132279,
+      "step": 20748
+    },
+    {
+      "epoch": 41.498,
+      "grad_norm": 1.2954548597335815,
+      "learning_rate": 2e-05,
+      "loss": 0.04977329,
+      "step": 20749
+    },
+    {
+      "epoch": 41.5,
+      "grad_norm": 1.190082311630249,
+      "learning_rate": 2e-05,
+      "loss": 0.05261871,
+      "step": 20750
+    },
+    {
+      "epoch": 41.502,
+      "grad_norm": 1.1780047416687012,
+      "learning_rate": 2e-05,
+      "loss": 0.03741407,
+      "step": 20751
+    },
+    {
+      "epoch": 41.504,
+      "grad_norm": 1.4033395051956177,
+      "learning_rate": 2e-05,
+      "loss": 0.04731531,
+      "step": 20752
+    },
+    {
+      "epoch": 41.506,
+      "grad_norm": 0.8769941329956055,
+      "learning_rate": 2e-05,
+      "loss": 0.02578051,
+      "step": 20753
+    },
+    {
+      "epoch": 41.508,
+      "grad_norm": 1.502710223197937,
+      "learning_rate": 2e-05,
+      "loss": 0.05851678,
+      "step": 20754
+    },
+    {
+      "epoch": 41.51,
+      "grad_norm": 1.0792176723480225,
+      "learning_rate": 2e-05,
+      "loss": 0.03291167,
+      "step": 20755
+    },
+    {
+      "epoch": 41.512,
+      "grad_norm": 1.7319297790527344,
+      "learning_rate": 2e-05,
+      "loss": 0.05475036,
+      "step": 20756
+    },
+    {
+      "epoch": 41.514,
+      "grad_norm": 1.918562889099121,
+      "learning_rate": 2e-05,
+      "loss": 0.04272918,
+      "step": 20757
+    },
+    {
+      "epoch": 41.516,
+      "grad_norm": 2.9401967525482178,
+      "learning_rate": 2e-05,
+      "loss": 0.05610197,
+      "step": 20758
+    },
+    {
+      "epoch": 41.518,
+      "grad_norm": 0.8515689373016357,
+      "learning_rate": 2e-05,
+      "loss": 0.02853437,
+      "step": 20759
+    },
+    {
+      "epoch": 41.52,
+      "grad_norm": 1.914589285850525,
+      "learning_rate": 2e-05,
+      "loss": 0.05829733,
+      "step": 20760
+    },
+    {
+      "epoch": 41.522,
+      "grad_norm": 1.0378468036651611,
+      "learning_rate": 2e-05,
+      "loss": 0.03754158,
+      "step": 20761
+    },
+    {
+      "epoch": 41.524,
+      "grad_norm": 1.3927786350250244,
+      "learning_rate": 2e-05,
+      "loss": 0.05459459,
+      "step": 20762
+    },
+    {
+      "epoch": 41.526,
+      "grad_norm": 4.084712028503418,
+      "learning_rate": 2e-05,
+      "loss": 0.0557055,
+      "step": 20763
+    },
+    {
+      "epoch": 41.528,
+      "grad_norm": 1.2816001176834106,
+      "learning_rate": 2e-05,
+      "loss": 0.05118892,
+      "step": 20764
+    },
+    {
+      "epoch": 41.53,
+      "grad_norm": 1.243981122970581,
+      "learning_rate": 2e-05,
+      "loss": 0.053567,
+      "step": 20765
+    },
+    {
+      "epoch": 41.532,
+      "grad_norm": 1.179433822631836,
+      "learning_rate": 2e-05,
+      "loss": 0.04334904,
+      "step": 20766
+    },
+    {
+      "epoch": 41.534,
+      "grad_norm": 1.8389700651168823,
+      "learning_rate": 2e-05,
+      "loss": 0.03809846,
+      "step": 20767
+    },
+    {
+      "epoch": 41.536,
+      "grad_norm": 1.2287765741348267,
+      "learning_rate": 2e-05,
+      "loss": 0.05584181,
+      "step": 20768
+    },
+    {
+      "epoch": 41.538,
+      "grad_norm": 1.2944599390029907,
+      "learning_rate": 2e-05,
+      "loss": 0.04906465,
+      "step": 20769
+    },
+    {
+      "epoch": 41.54,
+      "grad_norm": 1.9301244020462036,
+      "learning_rate": 2e-05,
+      "loss": 0.04297593,
+      "step": 20770
+    },
+    {
+      "epoch": 41.542,
+      "grad_norm": 1.5391250848770142,
+      "learning_rate": 2e-05,
+      "loss": 0.05037332,
+      "step": 20771
+    },
+    {
+      "epoch": 41.544,
+      "grad_norm": 1.3913311958312988,
+      "learning_rate": 2e-05,
+      "loss": 0.0544315,
+      "step": 20772
+    },
+    {
+      "epoch": 41.546,
+      "grad_norm": 0.9265551567077637,
+      "learning_rate": 2e-05,
+      "loss": 0.0323727,
+      "step": 20773
+    },
+    {
+      "epoch": 41.548,
+      "grad_norm": 1.0124329328536987,
+      "learning_rate": 2e-05,
+      "loss": 0.0399297,
+      "step": 20774
+    },
+    {
+      "epoch": 41.55,
+      "grad_norm": 1.2719837427139282,
+      "learning_rate": 2e-05,
+      "loss": 0.04783987,
+      "step": 20775
+    },
+    {
+      "epoch": 41.552,
+      "grad_norm": 1.9327387809753418,
+      "learning_rate": 2e-05,
+      "loss": 0.04304921,
+      "step": 20776
+    },
+    {
+      "epoch": 41.554,
+      "grad_norm": 1.8879715204238892,
+      "learning_rate": 2e-05,
+      "loss": 0.03215152,
+      "step": 20777
+    },
+    {
+      "epoch": 41.556,
+      "grad_norm": 1.0443545579910278,
+      "learning_rate": 2e-05,
+      "loss": 0.04257246,
+      "step": 20778
+    },
+    {
+      "epoch": 41.558,
+      "grad_norm": 1.7671880722045898,
+      "learning_rate": 2e-05,
+      "loss": 0.04795214,
+      "step": 20779
+    },
+    {
+      "epoch": 41.56,
+      "grad_norm": 1.3257173299789429,
+      "learning_rate": 2e-05,
+      "loss": 0.05144795,
+      "step": 20780
+    },
+    {
+      "epoch": 41.562,
+      "grad_norm": 1.5164159536361694,
+      "learning_rate": 2e-05,
+      "loss": 0.05290364,
+      "step": 20781
+    },
+    {
+      "epoch": 41.564,
+      "grad_norm": 1.2703793048858643,
+      "learning_rate": 2e-05,
+      "loss": 0.05436259,
+      "step": 20782
+    },
+    {
+      "epoch": 41.566,
+      "grad_norm": 1.13009774684906,
+      "learning_rate": 2e-05,
+      "loss": 0.03706975,
+      "step": 20783
+    },
+    {
+      "epoch": 41.568,
+      "grad_norm": 1.3934050798416138,
+      "learning_rate": 2e-05,
+      "loss": 0.04990171,
+      "step": 20784
+    },
+    {
+      "epoch": 41.57,
+      "grad_norm": 1.6412134170532227,
+      "learning_rate": 2e-05,
+      "loss": 0.04792435,
+      "step": 20785
+    },
+    {
+      "epoch": 41.572,
+      "grad_norm": 1.3018213510513306,
+      "learning_rate": 2e-05,
+      "loss": 0.06655528,
+      "step": 20786
+    },
+    {
+      "epoch": 41.574,
+      "grad_norm": 1.3327934741973877,
+      "learning_rate": 2e-05,
+      "loss": 0.04949384,
+      "step": 20787
+    },
+    {
+      "epoch": 41.576,
+      "grad_norm": 1.173862338066101,
+      "learning_rate": 2e-05,
+      "loss": 0.05083285,
+      "step": 20788
+    },
+    {
+      "epoch": 41.578,
+      "grad_norm": 1.3000322580337524,
+      "learning_rate": 2e-05,
+      "loss": 0.05392583,
+      "step": 20789
+    },
+    {
+      "epoch": 41.58,
+      "grad_norm": 1.50508451461792,
+      "learning_rate": 2e-05,
+      "loss": 0.03634304,
+      "step": 20790
+    },
+    {
+      "epoch": 41.582,
+      "grad_norm": 1.342342734336853,
+      "learning_rate": 2e-05,
+      "loss": 0.05079833,
+      "step": 20791
+    },
+    {
+      "epoch": 41.584,
+      "grad_norm": 1.1908931732177734,
+      "learning_rate": 2e-05,
+      "loss": 0.0528966,
+      "step": 20792
+    },
+    {
+      "epoch": 41.586,
+      "grad_norm": 1.2644261121749878,
+      "learning_rate": 2e-05,
+      "loss": 0.04991703,
+      "step": 20793
+    },
+    {
+      "epoch": 41.588,
+      "grad_norm": 1.19503915309906,
+      "learning_rate": 2e-05,
+      "loss": 0.04296641,
+      "step": 20794
+    },
+    {
+      "epoch": 41.59,
+      "grad_norm": 1.3065739870071411,
+      "learning_rate": 2e-05,
+      "loss": 0.04288668,
+      "step": 20795
+    },
+    {
+      "epoch": 41.592,
+      "grad_norm": 1.5874643325805664,
+      "learning_rate": 2e-05,
+      "loss": 0.06031433,
+      "step": 20796
+    },
+    {
+      "epoch": 41.594,
+      "grad_norm": 1.4615437984466553,
+      "learning_rate": 2e-05,
+      "loss": 0.06782658,
+      "step": 20797
+    },
+    {
+      "epoch": 41.596,
+      "grad_norm": 1.0287331342697144,
+      "learning_rate": 2e-05,
+      "loss": 0.03726696,
+      "step": 20798
+    },
+    {
+      "epoch": 41.598,
+      "grad_norm": 2.5669822692871094,
+      "learning_rate": 2e-05,
+      "loss": 0.05073772,
+      "step": 20799
+    },
+    {
+      "epoch": 41.6,
+      "grad_norm": 1.9865378141403198,
+      "learning_rate": 2e-05,
+      "loss": 0.06835793,
+      "step": 20800
+    },
+    {
+      "epoch": 41.602,
+      "grad_norm": 1.1996169090270996,
+      "learning_rate": 2e-05,
+      "loss": 0.0441338,
+      "step": 20801
+    },
+    {
+      "epoch": 41.604,
+      "grad_norm": 1.1404577493667603,
+      "learning_rate": 2e-05,
+      "loss": 0.05091837,
+      "step": 20802
+    },
+    {
+      "epoch": 41.606,
+      "grad_norm": 1.366138219833374,
+      "learning_rate": 2e-05,
+      "loss": 0.05536352,
+      "step": 20803
+    },
+    {
+      "epoch": 41.608,
+      "grad_norm": 1.2868354320526123,
+      "learning_rate": 2e-05,
+      "loss": 0.05645128,
+      "step": 20804
+    },
+    {
+      "epoch": 41.61,
+      "grad_norm": 1.2163975238800049,
+      "learning_rate": 2e-05,
+      "loss": 0.04356652,
+      "step": 20805
+    },
+    {
+      "epoch": 41.612,
+      "grad_norm": 1.5239417552947998,
+      "learning_rate": 2e-05,
+      "loss": 0.05934812,
+      "step": 20806
+    },
+    {
+      "epoch": 41.614,
+      "grad_norm": 1.641875982284546,
+      "learning_rate": 2e-05,
+      "loss": 0.0451197,
+      "step": 20807
+    },
+    {
+      "epoch": 41.616,
+      "grad_norm": 1.4720548391342163,
+      "learning_rate": 2e-05,
+      "loss": 0.05895507,
+      "step": 20808
+    },
+    {
+      "epoch": 41.618,
+      "grad_norm": 1.20379638671875,
+      "learning_rate": 2e-05,
+      "loss": 0.05722809,
+      "step": 20809
+    },
+    {
+      "epoch": 41.62,
+      "grad_norm": 1.206185221672058,
+      "learning_rate": 2e-05,
+      "loss": 0.04924655,
+      "step": 20810
+    },
+    {
+      "epoch": 41.622,
+      "grad_norm": 1.3422046899795532,
+      "learning_rate": 2e-05,
+      "loss": 0.04705754,
+      "step": 20811
+    },
+    {
+      "epoch": 41.624,
+      "grad_norm": 1.251327395439148,
+      "learning_rate": 2e-05,
+      "loss": 0.03359656,
+      "step": 20812
+    },
+    {
+      "epoch": 41.626,
+      "grad_norm": 1.0493559837341309,
+      "learning_rate": 2e-05,
+      "loss": 0.03650389,
+      "step": 20813
+    },
+    {
+      "epoch": 41.628,
+      "grad_norm": 1.1283539533615112,
+      "learning_rate": 2e-05,
+      "loss": 0.04552049,
+      "step": 20814
+    },
+    {
+      "epoch": 41.63,
+      "grad_norm": 1.2148489952087402,
+      "learning_rate": 2e-05,
+      "loss": 0.04333513,
+      "step": 20815
+    },
+    {
+      "epoch": 41.632,
+      "grad_norm": 1.1849173307418823,
+      "learning_rate": 2e-05,
+      "loss": 0.04451859,
+      "step": 20816
+    },
+    {
+      "epoch": 41.634,
+      "grad_norm": 1.035466194152832,
+      "learning_rate": 2e-05,
+      "loss": 0.03730474,
+      "step": 20817
+    },
+    {
+      "epoch": 41.636,
+      "grad_norm": 1.3401261568069458,
+      "learning_rate": 2e-05,
+      "loss": 0.03913242,
+      "step": 20818
+    },
+    {
+      "epoch": 41.638,
+      "grad_norm": 1.3163193464279175,
+      "learning_rate": 2e-05,
+      "loss": 0.04887679,
+      "step": 20819
+    },
+    {
+      "epoch": 41.64,
+      "grad_norm": 1.2960766553878784,
+      "learning_rate": 2e-05,
+      "loss": 0.05596934,
+      "step": 20820
+    },
+    {
+      "epoch": 41.642,
+      "grad_norm": 1.2313705682754517,
+      "learning_rate": 2e-05,
+      "loss": 0.05113539,
+      "step": 20821
+    },
+    {
+      "epoch": 41.644,
+      "grad_norm": 1.16353178024292,
+      "learning_rate": 2e-05,
+      "loss": 0.04695116,
+      "step": 20822
+    },
+    {
+      "epoch": 41.646,
+      "grad_norm": 1.308196783065796,
+      "learning_rate": 2e-05,
+      "loss": 0.05837334,
+      "step": 20823
+    },
+    {
+      "epoch": 41.648,
+      "grad_norm": 1.0429937839508057,
+      "learning_rate": 2e-05,
+      "loss": 0.0384528,
+      "step": 20824
+    },
+    {
+      "epoch": 41.65,
+      "grad_norm": 1.335108995437622,
+      "learning_rate": 2e-05,
+      "loss": 0.04670976,
+      "step": 20825
+    },
+    {
+      "epoch": 41.652,
+      "grad_norm": 1.6604396104812622,
+      "learning_rate": 2e-05,
+      "loss": 0.04694479,
+      "step": 20826
+    },
+    {
+      "epoch": 41.654,
+      "grad_norm": 1.2151720523834229,
+      "learning_rate": 2e-05,
+      "loss": 0.0505834,
+      "step": 20827
+    },
+    {
+      "epoch": 41.656,
+      "grad_norm": 1.6010440587997437,
+      "learning_rate": 2e-05,
+      "loss": 0.0715237,
+      "step": 20828
+    },
+    {
+      "epoch": 41.658,
+      "grad_norm": 1.2786024808883667,
+      "learning_rate": 2e-05,
+      "loss": 0.06591807,
+      "step": 20829
+    },
+    {
+      "epoch": 41.66,
+      "grad_norm": 1.1999880075454712,
+      "learning_rate": 2e-05,
+      "loss": 0.05343919,
+      "step": 20830
+    },
+    {
+      "epoch": 41.662,
+      "grad_norm": 1.5012918710708618,
+      "learning_rate": 2e-05,
+      "loss": 0.05417911,
+      "step": 20831
+    },
+    {
+      "epoch": 41.664,
+      "grad_norm": 1.03823983669281,
+      "learning_rate": 2e-05,
+      "loss": 0.03913938,
+      "step": 20832
+    },
+    {
+      "epoch": 41.666,
+      "grad_norm": 1.4988044500350952,
+      "learning_rate": 2e-05,
+      "loss": 0.05051984,
+      "step": 20833
+    },
+    {
+      "epoch": 41.668,
+      "grad_norm": 1.3459997177124023,
+      "learning_rate": 2e-05,
+      "loss": 0.05139535,
+      "step": 20834
+    },
+    {
+      "epoch": 41.67,
+      "grad_norm": 1.1768592596054077,
+      "learning_rate": 2e-05,
+      "loss": 0.05571685,
+      "step": 20835
+    },
+    {
+      "epoch": 41.672,
+      "grad_norm": 1.1096575260162354,
+      "learning_rate": 2e-05,
+      "loss": 0.04181407,
+      "step": 20836
+    },
+    {
+      "epoch": 41.674,
+      "grad_norm": 1.260703444480896,
+      "learning_rate": 2e-05,
+      "loss": 0.05937716,
+      "step": 20837
+    },
+    {
+      "epoch": 41.676,
+      "grad_norm": 1.1465245485305786,
+      "learning_rate": 2e-05,
+      "loss": 0.05365495,
+      "step": 20838
+    },
+    {
+      "epoch": 41.678,
+      "grad_norm": 1.7328318357467651,
+      "learning_rate": 2e-05,
+      "loss": 0.06675532,
+      "step": 20839
+    },
+    {
+      "epoch": 41.68,
+      "grad_norm": 1.3580307960510254,
+      "learning_rate": 2e-05,
+      "loss": 0.06565544,
+      "step": 20840
+    },
+    {
+      "epoch": 41.682,
+      "grad_norm": 1.1711437702178955,
+      "learning_rate": 2e-05,
+      "loss": 0.04366954,
+      "step": 20841
+    },
+    {
+      "epoch": 41.684,
+      "grad_norm": 1.7624998092651367,
+      "learning_rate": 2e-05,
+      "loss": 0.05125007,
+      "step": 20842
+    },
+    {
+      "epoch": 41.686,
+      "grad_norm": 1.7056723833084106,
+      "learning_rate": 2e-05,
+      "loss": 0.06327073,
+      "step": 20843
+    },
+    {
+      "epoch": 41.688,
+      "grad_norm": 1.4002678394317627,
+      "learning_rate": 2e-05,
+      "loss": 0.05082846,
+      "step": 20844
+    },
+    {
+      "epoch": 41.69,
+      "grad_norm": 1.8663007020950317,
+      "learning_rate": 2e-05,
+      "loss": 0.06116794,
+      "step": 20845
+    },
+    {
+      "epoch": 41.692,
+      "grad_norm": 1.311848759651184,
+      "learning_rate": 2e-05,
+      "loss": 0.05300967,
+      "step": 20846
+    },
+    {
+      "epoch": 41.694,
+      "grad_norm": 1.2102714776992798,
+      "learning_rate": 2e-05,
+      "loss": 0.05376907,
+      "step": 20847
+    },
+    {
+      "epoch": 41.696,
+      "grad_norm": 1.0833280086517334,
+      "learning_rate": 2e-05,
+      "loss": 0.04561132,
+      "step": 20848
+    },
+    {
+      "epoch": 41.698,
+      "grad_norm": 1.1252832412719727,
+      "learning_rate": 2e-05,
+      "loss": 0.04921568,
+      "step": 20849
+    },
+    {
+      "epoch": 41.7,
+      "grad_norm": 1.2025973796844482,
+      "learning_rate": 2e-05,
+      "loss": 0.04832213,
+      "step": 20850
+    },
+    {
+      "epoch": 41.702,
+      "grad_norm": 1.0918816328048706,
+      "learning_rate": 2e-05,
+      "loss": 0.04648707,
+      "step": 20851
+    },
+    {
+      "epoch": 41.704,
+      "grad_norm": 1.1413060426712036,
+      "learning_rate": 2e-05,
+      "loss": 0.05401688,
+      "step": 20852
+    },
+    {
+      "epoch": 41.706,
+      "grad_norm": 1.722861886024475,
+      "learning_rate": 2e-05,
+      "loss": 0.04710011,
+      "step": 20853
+    },
+    {
+      "epoch": 41.708,
+      "grad_norm": 1.183224081993103,
+      "learning_rate": 2e-05,
+      "loss": 0.04919672,
+      "step": 20854
+    },
+    {
+      "epoch": 41.71,
+      "grad_norm": 1.0682933330535889,
+      "learning_rate": 2e-05,
+      "loss": 0.04360975,
+      "step": 20855
+    },
+    {
+      "epoch": 41.712,
+      "grad_norm": 1.3909543752670288,
+      "learning_rate": 2e-05,
+      "loss": 0.03963952,
+      "step": 20856
+    },
+    {
+      "epoch": 41.714,
+      "grad_norm": 1.2624576091766357,
+      "learning_rate": 2e-05,
+      "loss": 0.04159348,
+      "step": 20857
+    },
+    {
+      "epoch": 41.716,
+      "grad_norm": 1.2852903604507446,
+      "learning_rate": 2e-05,
+      "loss": 0.06388488,
+      "step": 20858
+    },
+    {
+      "epoch": 41.718,
+      "grad_norm": 1.2193105220794678,
+      "learning_rate": 2e-05,
+      "loss": 0.04823635,
+      "step": 20859
+    },
+    {
+      "epoch": 41.72,
+      "grad_norm": 1.3638604879379272,
+      "learning_rate": 2e-05,
+      "loss": 0.0435699,
+      "step": 20860
+    },
+    {
+      "epoch": 41.722,
+      "grad_norm": 1.2816431522369385,
+      "learning_rate": 2e-05,
+      "loss": 0.05209901,
+      "step": 20861
+    },
+    {
+      "epoch": 41.724,
+      "grad_norm": 1.994900107383728,
+      "learning_rate": 2e-05,
+      "loss": 0.0689984,
+      "step": 20862
+    },
+    {
+      "epoch": 41.726,
+      "grad_norm": 1.2706905603408813,
+      "learning_rate": 2e-05,
+      "loss": 0.04907268,
+      "step": 20863
+    },
+    {
+      "epoch": 41.728,
+      "grad_norm": 1.134853482246399,
+      "learning_rate": 2e-05,
+      "loss": 0.03874713,
+      "step": 20864
+    },
+    {
+      "epoch": 41.73,
+      "grad_norm": 1.305557370185852,
+      "learning_rate": 2e-05,
+      "loss": 0.0443382,
+      "step": 20865
+    },
+    {
+      "epoch": 41.732,
+      "grad_norm": 1.287753701210022,
+      "learning_rate": 2e-05,
+      "loss": 0.05005742,
+      "step": 20866
+    },
+    {
+      "epoch": 41.734,
+      "grad_norm": 1.2980296611785889,
+      "learning_rate": 2e-05,
+      "loss": 0.04606131,
+      "step": 20867
+    },
+    {
+      "epoch": 41.736,
+      "grad_norm": 1.003082036972046,
+      "learning_rate": 2e-05,
+      "loss": 0.03791095,
+      "step": 20868
+    },
+    {
+      "epoch": 41.738,
+      "grad_norm": 1.2747972011566162,
+      "learning_rate": 2e-05,
+      "loss": 0.04731661,
+      "step": 20869
+    },
+    {
+      "epoch": 41.74,
+      "grad_norm": 1.4059251546859741,
+      "learning_rate": 2e-05,
+      "loss": 0.03448366,
+      "step": 20870
+    },
+    {
+      "epoch": 41.742,
+      "grad_norm": 2.1521401405334473,
+      "learning_rate": 2e-05,
+      "loss": 0.06867001,
+      "step": 20871
+    },
+    {
+      "epoch": 41.744,
+      "grad_norm": 1.4154833555221558,
+      "learning_rate": 2e-05,
+      "loss": 0.03487004,
+      "step": 20872
+    },
+    {
+      "epoch": 41.746,
+      "grad_norm": 1.2275240421295166,
+      "learning_rate": 2e-05,
+      "loss": 0.05700513,
+      "step": 20873
+    },
+    {
+      "epoch": 41.748,
+      "grad_norm": 2.4994146823883057,
+      "learning_rate": 2e-05,
+      "loss": 0.05144731,
+      "step": 20874
+    },
+    {
+      "epoch": 41.75,
+      "grad_norm": 1.8417681455612183,
+      "learning_rate": 2e-05,
+      "loss": 0.05225058,
+      "step": 20875
+    },
+    {
+      "epoch": 41.752,
+      "grad_norm": 1.2966296672821045,
+      "learning_rate": 2e-05,
+      "loss": 0.04987405,
+      "step": 20876
+    },
+    {
+      "epoch": 41.754,
+      "grad_norm": 1.2915031909942627,
+      "learning_rate": 2e-05,
+      "loss": 0.05634173,
+      "step": 20877
+    },
+    {
+      "epoch": 41.756,
+      "grad_norm": 1.4315176010131836,
+      "learning_rate": 2e-05,
+      "loss": 0.05782579,
+      "step": 20878
+    },
+    {
+      "epoch": 41.758,
+      "grad_norm": 1.8965885639190674,
+      "learning_rate": 2e-05,
+      "loss": 0.05472075,
+      "step": 20879
+    },
+    {
+      "epoch": 41.76,
+      "grad_norm": 1.0366774797439575,
+      "learning_rate": 2e-05,
+      "loss": 0.03605853,
+      "step": 20880
+    },
+    {
+      "epoch": 41.762,
+      "grad_norm": 1.6545394659042358,
+      "learning_rate": 2e-05,
+      "loss": 0.05220097,
+      "step": 20881
+    },
+    {
+      "epoch": 41.764,
+      "grad_norm": 1.9253038167953491,
+      "learning_rate": 2e-05,
+      "loss": 0.04118697,
+      "step": 20882
+    },
+    {
+      "epoch": 41.766,
+      "grad_norm": 1.229835033416748,
+      "learning_rate": 2e-05,
+      "loss": 0.05954335,
+      "step": 20883
+    },
+    {
+      "epoch": 41.768,
+      "grad_norm": 2.4672608375549316,
+      "learning_rate": 2e-05,
+      "loss": 0.03520424,
+      "step": 20884
+    },
+    {
+      "epoch": 41.77,
+      "grad_norm": 1.317997694015503,
+      "learning_rate": 2e-05,
+      "loss": 0.05394396,
+      "step": 20885
+    },
+    {
+      "epoch": 41.772,
+      "grad_norm": 1.1028608083724976,
+      "learning_rate": 2e-05,
+      "loss": 0.05019546,
+      "step": 20886
+    },
+    {
+      "epoch": 41.774,
+      "grad_norm": 1.4625998735427856,
+      "learning_rate": 2e-05,
+      "loss": 0.05548112,
+      "step": 20887
+    },
+    {
+      "epoch": 41.776,
+      "grad_norm": 1.5702537298202515,
+      "learning_rate": 2e-05,
+      "loss": 0.0497899,
+      "step": 20888
+    },
+    {
+      "epoch": 41.778,
+      "grad_norm": 1.258571743965149,
+      "learning_rate": 2e-05,
+      "loss": 0.04912242,
+      "step": 20889
+    },
+    {
+      "epoch": 41.78,
+      "grad_norm": 1.0844334363937378,
+      "learning_rate": 2e-05,
+      "loss": 0.04057037,
+      "step": 20890
+    },
+    {
+      "epoch": 41.782,
+      "grad_norm": 1.1522717475891113,
+      "learning_rate": 2e-05,
+      "loss": 0.04750651,
+      "step": 20891
+    },
+    {
+      "epoch": 41.784,
+      "grad_norm": 1.290791630744934,
+      "learning_rate": 2e-05,
+      "loss": 0.06190044,
+      "step": 20892
+    },
+    {
+      "epoch": 41.786,
+      "grad_norm": 1.1737576723098755,
+      "learning_rate": 2e-05,
+      "loss": 0.03324655,
+      "step": 20893
+    },
+    {
+      "epoch": 41.788,
+      "grad_norm": 1.2524381875991821,
+      "learning_rate": 2e-05,
+      "loss": 0.03917806,
+      "step": 20894
+    },
+    {
+      "epoch": 41.79,
+      "grad_norm": 1.3281619548797607,
+      "learning_rate": 2e-05,
+      "loss": 0.04660825,
+      "step": 20895
+    },
+    {
+      "epoch": 41.792,
+      "grad_norm": 1.430392861366272,
+      "learning_rate": 2e-05,
+      "loss": 0.07056463,
+      "step": 20896
+    },
+    {
+      "epoch": 41.794,
+      "grad_norm": 1.4818058013916016,
+      "learning_rate": 2e-05,
+      "loss": 0.05624013,
+      "step": 20897
+    },
+    {
+      "epoch": 41.796,
+      "grad_norm": 1.2507944107055664,
+      "learning_rate": 2e-05,
+      "loss": 0.05323803,
+      "step": 20898
+    },
+    {
+      "epoch": 41.798,
+      "grad_norm": 1.244249701499939,
+      "learning_rate": 2e-05,
+      "loss": 0.04751854,
+      "step": 20899
+    },
+    {
+      "epoch": 41.8,
+      "grad_norm": 2.58414363861084,
+      "learning_rate": 2e-05,
+      "loss": 0.05688569,
+      "step": 20900
+    },
+    {
+      "epoch": 41.802,
+      "grad_norm": 1.0846785306930542,
+      "learning_rate": 2e-05,
+      "loss": 0.03785557,
+      "step": 20901
+    },
+    {
+      "epoch": 41.804,
+      "grad_norm": 1.3131710290908813,
+      "learning_rate": 2e-05,
+      "loss": 0.05168329,
+      "step": 20902
+    },
+    {
+      "epoch": 41.806,
+      "grad_norm": 1.262276530265808,
+      "learning_rate": 2e-05,
+      "loss": 0.04058243,
+      "step": 20903
+    },
+    {
+      "epoch": 41.808,
+      "grad_norm": 1.134749412536621,
+      "learning_rate": 2e-05,
+      "loss": 0.033862,
+      "step": 20904
+    },
+    {
+      "epoch": 41.81,
+      "grad_norm": 1.2034157514572144,
+      "learning_rate": 2e-05,
+      "loss": 0.0472676,
+      "step": 20905
+    },
+    {
+      "epoch": 41.812,
+      "grad_norm": 2.1183056831359863,
+      "learning_rate": 2e-05,
+      "loss": 0.04632457,
+      "step": 20906
+    },
+    {
+      "epoch": 41.814,
+      "grad_norm": 1.1431421041488647,
+      "learning_rate": 2e-05,
+      "loss": 0.04334393,
+      "step": 20907
+    },
+    {
+      "epoch": 41.816,
+      "grad_norm": 1.328840732574463,
+      "learning_rate": 2e-05,
+      "loss": 0.04690224,
+      "step": 20908
+    },
+    {
+      "epoch": 41.818,
+      "grad_norm": 0.96977299451828,
+      "learning_rate": 2e-05,
+      "loss": 0.03683947,
+      "step": 20909
+    },
+    {
+      "epoch": 41.82,
+      "grad_norm": 1.1125136613845825,
+      "learning_rate": 2e-05,
+      "loss": 0.04391717,
+      "step": 20910
+    },
+    {
+      "epoch": 41.822,
+      "grad_norm": 1.0546776056289673,
+      "learning_rate": 2e-05,
+      "loss": 0.0294345,
+      "step": 20911
+    },
+    {
+      "epoch": 41.824,
+      "grad_norm": 1.0246164798736572,
+      "learning_rate": 2e-05,
+      "loss": 0.04162443,
+      "step": 20912
+    },
+    {
+      "epoch": 41.826,
+      "grad_norm": 1.2705532312393188,
+      "learning_rate": 2e-05,
+      "loss": 0.06132779,
+      "step": 20913
+    },
+    {
+      "epoch": 41.828,
+      "grad_norm": 1.0475337505340576,
+      "learning_rate": 2e-05,
+      "loss": 0.03245585,
+      "step": 20914
+    },
+    {
+      "epoch": 41.83,
+      "grad_norm": 1.4540488719940186,
+      "learning_rate": 2e-05,
+      "loss": 0.04266465,
+      "step": 20915
+    },
+    {
+      "epoch": 41.832,
+      "grad_norm": 1.089884638786316,
+      "learning_rate": 2e-05,
+      "loss": 0.03583428,
+      "step": 20916
+    },
+    {
+      "epoch": 41.834,
+      "grad_norm": 1.248226284980774,
+      "learning_rate": 2e-05,
+      "loss": 0.04630677,
+      "step": 20917
+    },
+    {
+      "epoch": 41.836,
+      "grad_norm": 1.2985204458236694,
+      "learning_rate": 2e-05,
+      "loss": 0.05467469,
+      "step": 20918
+    },
+    {
+      "epoch": 41.838,
+      "grad_norm": 1.6257275342941284,
+      "learning_rate": 2e-05,
+      "loss": 0.05300467,
+      "step": 20919
+    },
+    {
+      "epoch": 41.84,
+      "grad_norm": 1.0730141401290894,
+      "learning_rate": 2e-05,
+      "loss": 0.04086423,
+      "step": 20920
+    },
+    {
+      "epoch": 41.842,
+      "grad_norm": 1.7589372396469116,
+      "learning_rate": 2e-05,
+      "loss": 0.07659753,
+      "step": 20921
+    },
+    {
+      "epoch": 41.844,
+      "grad_norm": 1.650455355644226,
+      "learning_rate": 2e-05,
+      "loss": 0.04704353,
+      "step": 20922
+    },
+    {
+      "epoch": 41.846,
+      "grad_norm": 0.9814362525939941,
+      "learning_rate": 2e-05,
+      "loss": 0.04067108,
+      "step": 20923
+    },
+    {
+      "epoch": 41.848,
+      "grad_norm": 2.033730983734131,
+      "learning_rate": 2e-05,
+      "loss": 0.03695693,
+      "step": 20924
+    },
+    {
+      "epoch": 41.85,
+      "grad_norm": 1.1429461240768433,
+      "learning_rate": 2e-05,
+      "loss": 0.04535512,
+      "step": 20925
+    },
+    {
+      "epoch": 41.852,
+      "grad_norm": 1.354186773300171,
+      "learning_rate": 2e-05,
+      "loss": 0.05408484,
+      "step": 20926
+    },
+    {
+      "epoch": 41.854,
+      "grad_norm": 2.6549506187438965,
+      "learning_rate": 2e-05,
+      "loss": 0.05873822,
+      "step": 20927
+    },
+    {
+      "epoch": 41.856,
+      "grad_norm": 1.1592339277267456,
+      "learning_rate": 2e-05,
+      "loss": 0.04995368,
+      "step": 20928
+    },
+    {
+      "epoch": 41.858,
+      "grad_norm": 1.7437788248062134,
+      "learning_rate": 2e-05,
+      "loss": 0.07510689,
+      "step": 20929
+    },
+    {
+      "epoch": 41.86,
+      "grad_norm": 1.1084489822387695,
+      "learning_rate": 2e-05,
+      "loss": 0.04345679,
+      "step": 20930
+    },
+    {
+      "epoch": 41.862,
+      "grad_norm": 1.3295727968215942,
+      "learning_rate": 2e-05,
+      "loss": 0.04446131,
+      "step": 20931
+    },
+    {
+      "epoch": 41.864,
+      "grad_norm": 1.3433420658111572,
+      "learning_rate": 2e-05,
+      "loss": 0.05595826,
+      "step": 20932
+    },
+    {
+      "epoch": 41.866,
+      "grad_norm": 1.3462584018707275,
+      "learning_rate": 2e-05,
+      "loss": 0.05649468,
+      "step": 20933
+    },
+    {
+      "epoch": 41.868,
+      "grad_norm": 1.4262540340423584,
+      "learning_rate": 2e-05,
+      "loss": 0.04885809,
+      "step": 20934
+    },
+    {
+      "epoch": 41.87,
+      "grad_norm": 2.7343051433563232,
+      "learning_rate": 2e-05,
+      "loss": 0.06057374,
+      "step": 20935
+    },
+    {
+      "epoch": 41.872,
+      "grad_norm": 1.2475897073745728,
+      "learning_rate": 2e-05,
+      "loss": 0.05510751,
+      "step": 20936
+    },
+    {
+      "epoch": 41.874,
+      "grad_norm": 1.5104964971542358,
+      "learning_rate": 2e-05,
+      "loss": 0.05848288,
+      "step": 20937
+    },
+    {
+      "epoch": 41.876,
+      "grad_norm": 1.10996675491333,
+      "learning_rate": 2e-05,
+      "loss": 0.04091154,
+      "step": 20938
+    },
+    {
+      "epoch": 41.878,
+      "grad_norm": 1.2295801639556885,
+      "learning_rate": 2e-05,
+      "loss": 0.05182572,
+      "step": 20939
+    },
+    {
+      "epoch": 41.88,
+      "grad_norm": 1.200059413909912,
+      "learning_rate": 2e-05,
+      "loss": 0.04649248,
+      "step": 20940
+    },
+    {
+      "epoch": 41.882,
+      "grad_norm": 1.1000398397445679,
+      "learning_rate": 2e-05,
+      "loss": 0.03840358,
+      "step": 20941
+    },
+    {
+      "epoch": 41.884,
+      "grad_norm": 1.3613183498382568,
+      "learning_rate": 2e-05,
+      "loss": 0.05442974,
+      "step": 20942
+    },
+    {
+      "epoch": 41.886,
+      "grad_norm": 1.567658543586731,
+      "learning_rate": 2e-05,
+      "loss": 0.05978185,
+      "step": 20943
+    },
+    {
+      "epoch": 41.888,
+      "grad_norm": 1.0944557189941406,
+      "learning_rate": 2e-05,
+      "loss": 0.03656168,
+      "step": 20944
+    },
+    {
+      "epoch": 41.89,
+      "grad_norm": 1.177577018737793,
+      "learning_rate": 2e-05,
+      "loss": 0.04644332,
+      "step": 20945
+    },
+    {
+      "epoch": 41.892,
+      "grad_norm": 1.2059484720230103,
+      "learning_rate": 2e-05,
+      "loss": 0.03849885,
+      "step": 20946
+    },
+    {
+      "epoch": 41.894,
+      "grad_norm": 4.2382283210754395,
+      "learning_rate": 2e-05,
+      "loss": 0.05498065,
+      "step": 20947
+    },
+    {
+      "epoch": 41.896,
+      "grad_norm": 1.1419352293014526,
+      "learning_rate": 2e-05,
+      "loss": 0.04763758,
+      "step": 20948
+    },
+    {
+      "epoch": 41.898,
+      "grad_norm": 2.648472785949707,
+      "learning_rate": 2e-05,
+      "loss": 0.06559248,
+      "step": 20949
+    },
+    {
+      "epoch": 41.9,
+      "grad_norm": 1.1834135055541992,
+      "learning_rate": 2e-05,
+      "loss": 0.06024366,
+      "step": 20950
+    },
+    {
+      "epoch": 41.902,
+      "grad_norm": 1.1037529706954956,
+      "learning_rate": 2e-05,
+      "loss": 0.05040961,
+      "step": 20951
+    },
+    {
+      "epoch": 41.904,
+      "grad_norm": 1.312083125114441,
+      "learning_rate": 2e-05,
+      "loss": 0.05827889,
+      "step": 20952
+    },
+    {
+      "epoch": 41.906,
+      "grad_norm": 1.707861065864563,
+      "learning_rate": 2e-05,
+      "loss": 0.05063179,
+      "step": 20953
+    },
+    {
+      "epoch": 41.908,
+      "grad_norm": 1.2550243139266968,
+      "learning_rate": 2e-05,
+      "loss": 0.04975577,
+      "step": 20954
+    },
+    {
+      "epoch": 41.91,
+      "grad_norm": 1.1055585145950317,
+      "learning_rate": 2e-05,
+      "loss": 0.04447905,
+      "step": 20955
+    },
+    {
+      "epoch": 41.912,
+      "grad_norm": 1.2128030061721802,
+      "learning_rate": 2e-05,
+      "loss": 0.04356065,
+      "step": 20956
+    },
+    {
+      "epoch": 41.914,
+      "grad_norm": 2.053502321243286,
+      "learning_rate": 2e-05,
+      "loss": 0.04877847,
+      "step": 20957
+    },
+    {
+      "epoch": 41.916,
+      "grad_norm": 1.2415821552276611,
+      "learning_rate": 2e-05,
+      "loss": 0.05251724,
+      "step": 20958
+    },
+    {
+      "epoch": 41.918,
+      "grad_norm": 1.246356725692749,
+      "learning_rate": 2e-05,
+      "loss": 0.03539124,
+      "step": 20959
+    },
+    {
+      "epoch": 41.92,
+      "grad_norm": 3.0229344367980957,
+      "learning_rate": 2e-05,
+      "loss": 0.04473191,
+      "step": 20960
+    },
+    {
+      "epoch": 41.922,
+      "grad_norm": 1.5012351274490356,
+      "learning_rate": 2e-05,
+      "loss": 0.04717286,
+      "step": 20961
+    },
+    {
+      "epoch": 41.924,
+      "grad_norm": 1.6548337936401367,
+      "learning_rate": 2e-05,
+      "loss": 0.05977109,
+      "step": 20962
+    },
+    {
+      "epoch": 41.926,
+      "grad_norm": 1.0772522687911987,
+      "learning_rate": 2e-05,
+      "loss": 0.04247387,
+      "step": 20963
+    },
+    {
+      "epoch": 41.928,
+      "grad_norm": 1.259922981262207,
+      "learning_rate": 2e-05,
+      "loss": 0.05617234,
+      "step": 20964
+    },
+    {
+      "epoch": 41.93,
+      "grad_norm": 1.1000100374221802,
+      "learning_rate": 2e-05,
+      "loss": 0.04250489,
+      "step": 20965
+    },
+    {
+      "epoch": 41.932,
+      "grad_norm": 0.9738196730613708,
+      "learning_rate": 2e-05,
+      "loss": 0.02657899,
+      "step": 20966
+    },
+    {
+      "epoch": 41.934,
+      "grad_norm": 1.1028367280960083,
+      "learning_rate": 2e-05,
+      "loss": 0.03186112,
+      "step": 20967
+    },
+    {
+      "epoch": 41.936,
+      "grad_norm": 1.2720319032669067,
+      "learning_rate": 2e-05,
+      "loss": 0.04869255,
+      "step": 20968
+    },
+    {
+      "epoch": 41.938,
+      "grad_norm": 1.1687405109405518,
+      "learning_rate": 2e-05,
+      "loss": 0.03809325,
+      "step": 20969
+    },
+    {
+      "epoch": 41.94,
+      "grad_norm": 0.9612535238265991,
+      "learning_rate": 2e-05,
+      "loss": 0.02938464,
+      "step": 20970
+    },
+    {
+      "epoch": 41.942,
+      "grad_norm": 1.5162972211837769,
+      "learning_rate": 2e-05,
+      "loss": 0.06188224,
+      "step": 20971
+    },
+    {
+      "epoch": 41.944,
+      "grad_norm": 2.4762864112854004,
+      "learning_rate": 2e-05,
+      "loss": 0.05203952,
+      "step": 20972
+    },
+    {
+      "epoch": 41.946,
+      "grad_norm": 1.2529581785202026,
+      "learning_rate": 2e-05,
+      "loss": 0.04696166,
+      "step": 20973
+    },
+    {
+      "epoch": 41.948,
+      "grad_norm": 1.1282131671905518,
+      "learning_rate": 2e-05,
+      "loss": 0.03645737,
+      "step": 20974
+    },
+    {
+      "epoch": 41.95,
+      "grad_norm": 1.2952510118484497,
+      "learning_rate": 2e-05,
+      "loss": 0.04191312,
+      "step": 20975
+    },
+    {
+      "epoch": 41.952,
+      "grad_norm": 1.780375599861145,
+      "learning_rate": 2e-05,
+      "loss": 0.05335905,
+      "step": 20976
+    },
+    {
+      "epoch": 41.954,
+      "grad_norm": 1.3107290267944336,
+      "learning_rate": 2e-05,
+      "loss": 0.04631649,
+      "step": 20977
+    },
+    {
+      "epoch": 41.956,
+      "grad_norm": 1.3172683715820312,
+      "learning_rate": 2e-05,
+      "loss": 0.05279814,
+      "step": 20978
+    },
+    {
+      "epoch": 41.958,
+      "grad_norm": 1.3357253074645996,
+      "learning_rate": 2e-05,
+      "loss": 0.05109545,
+      "step": 20979
+    },
+    {
+      "epoch": 41.96,
+      "grad_norm": 1.298035740852356,
+      "learning_rate": 2e-05,
+      "loss": 0.05070015,
+      "step": 20980
+    },
+    {
+      "epoch": 41.962,
+      "grad_norm": 1.3245327472686768,
+      "learning_rate": 2e-05,
+      "loss": 0.05043676,
+      "step": 20981
+    },
+    {
+      "epoch": 41.964,
+      "grad_norm": 1.8343544006347656,
+      "learning_rate": 2e-05,
+      "loss": 0.05134499,
+      "step": 20982
+    },
+    {
+      "epoch": 41.966,
+      "grad_norm": 1.2885112762451172,
+      "learning_rate": 2e-05,
+      "loss": 0.04808827,
+      "step": 20983
+    },
+    {
+      "epoch": 41.968,
+      "grad_norm": 1.0060571432113647,
+      "learning_rate": 2e-05,
+      "loss": 0.04182211,
+      "step": 20984
+    },
+    {
+      "epoch": 41.97,
+      "grad_norm": 1.8609304428100586,
+      "learning_rate": 2e-05,
+      "loss": 0.05335212,
+      "step": 20985
+    },
+    {
+      "epoch": 41.972,
+      "grad_norm": 1.4980871677398682,
+      "learning_rate": 2e-05,
+      "loss": 0.04966426,
+      "step": 20986
+    },
+    {
+      "epoch": 41.974,
+      "grad_norm": 1.4951345920562744,
+      "learning_rate": 2e-05,
+      "loss": 0.04603991,
+      "step": 20987
+    },
+    {
+      "epoch": 41.976,
+      "grad_norm": 0.9615522623062134,
+      "learning_rate": 2e-05,
+      "loss": 0.02445277,
+      "step": 20988
+    },
+    {
+      "epoch": 41.978,
+      "grad_norm": 1.2384010553359985,
+      "learning_rate": 2e-05,
+      "loss": 0.04333053,
+      "step": 20989
+    },
+    {
+      "epoch": 41.98,
+      "grad_norm": 1.366593837738037,
+      "learning_rate": 2e-05,
+      "loss": 0.06016916,
+      "step": 20990
+    },
+    {
+      "epoch": 41.982,
+      "grad_norm": 1.1202651262283325,
+      "learning_rate": 2e-05,
+      "loss": 0.0522389,
+      "step": 20991
+    },
+    {
+      "epoch": 41.984,
+      "grad_norm": 1.1914691925048828,
+      "learning_rate": 2e-05,
+      "loss": 0.04286882,
+      "step": 20992
+    },
+    {
+      "epoch": 41.986,
+      "grad_norm": 1.3832005262374878,
+      "learning_rate": 2e-05,
+      "loss": 0.05637228,
+      "step": 20993
+    },
+    {
+      "epoch": 41.988,
+      "grad_norm": 1.253294587135315,
+      "learning_rate": 2e-05,
+      "loss": 0.04640395,
+      "step": 20994
+    },
+    {
+      "epoch": 41.99,
+      "grad_norm": 2.2488229274749756,
+      "learning_rate": 2e-05,
+      "loss": 0.04534402,
+      "step": 20995
+    },
+    {
+      "epoch": 41.992,
+      "grad_norm": 1.515738606452942,
+      "learning_rate": 2e-05,
+      "loss": 0.0565161,
+      "step": 20996
+    },
+    {
+      "epoch": 41.994,
+      "grad_norm": 2.047625780105591,
+      "learning_rate": 2e-05,
+      "loss": 0.04121755,
+      "step": 20997
+    },
+    {
+      "epoch": 41.996,
+      "grad_norm": 2.0094411373138428,
+      "learning_rate": 2e-05,
+      "loss": 0.06695313,
+      "step": 20998
+    },
+    {
+      "epoch": 41.998,
+      "grad_norm": 1.419292688369751,
+      "learning_rate": 2e-05,
+      "loss": 0.04039873,
+      "step": 20999
+    },
+    {
+      "epoch": 42.0,
+      "grad_norm": 2.344158887863159,
+      "learning_rate": 2e-05,
+      "loss": 0.047058,
+      "step": 21000
+    },
+    {
+      "epoch": 42.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.994,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9820359281437125,
+        "Equal_1": 0.996,
+        "Equal_2": 0.9820359281437125,
+        "Equal_3": 0.9880239520958084,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.9940119760479041,
+        "Parallel_1": 0.9959919839679359,
+        "Parallel_2": 0.9979959919839679,
+        "Parallel_3": 0.996,
+        "Perpendicular_1": 1.0,
+        "Perpendicular_2": 0.998,
+        "Perpendicular_3": 0.9038076152304609,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 1.0,
+        "PointLiesOnCircle_3": 0.9956,
+        "PointLiesOnLine_1": 0.9939879759519038,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9860279441117764
+      },
+      "eval_runtime": 320.4536,
+      "eval_samples_per_second": 32.766,
+      "eval_steps_per_second": 0.655,
+      "step": 21000
+    },
+    {
+      "epoch": 42.002,
+      "grad_norm": 2.0079424381256104,
+      "learning_rate": 2e-05,
+      "loss": 0.03449755,
+      "step": 21001
+    },
+    {
+      "epoch": 42.004,
+      "grad_norm": 1.365389108657837,
+      "learning_rate": 2e-05,
+      "loss": 0.0548946,
+      "step": 21002
+    },
+    {
+      "epoch": 42.006,
+      "grad_norm": 2.7778944969177246,
+      "learning_rate": 2e-05,
+      "loss": 0.04609371,
+      "step": 21003
+    },
+    {
+      "epoch": 42.008,
+      "grad_norm": 1.0593626499176025,
+      "learning_rate": 2e-05,
+      "loss": 0.0348668,
+      "step": 21004
+    },
+    {
+      "epoch": 42.01,
+      "grad_norm": 1.3281819820404053,
+      "learning_rate": 2e-05,
+      "loss": 0.05575937,
+      "step": 21005
+    },
+    {
+      "epoch": 42.012,
+      "grad_norm": 1.5199637413024902,
+      "learning_rate": 2e-05,
+      "loss": 0.04298349,
+      "step": 21006
+    },
+    {
+      "epoch": 42.014,
+      "grad_norm": 1.4977668523788452,
+      "learning_rate": 2e-05,
+      "loss": 0.03856365,
+      "step": 21007
+    },
+    {
+      "epoch": 42.016,
+      "grad_norm": 1.0063074827194214,
+      "learning_rate": 2e-05,
+      "loss": 0.02791759,
+      "step": 21008
+    },
+    {
+      "epoch": 42.018,
+      "grad_norm": 1.3319441080093384,
+      "learning_rate": 2e-05,
+      "loss": 0.05699763,
+      "step": 21009
+    },
+    {
+      "epoch": 42.02,
+      "grad_norm": 1.0372650623321533,
+      "learning_rate": 2e-05,
+      "loss": 0.04131971,
+      "step": 21010
+    },
+    {
+      "epoch": 42.022,
+      "grad_norm": 1.1082048416137695,
+      "learning_rate": 2e-05,
+      "loss": 0.04324374,
+      "step": 21011
+    },
+    {
+      "epoch": 42.024,
+      "grad_norm": 1.2265472412109375,
+      "learning_rate": 2e-05,
+      "loss": 0.04790197,
+      "step": 21012
+    },
+    {
+      "epoch": 42.026,
+      "grad_norm": 1.5326513051986694,
+      "learning_rate": 2e-05,
+      "loss": 0.05717977,
+      "step": 21013
+    },
+    {
+      "epoch": 42.028,
+      "grad_norm": 1.2501510381698608,
+      "learning_rate": 2e-05,
+      "loss": 0.04369808,
+      "step": 21014
+    },
+    {
+      "epoch": 42.03,
+      "grad_norm": 0.9996318817138672,
+      "learning_rate": 2e-05,
+      "loss": 0.03566208,
+      "step": 21015
+    },
+    {
+      "epoch": 42.032,
+      "grad_norm": 1.2359881401062012,
+      "learning_rate": 2e-05,
+      "loss": 0.05509493,
+      "step": 21016
+    },
+    {
+      "epoch": 42.034,
+      "grad_norm": 1.4756462574005127,
+      "learning_rate": 2e-05,
+      "loss": 0.06215252,
+      "step": 21017
+    },
+    {
+      "epoch": 42.036,
+      "grad_norm": 1.2540290355682373,
+      "learning_rate": 2e-05,
+      "loss": 0.03848273,
+      "step": 21018
+    },
+    {
+      "epoch": 42.038,
+      "grad_norm": 1.3934910297393799,
+      "learning_rate": 2e-05,
+      "loss": 0.0600573,
+      "step": 21019
+    },
+    {
+      "epoch": 42.04,
+      "grad_norm": 1.3779077529907227,
+      "learning_rate": 2e-05,
+      "loss": 0.05684218,
+      "step": 21020
+    },
+    {
+      "epoch": 42.042,
+      "grad_norm": 1.3131364583969116,
+      "learning_rate": 2e-05,
+      "loss": 0.04408994,
+      "step": 21021
+    },
+    {
+      "epoch": 42.044,
+      "grad_norm": 1.862250804901123,
+      "learning_rate": 2e-05,
+      "loss": 0.05622796,
+      "step": 21022
+    },
+    {
+      "epoch": 42.046,
+      "grad_norm": 1.1200668811798096,
+      "learning_rate": 2e-05,
+      "loss": 0.03651465,
+      "step": 21023
+    },
+    {
+      "epoch": 42.048,
+      "grad_norm": 1.217587947845459,
+      "learning_rate": 2e-05,
+      "loss": 0.03259541,
+      "step": 21024
+    },
+    {
+      "epoch": 42.05,
+      "grad_norm": 1.5544425249099731,
+      "learning_rate": 2e-05,
+      "loss": 0.03680165,
+      "step": 21025
+    },
+    {
+      "epoch": 42.052,
+      "grad_norm": 1.274043083190918,
+      "learning_rate": 2e-05,
+      "loss": 0.04595189,
+      "step": 21026
+    },
+    {
+      "epoch": 42.054,
+      "grad_norm": 2.187490463256836,
+      "learning_rate": 2e-05,
+      "loss": 0.04805189,
+      "step": 21027
+    },
+    {
+      "epoch": 42.056,
+      "grad_norm": 1.2666808366775513,
+      "learning_rate": 2e-05,
+      "loss": 0.05547645,
+      "step": 21028
+    },
+    {
+      "epoch": 42.058,
+      "grad_norm": 1.4787068367004395,
+      "learning_rate": 2e-05,
+      "loss": 0.06114297,
+      "step": 21029
+    },
+    {
+      "epoch": 42.06,
+      "grad_norm": 1.4573709964752197,
+      "learning_rate": 2e-05,
+      "loss": 0.04937818,
+      "step": 21030
+    },
+    {
+      "epoch": 42.062,
+      "grad_norm": 1.2699495553970337,
+      "learning_rate": 2e-05,
+      "loss": 0.04790761,
+      "step": 21031
+    },
+    {
+      "epoch": 42.064,
+      "grad_norm": 1.1611040830612183,
+      "learning_rate": 2e-05,
+      "loss": 0.03836013,
+      "step": 21032
+    },
+    {
+      "epoch": 42.066,
+      "grad_norm": 1.162618637084961,
+      "learning_rate": 2e-05,
+      "loss": 0.04036701,
+      "step": 21033
+    },
+    {
+      "epoch": 42.068,
+      "grad_norm": 1.3911733627319336,
+      "learning_rate": 2e-05,
+      "loss": 0.05297327,
+      "step": 21034
+    },
+    {
+      "epoch": 42.07,
+      "grad_norm": 1.1646391153335571,
+      "learning_rate": 2e-05,
+      "loss": 0.05322537,
+      "step": 21035
+    },
+    {
+      "epoch": 42.072,
+      "grad_norm": 1.2405617237091064,
+      "learning_rate": 2e-05,
+      "loss": 0.03772672,
+      "step": 21036
+    },
+    {
+      "epoch": 42.074,
+      "grad_norm": 1.0591578483581543,
+      "learning_rate": 2e-05,
+      "loss": 0.03944032,
+      "step": 21037
+    },
+    {
+      "epoch": 42.076,
+      "grad_norm": 1.182476282119751,
+      "learning_rate": 2e-05,
+      "loss": 0.05112866,
+      "step": 21038
+    },
+    {
+      "epoch": 42.078,
+      "grad_norm": 1.7845994234085083,
+      "learning_rate": 2e-05,
+      "loss": 0.04743159,
+      "step": 21039
+    },
+    {
+      "epoch": 42.08,
+      "grad_norm": 1.1651856899261475,
+      "learning_rate": 2e-05,
+      "loss": 0.04408582,
+      "step": 21040
+    },
+    {
+      "epoch": 42.082,
+      "grad_norm": 1.3968032598495483,
+      "learning_rate": 2e-05,
+      "loss": 0.06172002,
+      "step": 21041
+    },
+    {
+      "epoch": 42.084,
+      "grad_norm": 1.2377245426177979,
+      "learning_rate": 2e-05,
+      "loss": 0.06204825,
+      "step": 21042
+    },
+    {
+      "epoch": 42.086,
+      "grad_norm": 1.1847991943359375,
+      "learning_rate": 2e-05,
+      "loss": 0.03809762,
+      "step": 21043
+    },
+    {
+      "epoch": 42.088,
+      "grad_norm": 1.6209152936935425,
+      "learning_rate": 2e-05,
+      "loss": 0.0557748,
+      "step": 21044
+    },
+    {
+      "epoch": 42.09,
+      "grad_norm": 1.1455878019332886,
+      "learning_rate": 2e-05,
+      "loss": 0.04782082,
+      "step": 21045
+    },
+    {
+      "epoch": 42.092,
+      "grad_norm": 1.1409224271774292,
+      "learning_rate": 2e-05,
+      "loss": 0.04084107,
+      "step": 21046
+    },
+    {
+      "epoch": 42.094,
+      "grad_norm": 1.6999880075454712,
+      "learning_rate": 2e-05,
+      "loss": 0.0430681,
+      "step": 21047
+    },
+    {
+      "epoch": 42.096,
+      "grad_norm": 1.3535499572753906,
+      "learning_rate": 2e-05,
+      "loss": 0.04580639,
+      "step": 21048
+    },
+    {
+      "epoch": 42.098,
+      "grad_norm": 1.6721376180648804,
+      "learning_rate": 2e-05,
+      "loss": 0.04870587,
+      "step": 21049
+    },
+    {
+      "epoch": 42.1,
+      "grad_norm": 3.2024240493774414,
+      "learning_rate": 2e-05,
+      "loss": 0.04476042,
+      "step": 21050
+    },
+    {
+      "epoch": 42.102,
+      "grad_norm": 0.9818119406700134,
+      "learning_rate": 2e-05,
+      "loss": 0.0436222,
+      "step": 21051
+    },
+    {
+      "epoch": 42.104,
+      "grad_norm": 1.2789314985275269,
+      "learning_rate": 2e-05,
+      "loss": 0.04346582,
+      "step": 21052
+    },
+    {
+      "epoch": 42.106,
+      "grad_norm": 1.2788130044937134,
+      "learning_rate": 2e-05,
+      "loss": 0.05709602,
+      "step": 21053
+    },
+    {
+      "epoch": 42.108,
+      "grad_norm": 2.102720260620117,
+      "learning_rate": 2e-05,
+      "loss": 0.06700805,
+      "step": 21054
+    },
+    {
+      "epoch": 42.11,
+      "grad_norm": 1.278295636177063,
+      "learning_rate": 2e-05,
+      "loss": 0.05733266,
+      "step": 21055
+    },
+    {
+      "epoch": 42.112,
+      "grad_norm": 1.3988900184631348,
+      "learning_rate": 2e-05,
+      "loss": 0.04855637,
+      "step": 21056
+    },
+    {
+      "epoch": 42.114,
+      "grad_norm": 1.2994776964187622,
+      "learning_rate": 2e-05,
+      "loss": 0.04653971,
+      "step": 21057
+    },
+    {
+      "epoch": 42.116,
+      "grad_norm": 1.2954206466674805,
+      "learning_rate": 2e-05,
+      "loss": 0.051019,
+      "step": 21058
+    },
+    {
+      "epoch": 42.118,
+      "grad_norm": 1.1361409425735474,
+      "learning_rate": 2e-05,
+      "loss": 0.03964051,
+      "step": 21059
+    },
+    {
+      "epoch": 42.12,
+      "grad_norm": 0.9857589602470398,
+      "learning_rate": 2e-05,
+      "loss": 0.0420249,
+      "step": 21060
+    },
+    {
+      "epoch": 42.122,
+      "grad_norm": 1.19060480594635,
+      "learning_rate": 2e-05,
+      "loss": 0.03876651,
+      "step": 21061
+    },
+    {
+      "epoch": 42.124,
+      "grad_norm": 1.0141627788543701,
+      "learning_rate": 2e-05,
+      "loss": 0.04508419,
+      "step": 21062
+    },
+    {
+      "epoch": 42.126,
+      "grad_norm": 1.3992326259613037,
+      "learning_rate": 2e-05,
+      "loss": 0.05133965,
+      "step": 21063
+    },
+    {
+      "epoch": 42.128,
+      "grad_norm": 1.2171835899353027,
+      "learning_rate": 2e-05,
+      "loss": 0.03061209,
+      "step": 21064
+    },
+    {
+      "epoch": 42.13,
+      "grad_norm": 0.9995740652084351,
+      "learning_rate": 2e-05,
+      "loss": 0.045886,
+      "step": 21065
+    },
+    {
+      "epoch": 42.132,
+      "grad_norm": 1.2155423164367676,
+      "learning_rate": 2e-05,
+      "loss": 0.05031782,
+      "step": 21066
+    },
+    {
+      "epoch": 42.134,
+      "grad_norm": 3.3475124835968018,
+      "learning_rate": 2e-05,
+      "loss": 0.05089858,
+      "step": 21067
+    },
+    {
+      "epoch": 42.136,
+      "grad_norm": 1.1842341423034668,
+      "learning_rate": 2e-05,
+      "loss": 0.03898914,
+      "step": 21068
+    },
+    {
+      "epoch": 42.138,
+      "grad_norm": 1.3484647274017334,
+      "learning_rate": 2e-05,
+      "loss": 0.05519821,
+      "step": 21069
+    },
+    {
+      "epoch": 42.14,
+      "grad_norm": 1.2679609060287476,
+      "learning_rate": 2e-05,
+      "loss": 0.04845376,
+      "step": 21070
+    },
+    {
+      "epoch": 42.142,
+      "grad_norm": 1.0729557275772095,
+      "learning_rate": 2e-05,
+      "loss": 0.03485009,
+      "step": 21071
+    },
+    {
+      "epoch": 42.144,
+      "grad_norm": 1.442091464996338,
+      "learning_rate": 2e-05,
+      "loss": 0.04761119,
+      "step": 21072
+    },
+    {
+      "epoch": 42.146,
+      "grad_norm": 1.2365809679031372,
+      "learning_rate": 2e-05,
+      "loss": 0.04327618,
+      "step": 21073
+    },
+    {
+      "epoch": 42.148,
+      "grad_norm": 1.1715463399887085,
+      "learning_rate": 2e-05,
+      "loss": 0.04974968,
+      "step": 21074
+    },
+    {
+      "epoch": 42.15,
+      "grad_norm": 1.1144353151321411,
+      "learning_rate": 2e-05,
+      "loss": 0.04348652,
+      "step": 21075
+    },
+    {
+      "epoch": 42.152,
+      "grad_norm": 1.4729726314544678,
+      "learning_rate": 2e-05,
+      "loss": 0.05958918,
+      "step": 21076
+    },
+    {
+      "epoch": 42.154,
+      "grad_norm": 1.1265980005264282,
+      "learning_rate": 2e-05,
+      "loss": 0.05273468,
+      "step": 21077
+    },
+    {
+      "epoch": 42.156,
+      "grad_norm": 1.2327011823654175,
+      "learning_rate": 2e-05,
+      "loss": 0.05835613,
+      "step": 21078
+    },
+    {
+      "epoch": 42.158,
+      "grad_norm": 1.3347694873809814,
+      "learning_rate": 2e-05,
+      "loss": 0.03290975,
+      "step": 21079
+    },
+    {
+      "epoch": 42.16,
+      "grad_norm": 1.2273168563842773,
+      "learning_rate": 2e-05,
+      "loss": 0.05718467,
+      "step": 21080
+    },
+    {
+      "epoch": 42.162,
+      "grad_norm": 1.1186082363128662,
+      "learning_rate": 2e-05,
+      "loss": 0.0496081,
+      "step": 21081
+    },
+    {
+      "epoch": 42.164,
+      "grad_norm": 1.4576385021209717,
+      "learning_rate": 2e-05,
+      "loss": 0.0473141,
+      "step": 21082
+    },
+    {
+      "epoch": 42.166,
+      "grad_norm": 1.3093568086624146,
+      "learning_rate": 2e-05,
+      "loss": 0.05596102,
+      "step": 21083
+    },
+    {
+      "epoch": 42.168,
+      "grad_norm": 1.2706962823867798,
+      "learning_rate": 2e-05,
+      "loss": 0.050145,
+      "step": 21084
+    },
+    {
+      "epoch": 42.17,
+      "grad_norm": 1.4902071952819824,
+      "learning_rate": 2e-05,
+      "loss": 0.04793771,
+      "step": 21085
+    },
+    {
+      "epoch": 42.172,
+      "grad_norm": 1.174926519393921,
+      "learning_rate": 2e-05,
+      "loss": 0.05029378,
+      "step": 21086
+    },
+    {
+      "epoch": 42.174,
+      "grad_norm": 1.2596014738082886,
+      "learning_rate": 2e-05,
+      "loss": 0.03948942,
+      "step": 21087
+    },
+    {
+      "epoch": 42.176,
+      "grad_norm": 1.0307563543319702,
+      "learning_rate": 2e-05,
+      "loss": 0.03940136,
+      "step": 21088
+    },
+    {
+      "epoch": 42.178,
+      "grad_norm": 1.1724040508270264,
+      "learning_rate": 2e-05,
+      "loss": 0.04482014,
+      "step": 21089
+    },
+    {
+      "epoch": 42.18,
+      "grad_norm": 1.143633484840393,
+      "learning_rate": 2e-05,
+      "loss": 0.03999674,
+      "step": 21090
+    },
+    {
+      "epoch": 42.182,
+      "grad_norm": 1.217869758605957,
+      "learning_rate": 2e-05,
+      "loss": 0.03607331,
+      "step": 21091
+    },
+    {
+      "epoch": 42.184,
+      "grad_norm": 1.5815616846084595,
+      "learning_rate": 2e-05,
+      "loss": 0.03352413,
+      "step": 21092
+    },
+    {
+      "epoch": 42.186,
+      "grad_norm": 1.3018903732299805,
+      "learning_rate": 2e-05,
+      "loss": 0.04986539,
+      "step": 21093
+    },
+    {
+      "epoch": 42.188,
+      "grad_norm": 0.9588651657104492,
+      "learning_rate": 2e-05,
+      "loss": 0.03948629,
+      "step": 21094
+    },
+    {
+      "epoch": 42.19,
+      "grad_norm": 1.2913947105407715,
+      "learning_rate": 2e-05,
+      "loss": 0.04663259,
+      "step": 21095
+    },
+    {
+      "epoch": 42.192,
+      "grad_norm": 1.219344139099121,
+      "learning_rate": 2e-05,
+      "loss": 0.0564451,
+      "step": 21096
+    },
+    {
+      "epoch": 42.194,
+      "grad_norm": 1.4023454189300537,
+      "learning_rate": 2e-05,
+      "loss": 0.06140105,
+      "step": 21097
+    },
+    {
+      "epoch": 42.196,
+      "grad_norm": 1.1653475761413574,
+      "learning_rate": 2e-05,
+      "loss": 0.03363206,
+      "step": 21098
+    },
+    {
+      "epoch": 42.198,
+      "grad_norm": 1.1109018325805664,
+      "learning_rate": 2e-05,
+      "loss": 0.0404594,
+      "step": 21099
+    },
+    {
+      "epoch": 42.2,
+      "grad_norm": 1.053236722946167,
+      "learning_rate": 2e-05,
+      "loss": 0.02968258,
+      "step": 21100
+    },
+    {
+      "epoch": 42.202,
+      "grad_norm": 1.2577223777770996,
+      "learning_rate": 2e-05,
+      "loss": 0.05090709,
+      "step": 21101
+    },
+    {
+      "epoch": 42.204,
+      "grad_norm": 1.3133550882339478,
+      "learning_rate": 2e-05,
+      "loss": 0.05599451,
+      "step": 21102
+    },
+    {
+      "epoch": 42.206,
+      "grad_norm": 2.2289819717407227,
+      "learning_rate": 2e-05,
+      "loss": 0.04835991,
+      "step": 21103
+    },
+    {
+      "epoch": 42.208,
+      "grad_norm": 1.7648617029190063,
+      "learning_rate": 2e-05,
+      "loss": 0.04039288,
+      "step": 21104
+    },
+    {
+      "epoch": 42.21,
+      "grad_norm": 1.5205116271972656,
+      "learning_rate": 2e-05,
+      "loss": 0.06169431,
+      "step": 21105
+    },
+    {
+      "epoch": 42.212,
+      "grad_norm": 1.573837161064148,
+      "learning_rate": 2e-05,
+      "loss": 0.06739548,
+      "step": 21106
+    },
+    {
+      "epoch": 42.214,
+      "grad_norm": 1.3160827159881592,
+      "learning_rate": 2e-05,
+      "loss": 0.04701831,
+      "step": 21107
+    },
+    {
+      "epoch": 42.216,
+      "grad_norm": 1.2954484224319458,
+      "learning_rate": 2e-05,
+      "loss": 0.05630885,
+      "step": 21108
+    },
+    {
+      "epoch": 42.218,
+      "grad_norm": 1.2000118494033813,
+      "learning_rate": 2e-05,
+      "loss": 0.0458281,
+      "step": 21109
+    },
+    {
+      "epoch": 42.22,
+      "grad_norm": 1.564306378364563,
+      "learning_rate": 2e-05,
+      "loss": 0.05043546,
+      "step": 21110
+    },
+    {
+      "epoch": 42.222,
+      "grad_norm": 1.0061910152435303,
+      "learning_rate": 2e-05,
+      "loss": 0.03526705,
+      "step": 21111
+    },
+    {
+      "epoch": 42.224,
+      "grad_norm": 1.4038654565811157,
+      "learning_rate": 2e-05,
+      "loss": 0.05683837,
+      "step": 21112
+    },
+    {
+      "epoch": 42.226,
+      "grad_norm": 1.0469942092895508,
+      "learning_rate": 2e-05,
+      "loss": 0.04538111,
+      "step": 21113
+    },
+    {
+      "epoch": 42.228,
+      "grad_norm": 1.2826262712478638,
+      "learning_rate": 2e-05,
+      "loss": 0.04978053,
+      "step": 21114
+    },
+    {
+      "epoch": 42.23,
+      "grad_norm": 1.3049055337905884,
+      "learning_rate": 2e-05,
+      "loss": 0.04201498,
+      "step": 21115
+    },
+    {
+      "epoch": 42.232,
+      "grad_norm": 1.076093077659607,
+      "learning_rate": 2e-05,
+      "loss": 0.04795041,
+      "step": 21116
+    },
+    {
+      "epoch": 42.234,
+      "grad_norm": 1.1158887147903442,
+      "learning_rate": 2e-05,
+      "loss": 0.03639564,
+      "step": 21117
+    },
+    {
+      "epoch": 42.236,
+      "grad_norm": 1.251383900642395,
+      "learning_rate": 2e-05,
+      "loss": 0.04058103,
+      "step": 21118
+    },
+    {
+      "epoch": 42.238,
+      "grad_norm": 1.0459401607513428,
+      "learning_rate": 2e-05,
+      "loss": 0.04287563,
+      "step": 21119
+    },
+    {
+      "epoch": 42.24,
+      "grad_norm": 1.3883095979690552,
+      "learning_rate": 2e-05,
+      "loss": 0.06586547,
+      "step": 21120
+    },
+    {
+      "epoch": 42.242,
+      "grad_norm": 1.1663405895233154,
+      "learning_rate": 2e-05,
+      "loss": 0.03624522,
+      "step": 21121
+    },
+    {
+      "epoch": 42.244,
+      "grad_norm": 1.8142751455307007,
+      "learning_rate": 2e-05,
+      "loss": 0.0545116,
+      "step": 21122
+    },
+    {
+      "epoch": 42.246,
+      "grad_norm": 2.438311815261841,
+      "learning_rate": 2e-05,
+      "loss": 0.05591767,
+      "step": 21123
+    },
+    {
+      "epoch": 42.248,
+      "grad_norm": 1.2450895309448242,
+      "learning_rate": 2e-05,
+      "loss": 0.04676469,
+      "step": 21124
+    },
+    {
+      "epoch": 42.25,
+      "grad_norm": 1.227967619895935,
+      "learning_rate": 2e-05,
+      "loss": 0.05129366,
+      "step": 21125
+    },
+    {
+      "epoch": 42.252,
+      "grad_norm": 1.1938300132751465,
+      "learning_rate": 2e-05,
+      "loss": 0.05615389,
+      "step": 21126
+    },
+    {
+      "epoch": 42.254,
+      "grad_norm": 1.1461693048477173,
+      "learning_rate": 2e-05,
+      "loss": 0.04713152,
+      "step": 21127
+    },
+    {
+      "epoch": 42.256,
+      "grad_norm": 1.187277913093567,
+      "learning_rate": 2e-05,
+      "loss": 0.04280775,
+      "step": 21128
+    },
+    {
+      "epoch": 42.258,
+      "grad_norm": 1.5151171684265137,
+      "learning_rate": 2e-05,
+      "loss": 0.04615886,
+      "step": 21129
+    },
+    {
+      "epoch": 42.26,
+      "grad_norm": 1.1889163255691528,
+      "learning_rate": 2e-05,
+      "loss": 0.0394232,
+      "step": 21130
+    },
+    {
+      "epoch": 42.262,
+      "grad_norm": 1.3880109786987305,
+      "learning_rate": 2e-05,
+      "loss": 0.06245904,
+      "step": 21131
+    },
+    {
+      "epoch": 42.264,
+      "grad_norm": 1.1990442276000977,
+      "learning_rate": 2e-05,
+      "loss": 0.05073933,
+      "step": 21132
+    },
+    {
+      "epoch": 42.266,
+      "grad_norm": 1.3353129625320435,
+      "learning_rate": 2e-05,
+      "loss": 0.04800763,
+      "step": 21133
+    },
+    {
+      "epoch": 42.268,
+      "grad_norm": 1.1701955795288086,
+      "learning_rate": 2e-05,
+      "loss": 0.05479547,
+      "step": 21134
+    },
+    {
+      "epoch": 42.27,
+      "grad_norm": 1.744545578956604,
+      "learning_rate": 2e-05,
+      "loss": 0.04801206,
+      "step": 21135
+    },
+    {
+      "epoch": 42.272,
+      "grad_norm": 1.2501215934753418,
+      "learning_rate": 2e-05,
+      "loss": 0.05587947,
+      "step": 21136
+    },
+    {
+      "epoch": 42.274,
+      "grad_norm": 1.1730159521102905,
+      "learning_rate": 2e-05,
+      "loss": 0.03659301,
+      "step": 21137
+    },
+    {
+      "epoch": 42.276,
+      "grad_norm": 1.2556431293487549,
+      "learning_rate": 2e-05,
+      "loss": 0.06406794,
+      "step": 21138
+    },
+    {
+      "epoch": 42.278,
+      "grad_norm": 0.9709025621414185,
+      "learning_rate": 2e-05,
+      "loss": 0.02665532,
+      "step": 21139
+    },
+    {
+      "epoch": 42.28,
+      "grad_norm": 1.0300137996673584,
+      "learning_rate": 2e-05,
+      "loss": 0.0415834,
+      "step": 21140
+    },
+    {
+      "epoch": 42.282,
+      "grad_norm": 1.334498643875122,
+      "learning_rate": 2e-05,
+      "loss": 0.04262113,
+      "step": 21141
+    },
+    {
+      "epoch": 42.284,
+      "grad_norm": 0.9960622191429138,
+      "learning_rate": 2e-05,
+      "loss": 0.03293207,
+      "step": 21142
+    },
+    {
+      "epoch": 42.286,
+      "grad_norm": 1.19757878780365,
+      "learning_rate": 2e-05,
+      "loss": 0.05076621,
+      "step": 21143
+    },
+    {
+      "epoch": 42.288,
+      "grad_norm": 1.1853511333465576,
+      "learning_rate": 2e-05,
+      "loss": 0.05421963,
+      "step": 21144
+    },
+    {
+      "epoch": 42.29,
+      "grad_norm": 1.0343493223190308,
+      "learning_rate": 2e-05,
+      "loss": 0.03314459,
+      "step": 21145
+    },
+    {
+      "epoch": 42.292,
+      "grad_norm": 6.21450662612915,
+      "learning_rate": 2e-05,
+      "loss": 0.05752373,
+      "step": 21146
+    },
+    {
+      "epoch": 42.294,
+      "grad_norm": 1.1545538902282715,
+      "learning_rate": 2e-05,
+      "loss": 0.04223744,
+      "step": 21147
+    },
+    {
+      "epoch": 42.296,
+      "grad_norm": 1.1654874086380005,
+      "learning_rate": 2e-05,
+      "loss": 0.03923909,
+      "step": 21148
+    },
+    {
+      "epoch": 42.298,
+      "grad_norm": 1.1227787733078003,
+      "learning_rate": 2e-05,
+      "loss": 0.03867494,
+      "step": 21149
+    },
+    {
+      "epoch": 42.3,
+      "grad_norm": 1.422080397605896,
+      "learning_rate": 2e-05,
+      "loss": 0.05246984,
+      "step": 21150
+    },
+    {
+      "epoch": 42.302,
+      "grad_norm": 1.0861878395080566,
+      "learning_rate": 2e-05,
+      "loss": 0.03628934,
+      "step": 21151
+    },
+    {
+      "epoch": 42.304,
+      "grad_norm": 1.113157868385315,
+      "learning_rate": 2e-05,
+      "loss": 0.03985634,
+      "step": 21152
+    },
+    {
+      "epoch": 42.306,
+      "grad_norm": 1.102103352546692,
+      "learning_rate": 2e-05,
+      "loss": 0.04545946,
+      "step": 21153
+    },
+    {
+      "epoch": 42.308,
+      "grad_norm": 2.6324360370635986,
+      "learning_rate": 2e-05,
+      "loss": 0.052878,
+      "step": 21154
+    },
+    {
+      "epoch": 42.31,
+      "grad_norm": 3.8525304794311523,
+      "learning_rate": 2e-05,
+      "loss": 0.06736265,
+      "step": 21155
+    },
+    {
+      "epoch": 42.312,
+      "grad_norm": 1.187397837638855,
+      "learning_rate": 2e-05,
+      "loss": 0.04645222,
+      "step": 21156
+    },
+    {
+      "epoch": 42.314,
+      "grad_norm": 0.9652663469314575,
+      "learning_rate": 2e-05,
+      "loss": 0.03505144,
+      "step": 21157
+    },
+    {
+      "epoch": 42.316,
+      "grad_norm": 0.9658792018890381,
+      "learning_rate": 2e-05,
+      "loss": 0.0281552,
+      "step": 21158
+    },
+    {
+      "epoch": 42.318,
+      "grad_norm": 2.401021718978882,
+      "learning_rate": 2e-05,
+      "loss": 0.05007901,
+      "step": 21159
+    },
+    {
+      "epoch": 42.32,
+      "grad_norm": 1.2055039405822754,
+      "learning_rate": 2e-05,
+      "loss": 0.05365395,
+      "step": 21160
+    },
+    {
+      "epoch": 42.322,
+      "grad_norm": 3.1883809566497803,
+      "learning_rate": 2e-05,
+      "loss": 0.05920103,
+      "step": 21161
+    },
+    {
+      "epoch": 42.324,
+      "grad_norm": 1.5328859090805054,
+      "learning_rate": 2e-05,
+      "loss": 0.04159598,
+      "step": 21162
+    },
+    {
+      "epoch": 42.326,
+      "grad_norm": 1.2868179082870483,
+      "learning_rate": 2e-05,
+      "loss": 0.05470748,
+      "step": 21163
+    },
+    {
+      "epoch": 42.328,
+      "grad_norm": 1.2141040563583374,
+      "learning_rate": 2e-05,
+      "loss": 0.0402927,
+      "step": 21164
+    },
+    {
+      "epoch": 42.33,
+      "grad_norm": 1.7615190744400024,
+      "learning_rate": 2e-05,
+      "loss": 0.03097968,
+      "step": 21165
+    },
+    {
+      "epoch": 42.332,
+      "grad_norm": 1.5224100351333618,
+      "learning_rate": 2e-05,
+      "loss": 0.04342334,
+      "step": 21166
+    },
+    {
+      "epoch": 42.334,
+      "grad_norm": 1.1339821815490723,
+      "learning_rate": 2e-05,
+      "loss": 0.03738585,
+      "step": 21167
+    },
+    {
+      "epoch": 42.336,
+      "grad_norm": 1.207984209060669,
+      "learning_rate": 2e-05,
+      "loss": 0.0441434,
+      "step": 21168
+    },
+    {
+      "epoch": 42.338,
+      "grad_norm": 3.2836313247680664,
+      "learning_rate": 2e-05,
+      "loss": 0.03989324,
+      "step": 21169
+    },
+    {
+      "epoch": 42.34,
+      "grad_norm": 1.7195615768432617,
+      "learning_rate": 2e-05,
+      "loss": 0.04979802,
+      "step": 21170
+    },
+    {
+      "epoch": 42.342,
+      "grad_norm": 1.368320107460022,
+      "learning_rate": 2e-05,
+      "loss": 0.0440874,
+      "step": 21171
+    },
+    {
+      "epoch": 42.344,
+      "grad_norm": 1.2935134172439575,
+      "learning_rate": 2e-05,
+      "loss": 0.06269786,
+      "step": 21172
+    },
+    {
+      "epoch": 42.346,
+      "grad_norm": 1.5454837083816528,
+      "learning_rate": 2e-05,
+      "loss": 0.05150343,
+      "step": 21173
+    },
+    {
+      "epoch": 42.348,
+      "grad_norm": 1.0489261150360107,
+      "learning_rate": 2e-05,
+      "loss": 0.03613845,
+      "step": 21174
+    },
+    {
+      "epoch": 42.35,
+      "grad_norm": 2.1777830123901367,
+      "learning_rate": 2e-05,
+      "loss": 0.05178384,
+      "step": 21175
+    },
+    {
+      "epoch": 42.352,
+      "grad_norm": 1.3198027610778809,
+      "learning_rate": 2e-05,
+      "loss": 0.04619918,
+      "step": 21176
+    },
+    {
+      "epoch": 42.354,
+      "grad_norm": 1.1379380226135254,
+      "learning_rate": 2e-05,
+      "loss": 0.0368214,
+      "step": 21177
+    },
+    {
+      "epoch": 42.356,
+      "grad_norm": 2.0367469787597656,
+      "learning_rate": 2e-05,
+      "loss": 0.04297212,
+      "step": 21178
+    },
+    {
+      "epoch": 42.358,
+      "grad_norm": 1.455183506011963,
+      "learning_rate": 2e-05,
+      "loss": 0.06377006,
+      "step": 21179
+    },
+    {
+      "epoch": 42.36,
+      "grad_norm": 1.1051256656646729,
+      "learning_rate": 2e-05,
+      "loss": 0.03730104,
+      "step": 21180
+    },
+    {
+      "epoch": 42.362,
+      "grad_norm": 1.163114070892334,
+      "learning_rate": 2e-05,
+      "loss": 0.04407148,
+      "step": 21181
+    },
+    {
+      "epoch": 42.364,
+      "grad_norm": 1.0142223834991455,
+      "learning_rate": 2e-05,
+      "loss": 0.03786562,
+      "step": 21182
+    },
+    {
+      "epoch": 42.366,
+      "grad_norm": 1.0811436176300049,
+      "learning_rate": 2e-05,
+      "loss": 0.03777923,
+      "step": 21183
+    },
+    {
+      "epoch": 42.368,
+      "grad_norm": 1.1761528253555298,
+      "learning_rate": 2e-05,
+      "loss": 0.04415475,
+      "step": 21184
+    },
+    {
+      "epoch": 42.37,
+      "grad_norm": 0.9909497499465942,
+      "learning_rate": 2e-05,
+      "loss": 0.04150863,
+      "step": 21185
+    },
+    {
+      "epoch": 42.372,
+      "grad_norm": 1.077889323234558,
+      "learning_rate": 2e-05,
+      "loss": 0.04909585,
+      "step": 21186
+    },
+    {
+      "epoch": 42.374,
+      "grad_norm": 1.0327900648117065,
+      "learning_rate": 2e-05,
+      "loss": 0.03843714,
+      "step": 21187
+    },
+    {
+      "epoch": 42.376,
+      "grad_norm": 1.582082748413086,
+      "learning_rate": 2e-05,
+      "loss": 0.04728288,
+      "step": 21188
+    },
+    {
+      "epoch": 42.378,
+      "grad_norm": 1.148047924041748,
+      "learning_rate": 2e-05,
+      "loss": 0.04307776,
+      "step": 21189
+    },
+    {
+      "epoch": 42.38,
+      "grad_norm": 1.0733706951141357,
+      "learning_rate": 2e-05,
+      "loss": 0.04168211,
+      "step": 21190
+    },
+    {
+      "epoch": 42.382,
+      "grad_norm": 1.2072162628173828,
+      "learning_rate": 2e-05,
+      "loss": 0.05512146,
+      "step": 21191
+    },
+    {
+      "epoch": 42.384,
+      "grad_norm": 1.1533020734786987,
+      "learning_rate": 2e-05,
+      "loss": 0.03808629,
+      "step": 21192
+    },
+    {
+      "epoch": 42.386,
+      "grad_norm": 1.307199478149414,
+      "learning_rate": 2e-05,
+      "loss": 0.05116679,
+      "step": 21193
+    },
+    {
+      "epoch": 42.388,
+      "grad_norm": 1.3861874341964722,
+      "learning_rate": 2e-05,
+      "loss": 0.06823573,
+      "step": 21194
+    },
+    {
+      "epoch": 42.39,
+      "grad_norm": 1.1966004371643066,
+      "learning_rate": 2e-05,
+      "loss": 0.0478618,
+      "step": 21195
+    },
+    {
+      "epoch": 42.392,
+      "grad_norm": 1.1855981349945068,
+      "learning_rate": 2e-05,
+      "loss": 0.0414066,
+      "step": 21196
+    },
+    {
+      "epoch": 42.394,
+      "grad_norm": 1.1840327978134155,
+      "learning_rate": 2e-05,
+      "loss": 0.05320561,
+      "step": 21197
+    },
+    {
+      "epoch": 42.396,
+      "grad_norm": 1.4169137477874756,
+      "learning_rate": 2e-05,
+      "loss": 0.05518955,
+      "step": 21198
+    },
+    {
+      "epoch": 42.398,
+      "grad_norm": 1.1781421899795532,
+      "learning_rate": 2e-05,
+      "loss": 0.04477818,
+      "step": 21199
+    },
+    {
+      "epoch": 42.4,
+      "grad_norm": 1.092429757118225,
+      "learning_rate": 2e-05,
+      "loss": 0.04341368,
+      "step": 21200
+    },
+    {
+      "epoch": 42.402,
+      "grad_norm": 1.3216253519058228,
+      "learning_rate": 2e-05,
+      "loss": 0.05887017,
+      "step": 21201
+    },
+    {
+      "epoch": 42.404,
+      "grad_norm": 1.205215573310852,
+      "learning_rate": 2e-05,
+      "loss": 0.04666806,
+      "step": 21202
+    },
+    {
+      "epoch": 42.406,
+      "grad_norm": 1.2955050468444824,
+      "learning_rate": 2e-05,
+      "loss": 0.05575846,
+      "step": 21203
+    },
+    {
+      "epoch": 42.408,
+      "grad_norm": 2.5864737033843994,
+      "learning_rate": 2e-05,
+      "loss": 0.03791555,
+      "step": 21204
+    },
+    {
+      "epoch": 42.41,
+      "grad_norm": 1.188843011856079,
+      "learning_rate": 2e-05,
+      "loss": 0.05555812,
+      "step": 21205
+    },
+    {
+      "epoch": 42.412,
+      "grad_norm": 0.9318320751190186,
+      "learning_rate": 2e-05,
+      "loss": 0.02653901,
+      "step": 21206
+    },
+    {
+      "epoch": 42.414,
+      "grad_norm": 1.1733531951904297,
+      "learning_rate": 2e-05,
+      "loss": 0.04083906,
+      "step": 21207
+    },
+    {
+      "epoch": 42.416,
+      "grad_norm": 1.1449897289276123,
+      "learning_rate": 2e-05,
+      "loss": 0.04881506,
+      "step": 21208
+    },
+    {
+      "epoch": 42.418,
+      "grad_norm": 1.965896487236023,
+      "learning_rate": 2e-05,
+      "loss": 0.04963106,
+      "step": 21209
+    },
+    {
+      "epoch": 42.42,
+      "grad_norm": 1.0945767164230347,
+      "learning_rate": 2e-05,
+      "loss": 0.04849849,
+      "step": 21210
+    },
+    {
+      "epoch": 42.422,
+      "grad_norm": 1.308809757232666,
+      "learning_rate": 2e-05,
+      "loss": 0.06706718,
+      "step": 21211
+    },
+    {
+      "epoch": 42.424,
+      "grad_norm": 1.53829026222229,
+      "learning_rate": 2e-05,
+      "loss": 0.06904881,
+      "step": 21212
+    },
+    {
+      "epoch": 42.426,
+      "grad_norm": 1.2670159339904785,
+      "learning_rate": 2e-05,
+      "loss": 0.05303366,
+      "step": 21213
+    },
+    {
+      "epoch": 42.428,
+      "grad_norm": 1.2741113901138306,
+      "learning_rate": 2e-05,
+      "loss": 0.0492133,
+      "step": 21214
+    },
+    {
+      "epoch": 42.43,
+      "grad_norm": 1.2569606304168701,
+      "learning_rate": 2e-05,
+      "loss": 0.06223689,
+      "step": 21215
+    },
+    {
+      "epoch": 42.432,
+      "grad_norm": 1.107643485069275,
+      "learning_rate": 2e-05,
+      "loss": 0.04277761,
+      "step": 21216
+    },
+    {
+      "epoch": 42.434,
+      "grad_norm": 1.122188925743103,
+      "learning_rate": 2e-05,
+      "loss": 0.04128901,
+      "step": 21217
+    },
+    {
+      "epoch": 42.436,
+      "grad_norm": 1.1837259531021118,
+      "learning_rate": 2e-05,
+      "loss": 0.04776137,
+      "step": 21218
+    },
+    {
+      "epoch": 42.438,
+      "grad_norm": 3.169210910797119,
+      "learning_rate": 2e-05,
+      "loss": 0.05383972,
+      "step": 21219
+    },
+    {
+      "epoch": 42.44,
+      "grad_norm": 0.9989545941352844,
+      "learning_rate": 2e-05,
+      "loss": 0.03308531,
+      "step": 21220
+    },
+    {
+      "epoch": 42.442,
+      "grad_norm": 1.1423425674438477,
+      "learning_rate": 2e-05,
+      "loss": 0.04700923,
+      "step": 21221
+    },
+    {
+      "epoch": 42.444,
+      "grad_norm": 1.0154725313186646,
+      "learning_rate": 2e-05,
+      "loss": 0.03601583,
+      "step": 21222
+    },
+    {
+      "epoch": 42.446,
+      "grad_norm": 1.450246810913086,
+      "learning_rate": 2e-05,
+      "loss": 0.0615345,
+      "step": 21223
+    },
+    {
+      "epoch": 42.448,
+      "grad_norm": 1.1445060968399048,
+      "learning_rate": 2e-05,
+      "loss": 0.03595731,
+      "step": 21224
+    },
+    {
+      "epoch": 42.45,
+      "grad_norm": 1.6389358043670654,
+      "learning_rate": 2e-05,
+      "loss": 0.0496392,
+      "step": 21225
+    },
+    {
+      "epoch": 42.452,
+      "grad_norm": 1.141597867012024,
+      "learning_rate": 2e-05,
+      "loss": 0.03886272,
+      "step": 21226
+    },
+    {
+      "epoch": 42.454,
+      "grad_norm": 2.0984458923339844,
+      "learning_rate": 2e-05,
+      "loss": 0.0533991,
+      "step": 21227
+    },
+    {
+      "epoch": 42.456,
+      "grad_norm": 1.5619630813598633,
+      "learning_rate": 2e-05,
+      "loss": 0.0451661,
+      "step": 21228
+    },
+    {
+      "epoch": 42.458,
+      "grad_norm": 1.1230692863464355,
+      "learning_rate": 2e-05,
+      "loss": 0.02888122,
+      "step": 21229
+    },
+    {
+      "epoch": 42.46,
+      "grad_norm": 1.5001888275146484,
+      "learning_rate": 2e-05,
+      "loss": 0.04741661,
+      "step": 21230
+    },
+    {
+      "epoch": 42.462,
+      "grad_norm": 1.1775398254394531,
+      "learning_rate": 2e-05,
+      "loss": 0.05408292,
+      "step": 21231
+    },
+    {
+      "epoch": 42.464,
+      "grad_norm": 1.1655455827713013,
+      "learning_rate": 2e-05,
+      "loss": 0.04785192,
+      "step": 21232
+    },
+    {
+      "epoch": 42.466,
+      "grad_norm": 1.3385460376739502,
+      "learning_rate": 2e-05,
+      "loss": 0.04888346,
+      "step": 21233
+    },
+    {
+      "epoch": 42.468,
+      "grad_norm": 1.5374442338943481,
+      "learning_rate": 2e-05,
+      "loss": 0.048363,
+      "step": 21234
+    },
+    {
+      "epoch": 42.47,
+      "grad_norm": 1.1293838024139404,
+      "learning_rate": 2e-05,
+      "loss": 0.03480672,
+      "step": 21235
+    },
+    {
+      "epoch": 42.472,
+      "grad_norm": 1.4131519794464111,
+      "learning_rate": 2e-05,
+      "loss": 0.059271,
+      "step": 21236
+    },
+    {
+      "epoch": 42.474,
+      "grad_norm": 1.091775894165039,
+      "learning_rate": 2e-05,
+      "loss": 0.04225942,
+      "step": 21237
+    },
+    {
+      "epoch": 42.476,
+      "grad_norm": 1.5270899534225464,
+      "learning_rate": 2e-05,
+      "loss": 0.05104741,
+      "step": 21238
+    },
+    {
+      "epoch": 42.478,
+      "grad_norm": 1.4533936977386475,
+      "learning_rate": 2e-05,
+      "loss": 0.04931673,
+      "step": 21239
+    },
+    {
+      "epoch": 42.48,
+      "grad_norm": 1.187991738319397,
+      "learning_rate": 2e-05,
+      "loss": 0.05418856,
+      "step": 21240
+    },
+    {
+      "epoch": 42.482,
+      "grad_norm": 1.3054866790771484,
+      "learning_rate": 2e-05,
+      "loss": 0.05543121,
+      "step": 21241
+    },
+    {
+      "epoch": 42.484,
+      "grad_norm": 1.1094039678573608,
+      "learning_rate": 2e-05,
+      "loss": 0.03574315,
+      "step": 21242
+    },
+    {
+      "epoch": 42.486,
+      "grad_norm": 1.1816613674163818,
+      "learning_rate": 2e-05,
+      "loss": 0.04213757,
+      "step": 21243
+    },
+    {
+      "epoch": 42.488,
+      "grad_norm": 1.0031068325042725,
+      "learning_rate": 2e-05,
+      "loss": 0.03389908,
+      "step": 21244
+    },
+    {
+      "epoch": 42.49,
+      "grad_norm": 1.4081437587738037,
+      "learning_rate": 2e-05,
+      "loss": 0.05755498,
+      "step": 21245
+    },
+    {
+      "epoch": 42.492,
+      "grad_norm": 1.0443217754364014,
+      "learning_rate": 2e-05,
+      "loss": 0.0272135,
+      "step": 21246
+    },
+    {
+      "epoch": 42.494,
+      "grad_norm": 2.372973918914795,
+      "learning_rate": 2e-05,
+      "loss": 0.0727305,
+      "step": 21247
+    },
+    {
+      "epoch": 42.496,
+      "grad_norm": 1.1258074045181274,
+      "learning_rate": 2e-05,
+      "loss": 0.04463324,
+      "step": 21248
+    },
+    {
+      "epoch": 42.498,
+      "grad_norm": 1.0638490915298462,
+      "learning_rate": 2e-05,
+      "loss": 0.04064821,
+      "step": 21249
+    },
+    {
+      "epoch": 42.5,
+      "grad_norm": 2.074948787689209,
+      "learning_rate": 2e-05,
+      "loss": 0.04368262,
+      "step": 21250
+    },
+    {
+      "epoch": 42.502,
+      "grad_norm": 1.1462494134902954,
+      "learning_rate": 2e-05,
+      "loss": 0.05238108,
+      "step": 21251
+    },
+    {
+      "epoch": 42.504,
+      "grad_norm": 1.1089760065078735,
+      "learning_rate": 2e-05,
+      "loss": 0.05240737,
+      "step": 21252
+    },
+    {
+      "epoch": 42.506,
+      "grad_norm": 1.010260820388794,
+      "learning_rate": 2e-05,
+      "loss": 0.0344992,
+      "step": 21253
+    },
+    {
+      "epoch": 42.508,
+      "grad_norm": 1.2549633979797363,
+      "learning_rate": 2e-05,
+      "loss": 0.05022692,
+      "step": 21254
+    },
+    {
+      "epoch": 42.51,
+      "grad_norm": 1.307962417602539,
+      "learning_rate": 2e-05,
+      "loss": 0.05273689,
+      "step": 21255
+    },
+    {
+      "epoch": 42.512,
+      "grad_norm": 1.516259789466858,
+      "learning_rate": 2e-05,
+      "loss": 0.03713296,
+      "step": 21256
+    },
+    {
+      "epoch": 42.514,
+      "grad_norm": 1.1646299362182617,
+      "learning_rate": 2e-05,
+      "loss": 0.04224031,
+      "step": 21257
+    },
+    {
+      "epoch": 42.516,
+      "grad_norm": 1.3015379905700684,
+      "learning_rate": 2e-05,
+      "loss": 0.05681949,
+      "step": 21258
+    },
+    {
+      "epoch": 42.518,
+      "grad_norm": 1.2873107194900513,
+      "learning_rate": 2e-05,
+      "loss": 0.06247327,
+      "step": 21259
+    },
+    {
+      "epoch": 42.52,
+      "grad_norm": 1.2081667184829712,
+      "learning_rate": 2e-05,
+      "loss": 0.04261228,
+      "step": 21260
+    },
+    {
+      "epoch": 42.522,
+      "grad_norm": 2.0166215896606445,
+      "learning_rate": 2e-05,
+      "loss": 0.05025642,
+      "step": 21261
+    },
+    {
+      "epoch": 42.524,
+      "grad_norm": 1.4932124614715576,
+      "learning_rate": 2e-05,
+      "loss": 0.04905674,
+      "step": 21262
+    },
+    {
+      "epoch": 42.526,
+      "grad_norm": 1.2398021221160889,
+      "learning_rate": 2e-05,
+      "loss": 0.0445546,
+      "step": 21263
+    },
+    {
+      "epoch": 42.528,
+      "grad_norm": 7.126203536987305,
+      "learning_rate": 2e-05,
+      "loss": 0.05969048,
+      "step": 21264
+    },
+    {
+      "epoch": 42.53,
+      "grad_norm": 1.2210969924926758,
+      "learning_rate": 2e-05,
+      "loss": 0.03673073,
+      "step": 21265
+    },
+    {
+      "epoch": 42.532,
+      "grad_norm": 1.105908751487732,
+      "learning_rate": 2e-05,
+      "loss": 0.04691067,
+      "step": 21266
+    },
+    {
+      "epoch": 42.534,
+      "grad_norm": 1.393251657485962,
+      "learning_rate": 2e-05,
+      "loss": 0.04946836,
+      "step": 21267
+    },
+    {
+      "epoch": 42.536,
+      "grad_norm": 1.284510850906372,
+      "learning_rate": 2e-05,
+      "loss": 0.04197853,
+      "step": 21268
+    },
+    {
+      "epoch": 42.538,
+      "grad_norm": 1.0972687005996704,
+      "learning_rate": 2e-05,
+      "loss": 0.03820524,
+      "step": 21269
+    },
+    {
+      "epoch": 42.54,
+      "grad_norm": 1.1322238445281982,
+      "learning_rate": 2e-05,
+      "loss": 0.04426937,
+      "step": 21270
+    },
+    {
+      "epoch": 42.542,
+      "grad_norm": 2.50138783454895,
+      "learning_rate": 2e-05,
+      "loss": 0.06171612,
+      "step": 21271
+    },
+    {
+      "epoch": 42.544,
+      "grad_norm": 1.0176743268966675,
+      "learning_rate": 2e-05,
+      "loss": 0.03656572,
+      "step": 21272
+    },
+    {
+      "epoch": 42.546,
+      "grad_norm": 1.3137876987457275,
+      "learning_rate": 2e-05,
+      "loss": 0.05632983,
+      "step": 21273
+    },
+    {
+      "epoch": 42.548,
+      "grad_norm": 2.6000397205352783,
+      "learning_rate": 2e-05,
+      "loss": 0.04389685,
+      "step": 21274
+    },
+    {
+      "epoch": 42.55,
+      "grad_norm": 2.7255218029022217,
+      "learning_rate": 2e-05,
+      "loss": 0.06559823,
+      "step": 21275
+    },
+    {
+      "epoch": 42.552,
+      "grad_norm": 1.1543296575546265,
+      "learning_rate": 2e-05,
+      "loss": 0.05046273,
+      "step": 21276
+    },
+    {
+      "epoch": 42.554,
+      "grad_norm": 1.017396092414856,
+      "learning_rate": 2e-05,
+      "loss": 0.03115503,
+      "step": 21277
+    },
+    {
+      "epoch": 42.556,
+      "grad_norm": 1.3216530084609985,
+      "learning_rate": 2e-05,
+      "loss": 0.06245044,
+      "step": 21278
+    },
+    {
+      "epoch": 42.558,
+      "grad_norm": 12.367131233215332,
+      "learning_rate": 2e-05,
+      "loss": 0.05980074,
+      "step": 21279
+    },
+    {
+      "epoch": 42.56,
+      "grad_norm": 1.1518374681472778,
+      "learning_rate": 2e-05,
+      "loss": 0.03418801,
+      "step": 21280
+    },
+    {
+      "epoch": 42.562,
+      "grad_norm": 2.031050443649292,
+      "learning_rate": 2e-05,
+      "loss": 0.04848942,
+      "step": 21281
+    },
+    {
+      "epoch": 42.564,
+      "grad_norm": 1.0938751697540283,
+      "learning_rate": 2e-05,
+      "loss": 0.0451759,
+      "step": 21282
+    },
+    {
+      "epoch": 42.566,
+      "grad_norm": 1.1018202304840088,
+      "learning_rate": 2e-05,
+      "loss": 0.04848454,
+      "step": 21283
+    },
+    {
+      "epoch": 42.568,
+      "grad_norm": 1.2324109077453613,
+      "learning_rate": 2e-05,
+      "loss": 0.04794756,
+      "step": 21284
+    },
+    {
+      "epoch": 42.57,
+      "grad_norm": 2.7218592166900635,
+      "learning_rate": 2e-05,
+      "loss": 0.0646372,
+      "step": 21285
+    },
+    {
+      "epoch": 42.572,
+      "grad_norm": 1.2674838304519653,
+      "learning_rate": 2e-05,
+      "loss": 0.04550756,
+      "step": 21286
+    },
+    {
+      "epoch": 42.574,
+      "grad_norm": 1.5812036991119385,
+      "learning_rate": 2e-05,
+      "loss": 0.05703519,
+      "step": 21287
+    },
+    {
+      "epoch": 42.576,
+      "grad_norm": 1.1205166578292847,
+      "learning_rate": 2e-05,
+      "loss": 0.036104,
+      "step": 21288
+    },
+    {
+      "epoch": 42.578,
+      "grad_norm": 1.0537066459655762,
+      "learning_rate": 2e-05,
+      "loss": 0.03754063,
+      "step": 21289
+    },
+    {
+      "epoch": 42.58,
+      "grad_norm": 1.2235515117645264,
+      "learning_rate": 2e-05,
+      "loss": 0.05299861,
+      "step": 21290
+    },
+    {
+      "epoch": 42.582,
+      "grad_norm": 1.3739643096923828,
+      "learning_rate": 2e-05,
+      "loss": 0.0461156,
+      "step": 21291
+    },
+    {
+      "epoch": 42.584,
+      "grad_norm": 1.5320465564727783,
+      "learning_rate": 2e-05,
+      "loss": 0.04836542,
+      "step": 21292
+    },
+    {
+      "epoch": 42.586,
+      "grad_norm": 0.9427246451377869,
+      "learning_rate": 2e-05,
+      "loss": 0.0362893,
+      "step": 21293
+    },
+    {
+      "epoch": 42.588,
+      "grad_norm": 1.3789801597595215,
+      "learning_rate": 2e-05,
+      "loss": 0.06239626,
+      "step": 21294
+    },
+    {
+      "epoch": 42.59,
+      "grad_norm": 1.258785605430603,
+      "learning_rate": 2e-05,
+      "loss": 0.05440217,
+      "step": 21295
+    },
+    {
+      "epoch": 42.592,
+      "grad_norm": 1.2506073713302612,
+      "learning_rate": 2e-05,
+      "loss": 0.04535231,
+      "step": 21296
+    },
+    {
+      "epoch": 42.594,
+      "grad_norm": 1.300620436668396,
+      "learning_rate": 2e-05,
+      "loss": 0.0481655,
+      "step": 21297
+    },
+    {
+      "epoch": 42.596,
+      "grad_norm": 1.1247934103012085,
+      "learning_rate": 2e-05,
+      "loss": 0.03969868,
+      "step": 21298
+    },
+    {
+      "epoch": 42.598,
+      "grad_norm": 1.2077419757843018,
+      "learning_rate": 2e-05,
+      "loss": 0.04599006,
+      "step": 21299
+    },
+    {
+      "epoch": 42.6,
+      "grad_norm": 4.433143615722656,
+      "learning_rate": 2e-05,
+      "loss": 0.05563483,
+      "step": 21300
+    },
+    {
+      "epoch": 42.602,
+      "grad_norm": 1.223757266998291,
+      "learning_rate": 2e-05,
+      "loss": 0.04834705,
+      "step": 21301
+    },
+    {
+      "epoch": 42.604,
+      "grad_norm": 1.001859426498413,
+      "learning_rate": 2e-05,
+      "loss": 0.03891331,
+      "step": 21302
+    },
+    {
+      "epoch": 42.606,
+      "grad_norm": 1.4744184017181396,
+      "learning_rate": 2e-05,
+      "loss": 0.04635298,
+      "step": 21303
+    },
+    {
+      "epoch": 42.608,
+      "grad_norm": 1.0292778015136719,
+      "learning_rate": 2e-05,
+      "loss": 0.03995837,
+      "step": 21304
+    },
+    {
+      "epoch": 42.61,
+      "grad_norm": 1.2195087671279907,
+      "learning_rate": 2e-05,
+      "loss": 0.04556499,
+      "step": 21305
+    },
+    {
+      "epoch": 42.612,
+      "grad_norm": 2.625493049621582,
+      "learning_rate": 2e-05,
+      "loss": 0.0414276,
+      "step": 21306
+    },
+    {
+      "epoch": 42.614,
+      "grad_norm": 1.3295499086380005,
+      "learning_rate": 2e-05,
+      "loss": 0.05958395,
+      "step": 21307
+    },
+    {
+      "epoch": 42.616,
+      "grad_norm": 2.4811103343963623,
+      "learning_rate": 2e-05,
+      "loss": 0.08586189,
+      "step": 21308
+    },
+    {
+      "epoch": 42.618,
+      "grad_norm": 0.9455074667930603,
+      "learning_rate": 2e-05,
+      "loss": 0.02827884,
+      "step": 21309
+    },
+    {
+      "epoch": 42.62,
+      "grad_norm": 1.2621259689331055,
+      "learning_rate": 2e-05,
+      "loss": 0.04555801,
+      "step": 21310
+    },
+    {
+      "epoch": 42.622,
+      "grad_norm": 1.1620298624038696,
+      "learning_rate": 2e-05,
+      "loss": 0.0330307,
+      "step": 21311
+    },
+    {
+      "epoch": 42.624,
+      "grad_norm": 1.165947437286377,
+      "learning_rate": 2e-05,
+      "loss": 0.04790859,
+      "step": 21312
+    },
+    {
+      "epoch": 42.626,
+      "grad_norm": 2.1621484756469727,
+      "learning_rate": 2e-05,
+      "loss": 0.06904516,
+      "step": 21313
+    },
+    {
+      "epoch": 42.628,
+      "grad_norm": 1.2981723546981812,
+      "learning_rate": 2e-05,
+      "loss": 0.04969555,
+      "step": 21314
+    },
+    {
+      "epoch": 42.63,
+      "grad_norm": 1.206282377243042,
+      "learning_rate": 2e-05,
+      "loss": 0.05298881,
+      "step": 21315
+    },
+    {
+      "epoch": 42.632,
+      "grad_norm": 1.1706401109695435,
+      "learning_rate": 2e-05,
+      "loss": 0.04146463,
+      "step": 21316
+    },
+    {
+      "epoch": 42.634,
+      "grad_norm": 1.349630355834961,
+      "learning_rate": 2e-05,
+      "loss": 0.05339145,
+      "step": 21317
+    },
+    {
+      "epoch": 42.636,
+      "grad_norm": 1.3318768739700317,
+      "learning_rate": 2e-05,
+      "loss": 0.04014498,
+      "step": 21318
+    },
+    {
+      "epoch": 42.638,
+      "grad_norm": 1.787174105644226,
+      "learning_rate": 2e-05,
+      "loss": 0.07187255,
+      "step": 21319
+    },
+    {
+      "epoch": 42.64,
+      "grad_norm": 1.0085779428482056,
+      "learning_rate": 2e-05,
+      "loss": 0.03416509,
+      "step": 21320
+    },
+    {
+      "epoch": 42.642,
+      "grad_norm": 1.366184115409851,
+      "learning_rate": 2e-05,
+      "loss": 0.05623956,
+      "step": 21321
+    },
+    {
+      "epoch": 42.644,
+      "grad_norm": 1.2050182819366455,
+      "learning_rate": 2e-05,
+      "loss": 0.04033398,
+      "step": 21322
+    },
+    {
+      "epoch": 42.646,
+      "grad_norm": 1.4368079900741577,
+      "learning_rate": 2e-05,
+      "loss": 0.04438941,
+      "step": 21323
+    },
+    {
+      "epoch": 42.648,
+      "grad_norm": 1.231373906135559,
+      "learning_rate": 2e-05,
+      "loss": 0.04198126,
+      "step": 21324
+    },
+    {
+      "epoch": 42.65,
+      "grad_norm": 3.3663556575775146,
+      "learning_rate": 2e-05,
+      "loss": 0.05418935,
+      "step": 21325
+    },
+    {
+      "epoch": 42.652,
+      "grad_norm": 1.528099536895752,
+      "learning_rate": 2e-05,
+      "loss": 0.06321813,
+      "step": 21326
+    },
+    {
+      "epoch": 42.654,
+      "grad_norm": 1.117147445678711,
+      "learning_rate": 2e-05,
+      "loss": 0.0353119,
+      "step": 21327
+    },
+    {
+      "epoch": 42.656,
+      "grad_norm": 1.3439724445343018,
+      "learning_rate": 2e-05,
+      "loss": 0.05230685,
+      "step": 21328
+    },
+    {
+      "epoch": 42.658,
+      "grad_norm": 1.0160696506500244,
+      "learning_rate": 2e-05,
+      "loss": 0.04300717,
+      "step": 21329
+    },
+    {
+      "epoch": 42.66,
+      "grad_norm": 2.5469577312469482,
+      "learning_rate": 2e-05,
+      "loss": 0.04865529,
+      "step": 21330
+    },
+    {
+      "epoch": 42.662,
+      "grad_norm": 1.6859334707260132,
+      "learning_rate": 2e-05,
+      "loss": 0.04481193,
+      "step": 21331
+    },
+    {
+      "epoch": 42.664,
+      "grad_norm": 1.0845061540603638,
+      "learning_rate": 2e-05,
+      "loss": 0.03940871,
+      "step": 21332
+    },
+    {
+      "epoch": 42.666,
+      "grad_norm": 3.1730074882507324,
+      "learning_rate": 2e-05,
+      "loss": 0.06717253,
+      "step": 21333
+    },
+    {
+      "epoch": 42.668,
+      "grad_norm": 0.7828595638275146,
+      "learning_rate": 2e-05,
+      "loss": 0.02153579,
+      "step": 21334
+    },
+    {
+      "epoch": 42.67,
+      "grad_norm": 1.4767457246780396,
+      "learning_rate": 2e-05,
+      "loss": 0.05914707,
+      "step": 21335
+    },
+    {
+      "epoch": 42.672,
+      "grad_norm": 1.4632948637008667,
+      "learning_rate": 2e-05,
+      "loss": 0.0467616,
+      "step": 21336
+    },
+    {
+      "epoch": 42.674,
+      "grad_norm": 1.245827078819275,
+      "learning_rate": 2e-05,
+      "loss": 0.04130943,
+      "step": 21337
+    },
+    {
+      "epoch": 42.676,
+      "grad_norm": 1.1984535455703735,
+      "learning_rate": 2e-05,
+      "loss": 0.05097179,
+      "step": 21338
+    },
+    {
+      "epoch": 42.678,
+      "grad_norm": 1.018325686454773,
+      "learning_rate": 2e-05,
+      "loss": 0.04212132,
+      "step": 21339
+    },
+    {
+      "epoch": 42.68,
+      "grad_norm": 2.334608316421509,
+      "learning_rate": 2e-05,
+      "loss": 0.0457168,
+      "step": 21340
+    },
+    {
+      "epoch": 42.682,
+      "grad_norm": 1.2345631122589111,
+      "learning_rate": 2e-05,
+      "loss": 0.04911385,
+      "step": 21341
+    },
+    {
+      "epoch": 42.684,
+      "grad_norm": 1.1206731796264648,
+      "learning_rate": 2e-05,
+      "loss": 0.0280387,
+      "step": 21342
+    },
+    {
+      "epoch": 42.686,
+      "grad_norm": 1.4151371717453003,
+      "learning_rate": 2e-05,
+      "loss": 0.06245557,
+      "step": 21343
+    },
+    {
+      "epoch": 42.688,
+      "grad_norm": 1.2239195108413696,
+      "learning_rate": 2e-05,
+      "loss": 0.04748405,
+      "step": 21344
+    },
+    {
+      "epoch": 42.69,
+      "grad_norm": 1.0787338018417358,
+      "learning_rate": 2e-05,
+      "loss": 0.03307759,
+      "step": 21345
+    },
+    {
+      "epoch": 42.692,
+      "grad_norm": 1.2297214269638062,
+      "learning_rate": 2e-05,
+      "loss": 0.05256373,
+      "step": 21346
+    },
+    {
+      "epoch": 42.694,
+      "grad_norm": 1.2633470296859741,
+      "learning_rate": 2e-05,
+      "loss": 0.05449621,
+      "step": 21347
+    },
+    {
+      "epoch": 42.696,
+      "grad_norm": 1.178499460220337,
+      "learning_rate": 2e-05,
+      "loss": 0.06108705,
+      "step": 21348
+    },
+    {
+      "epoch": 42.698,
+      "grad_norm": 1.030000925064087,
+      "learning_rate": 2e-05,
+      "loss": 0.03802224,
+      "step": 21349
+    },
+    {
+      "epoch": 42.7,
+      "grad_norm": 1.2210856676101685,
+      "learning_rate": 2e-05,
+      "loss": 0.04589963,
+      "step": 21350
+    },
+    {
+      "epoch": 42.702,
+      "grad_norm": 2.0504300594329834,
+      "learning_rate": 2e-05,
+      "loss": 0.06983016,
+      "step": 21351
+    },
+    {
+      "epoch": 42.704,
+      "grad_norm": 1.2079181671142578,
+      "learning_rate": 2e-05,
+      "loss": 0.04810051,
+      "step": 21352
+    },
+    {
+      "epoch": 42.706,
+      "grad_norm": 1.1589932441711426,
+      "learning_rate": 2e-05,
+      "loss": 0.04745119,
+      "step": 21353
+    },
+    {
+      "epoch": 42.708,
+      "grad_norm": 4.482708930969238,
+      "learning_rate": 2e-05,
+      "loss": 0.0591661,
+      "step": 21354
+    },
+    {
+      "epoch": 42.71,
+      "grad_norm": 1.2510643005371094,
+      "learning_rate": 2e-05,
+      "loss": 0.05218077,
+      "step": 21355
+    },
+    {
+      "epoch": 42.712,
+      "grad_norm": 1.295291543006897,
+      "learning_rate": 2e-05,
+      "loss": 0.05780672,
+      "step": 21356
+    },
+    {
+      "epoch": 42.714,
+      "grad_norm": 1.1533124446868896,
+      "learning_rate": 2e-05,
+      "loss": 0.03984768,
+      "step": 21357
+    },
+    {
+      "epoch": 42.716,
+      "grad_norm": 3.040389060974121,
+      "learning_rate": 2e-05,
+      "loss": 0.04302094,
+      "step": 21358
+    },
+    {
+      "epoch": 42.718,
+      "grad_norm": 1.2422996759414673,
+      "learning_rate": 2e-05,
+      "loss": 0.03962889,
+      "step": 21359
+    },
+    {
+      "epoch": 42.72,
+      "grad_norm": 1.4480630159378052,
+      "learning_rate": 2e-05,
+      "loss": 0.06389065,
+      "step": 21360
+    },
+    {
+      "epoch": 42.722,
+      "grad_norm": 1.3611098527908325,
+      "learning_rate": 2e-05,
+      "loss": 0.0595374,
+      "step": 21361
+    },
+    {
+      "epoch": 42.724,
+      "grad_norm": 1.2345424890518188,
+      "learning_rate": 2e-05,
+      "loss": 0.03755029,
+      "step": 21362
+    },
+    {
+      "epoch": 42.726,
+      "grad_norm": 1.8066527843475342,
+      "learning_rate": 2e-05,
+      "loss": 0.05974579,
+      "step": 21363
+    },
+    {
+      "epoch": 42.728,
+      "grad_norm": 1.1717407703399658,
+      "learning_rate": 2e-05,
+      "loss": 0.04459912,
+      "step": 21364
+    },
+    {
+      "epoch": 42.73,
+      "grad_norm": 1.143455982208252,
+      "learning_rate": 2e-05,
+      "loss": 0.04196816,
+      "step": 21365
+    },
+    {
+      "epoch": 42.732,
+      "grad_norm": 1.2013553380966187,
+      "learning_rate": 2e-05,
+      "loss": 0.03594541,
+      "step": 21366
+    },
+    {
+      "epoch": 42.734,
+      "grad_norm": 1.1833301782608032,
+      "learning_rate": 2e-05,
+      "loss": 0.04475646,
+      "step": 21367
+    },
+    {
+      "epoch": 42.736,
+      "grad_norm": 1.0608525276184082,
+      "learning_rate": 2e-05,
+      "loss": 0.0312412,
+      "step": 21368
+    },
+    {
+      "epoch": 42.738,
+      "grad_norm": 1.3239516019821167,
+      "learning_rate": 2e-05,
+      "loss": 0.05588631,
+      "step": 21369
+    },
+    {
+      "epoch": 42.74,
+      "grad_norm": 1.2072561979293823,
+      "learning_rate": 2e-05,
+      "loss": 0.0523868,
+      "step": 21370
+    },
+    {
+      "epoch": 42.742,
+      "grad_norm": 2.1657562255859375,
+      "learning_rate": 2e-05,
+      "loss": 0.0605137,
+      "step": 21371
+    },
+    {
+      "epoch": 42.744,
+      "grad_norm": 1.158803105354309,
+      "learning_rate": 2e-05,
+      "loss": 0.04379743,
+      "step": 21372
+    },
+    {
+      "epoch": 42.746,
+      "grad_norm": 1.2670451402664185,
+      "learning_rate": 2e-05,
+      "loss": 0.04156856,
+      "step": 21373
+    },
+    {
+      "epoch": 42.748,
+      "grad_norm": 1.0824308395385742,
+      "learning_rate": 2e-05,
+      "loss": 0.04734112,
+      "step": 21374
+    },
+    {
+      "epoch": 42.75,
+      "grad_norm": 1.4957252740859985,
+      "learning_rate": 2e-05,
+      "loss": 0.04622442,
+      "step": 21375
+    },
+    {
+      "epoch": 42.752,
+      "grad_norm": 1.1113194227218628,
+      "learning_rate": 2e-05,
+      "loss": 0.03722591,
+      "step": 21376
+    },
+    {
+      "epoch": 42.754,
+      "grad_norm": 1.513335108757019,
+      "learning_rate": 2e-05,
+      "loss": 0.05863171,
+      "step": 21377
+    },
+    {
+      "epoch": 42.756,
+      "grad_norm": 1.463580846786499,
+      "learning_rate": 2e-05,
+      "loss": 0.0668181,
+      "step": 21378
+    },
+    {
+      "epoch": 42.758,
+      "grad_norm": 1.4309571981430054,
+      "learning_rate": 2e-05,
+      "loss": 0.03635763,
+      "step": 21379
+    },
+    {
+      "epoch": 42.76,
+      "grad_norm": 1.273354172706604,
+      "learning_rate": 2e-05,
+      "loss": 0.04854331,
+      "step": 21380
+    },
+    {
+      "epoch": 42.762,
+      "grad_norm": 1.6273373365402222,
+      "learning_rate": 2e-05,
+      "loss": 0.05022759,
+      "step": 21381
+    },
+    {
+      "epoch": 42.764,
+      "grad_norm": 1.2511389255523682,
+      "learning_rate": 2e-05,
+      "loss": 0.06143327,
+      "step": 21382
+    },
+    {
+      "epoch": 42.766,
+      "grad_norm": 2.7621679306030273,
+      "learning_rate": 2e-05,
+      "loss": 0.06704356,
+      "step": 21383
+    },
+    {
+      "epoch": 42.768,
+      "grad_norm": 1.5374120473861694,
+      "learning_rate": 2e-05,
+      "loss": 0.04694888,
+      "step": 21384
+    },
+    {
+      "epoch": 42.77,
+      "grad_norm": 1.2802921533584595,
+      "learning_rate": 2e-05,
+      "loss": 0.05271628,
+      "step": 21385
+    },
+    {
+      "epoch": 42.772,
+      "grad_norm": 1.4109792709350586,
+      "learning_rate": 2e-05,
+      "loss": 0.05482592,
+      "step": 21386
+    },
+    {
+      "epoch": 42.774,
+      "grad_norm": 1.5210480690002441,
+      "learning_rate": 2e-05,
+      "loss": 0.05803951,
+      "step": 21387
+    },
+    {
+      "epoch": 42.776,
+      "grad_norm": 1.1467036008834839,
+      "learning_rate": 2e-05,
+      "loss": 0.0471014,
+      "step": 21388
+    },
+    {
+      "epoch": 42.778,
+      "grad_norm": 1.3119735717773438,
+      "learning_rate": 2e-05,
+      "loss": 0.0533366,
+      "step": 21389
+    },
+    {
+      "epoch": 42.78,
+      "grad_norm": 1.1641420125961304,
+      "learning_rate": 2e-05,
+      "loss": 0.03897999,
+      "step": 21390
+    },
+    {
+      "epoch": 42.782,
+      "grad_norm": 1.1525665521621704,
+      "learning_rate": 2e-05,
+      "loss": 0.04090577,
+      "step": 21391
+    },
+    {
+      "epoch": 42.784,
+      "grad_norm": 1.2021316289901733,
+      "learning_rate": 2e-05,
+      "loss": 0.04947101,
+      "step": 21392
+    },
+    {
+      "epoch": 42.786,
+      "grad_norm": 1.2240577936172485,
+      "learning_rate": 2e-05,
+      "loss": 0.05535308,
+      "step": 21393
+    },
+    {
+      "epoch": 42.788,
+      "grad_norm": 1.2875986099243164,
+      "learning_rate": 2e-05,
+      "loss": 0.03962916,
+      "step": 21394
+    },
+    {
+      "epoch": 42.79,
+      "grad_norm": 1.526619553565979,
+      "learning_rate": 2e-05,
+      "loss": 0.03942203,
+      "step": 21395
+    },
+    {
+      "epoch": 42.792,
+      "grad_norm": 1.133887529373169,
+      "learning_rate": 2e-05,
+      "loss": 0.04529195,
+      "step": 21396
+    },
+    {
+      "epoch": 42.794,
+      "grad_norm": 1.195630669593811,
+      "learning_rate": 2e-05,
+      "loss": 0.04189469,
+      "step": 21397
+    },
+    {
+      "epoch": 42.796,
+      "grad_norm": 1.2251441478729248,
+      "learning_rate": 2e-05,
+      "loss": 0.04151393,
+      "step": 21398
+    },
+    {
+      "epoch": 42.798,
+      "grad_norm": 1.0438613891601562,
+      "learning_rate": 2e-05,
+      "loss": 0.03437603,
+      "step": 21399
+    },
+    {
+      "epoch": 42.8,
+      "grad_norm": 1.001349687576294,
+      "learning_rate": 2e-05,
+      "loss": 0.03562608,
+      "step": 21400
+    },
+    {
+      "epoch": 42.802,
+      "grad_norm": 1.156879186630249,
+      "learning_rate": 2e-05,
+      "loss": 0.03687527,
+      "step": 21401
+    },
+    {
+      "epoch": 42.804,
+      "grad_norm": 1.0168277025222778,
+      "learning_rate": 2e-05,
+      "loss": 0.04121634,
+      "step": 21402
+    },
+    {
+      "epoch": 42.806,
+      "grad_norm": 1.6192265748977661,
+      "learning_rate": 2e-05,
+      "loss": 0.05773157,
+      "step": 21403
+    },
+    {
+      "epoch": 42.808,
+      "grad_norm": 1.384158968925476,
+      "learning_rate": 2e-05,
+      "loss": 0.04894597,
+      "step": 21404
+    },
+    {
+      "epoch": 42.81,
+      "grad_norm": 1.0740807056427002,
+      "learning_rate": 2e-05,
+      "loss": 0.03360681,
+      "step": 21405
+    },
+    {
+      "epoch": 42.812,
+      "grad_norm": 1.36252760887146,
+      "learning_rate": 2e-05,
+      "loss": 0.04959939,
+      "step": 21406
+    },
+    {
+      "epoch": 42.814,
+      "grad_norm": 1.1320419311523438,
+      "learning_rate": 2e-05,
+      "loss": 0.03762558,
+      "step": 21407
+    },
+    {
+      "epoch": 42.816,
+      "grad_norm": 1.1921322345733643,
+      "learning_rate": 2e-05,
+      "loss": 0.04960704,
+      "step": 21408
+    },
+    {
+      "epoch": 42.818,
+      "grad_norm": 1.363591194152832,
+      "learning_rate": 2e-05,
+      "loss": 0.04697926,
+      "step": 21409
+    },
+    {
+      "epoch": 42.82,
+      "grad_norm": 1.330572485923767,
+      "learning_rate": 2e-05,
+      "loss": 0.03431287,
+      "step": 21410
+    },
+    {
+      "epoch": 42.822,
+      "grad_norm": 1.108099102973938,
+      "learning_rate": 2e-05,
+      "loss": 0.04299867,
+      "step": 21411
+    },
+    {
+      "epoch": 42.824,
+      "grad_norm": 1.28284752368927,
+      "learning_rate": 2e-05,
+      "loss": 0.05352758,
+      "step": 21412
+    },
+    {
+      "epoch": 42.826,
+      "grad_norm": 1.2229121923446655,
+      "learning_rate": 2e-05,
+      "loss": 0.0591327,
+      "step": 21413
+    },
+    {
+      "epoch": 42.828,
+      "grad_norm": 1.152987003326416,
+      "learning_rate": 2e-05,
+      "loss": 0.03737348,
+      "step": 21414
+    },
+    {
+      "epoch": 42.83,
+      "grad_norm": 1.1505484580993652,
+      "learning_rate": 2e-05,
+      "loss": 0.04128478,
+      "step": 21415
+    },
+    {
+      "epoch": 42.832,
+      "grad_norm": 1.2108104228973389,
+      "learning_rate": 2e-05,
+      "loss": 0.0475995,
+      "step": 21416
+    },
+    {
+      "epoch": 42.834,
+      "grad_norm": 1.315833330154419,
+      "learning_rate": 2e-05,
+      "loss": 0.04457718,
+      "step": 21417
+    },
+    {
+      "epoch": 42.836,
+      "grad_norm": 1.3256268501281738,
+      "learning_rate": 2e-05,
+      "loss": 0.04045897,
+      "step": 21418
+    },
+    {
+      "epoch": 42.838,
+      "grad_norm": 2.0043587684631348,
+      "learning_rate": 2e-05,
+      "loss": 0.05085915,
+      "step": 21419
+    },
+    {
+      "epoch": 42.84,
+      "grad_norm": 1.0381031036376953,
+      "learning_rate": 2e-05,
+      "loss": 0.03765831,
+      "step": 21420
+    },
+    {
+      "epoch": 42.842,
+      "grad_norm": 1.3118317127227783,
+      "learning_rate": 2e-05,
+      "loss": 0.05129515,
+      "step": 21421
+    },
+    {
+      "epoch": 42.844,
+      "grad_norm": 1.2408820390701294,
+      "learning_rate": 2e-05,
+      "loss": 0.05141382,
+      "step": 21422
+    },
+    {
+      "epoch": 42.846,
+      "grad_norm": 1.1298905611038208,
+      "learning_rate": 2e-05,
+      "loss": 0.03599843,
+      "step": 21423
+    },
+    {
+      "epoch": 42.848,
+      "grad_norm": 1.4461491107940674,
+      "learning_rate": 2e-05,
+      "loss": 0.06431194,
+      "step": 21424
+    },
+    {
+      "epoch": 42.85,
+      "grad_norm": 1.4541326761245728,
+      "learning_rate": 2e-05,
+      "loss": 0.0566168,
+      "step": 21425
+    },
+    {
+      "epoch": 42.852,
+      "grad_norm": 1.6744414567947388,
+      "learning_rate": 2e-05,
+      "loss": 0.04111119,
+      "step": 21426
+    },
+    {
+      "epoch": 42.854,
+      "grad_norm": 1.2035828828811646,
+      "learning_rate": 2e-05,
+      "loss": 0.05423025,
+      "step": 21427
+    },
+    {
+      "epoch": 42.856,
+      "grad_norm": 3.0272951126098633,
+      "learning_rate": 2e-05,
+      "loss": 0.05712648,
+      "step": 21428
+    },
+    {
+      "epoch": 42.858,
+      "grad_norm": 0.9234744310379028,
+      "learning_rate": 2e-05,
+      "loss": 0.03638177,
+      "step": 21429
+    },
+    {
+      "epoch": 42.86,
+      "grad_norm": 1.2166481018066406,
+      "learning_rate": 2e-05,
+      "loss": 0.04876867,
+      "step": 21430
+    },
+    {
+      "epoch": 42.862,
+      "grad_norm": 1.454555630683899,
+      "learning_rate": 2e-05,
+      "loss": 0.06422681,
+      "step": 21431
+    },
+    {
+      "epoch": 42.864,
+      "grad_norm": 1.2168675661087036,
+      "learning_rate": 2e-05,
+      "loss": 0.05015201,
+      "step": 21432
+    },
+    {
+      "epoch": 42.866,
+      "grad_norm": 1.659549355506897,
+      "learning_rate": 2e-05,
+      "loss": 0.03892962,
+      "step": 21433
+    },
+    {
+      "epoch": 42.868,
+      "grad_norm": 1.268166422843933,
+      "learning_rate": 2e-05,
+      "loss": 0.04936218,
+      "step": 21434
+    },
+    {
+      "epoch": 42.87,
+      "grad_norm": 1.0668729543685913,
+      "learning_rate": 2e-05,
+      "loss": 0.03783253,
+      "step": 21435
+    },
+    {
+      "epoch": 42.872,
+      "grad_norm": 1.1927694082260132,
+      "learning_rate": 2e-05,
+      "loss": 0.03492489,
+      "step": 21436
+    },
+    {
+      "epoch": 42.874,
+      "grad_norm": 1.2407193183898926,
+      "learning_rate": 2e-05,
+      "loss": 0.04802696,
+      "step": 21437
+    },
+    {
+      "epoch": 42.876,
+      "grad_norm": 1.156215786933899,
+      "learning_rate": 2e-05,
+      "loss": 0.0512966,
+      "step": 21438
+    },
+    {
+      "epoch": 42.878,
+      "grad_norm": 1.2110116481781006,
+      "learning_rate": 2e-05,
+      "loss": 0.03545962,
+      "step": 21439
+    },
+    {
+      "epoch": 42.88,
+      "grad_norm": 1.1005175113677979,
+      "learning_rate": 2e-05,
+      "loss": 0.04559709,
+      "step": 21440
+    },
+    {
+      "epoch": 42.882,
+      "grad_norm": 1.329023003578186,
+      "learning_rate": 2e-05,
+      "loss": 0.0600805,
+      "step": 21441
+    },
+    {
+      "epoch": 42.884,
+      "grad_norm": 1.2809315919876099,
+      "learning_rate": 2e-05,
+      "loss": 0.03948345,
+      "step": 21442
+    },
+    {
+      "epoch": 42.886,
+      "grad_norm": 1.0475729703903198,
+      "learning_rate": 2e-05,
+      "loss": 0.02779887,
+      "step": 21443
+    },
+    {
+      "epoch": 42.888,
+      "grad_norm": 1.2448220252990723,
+      "learning_rate": 2e-05,
+      "loss": 0.05767167,
+      "step": 21444
+    },
+    {
+      "epoch": 42.89,
+      "grad_norm": 1.1443229913711548,
+      "learning_rate": 2e-05,
+      "loss": 0.05177973,
+      "step": 21445
+    },
+    {
+      "epoch": 42.892,
+      "grad_norm": 1.5171314477920532,
+      "learning_rate": 2e-05,
+      "loss": 0.04922928,
+      "step": 21446
+    },
+    {
+      "epoch": 42.894,
+      "grad_norm": 1.217365026473999,
+      "learning_rate": 2e-05,
+      "loss": 0.05303325,
+      "step": 21447
+    },
+    {
+      "epoch": 42.896,
+      "grad_norm": 1.2820359468460083,
+      "learning_rate": 2e-05,
+      "loss": 0.06319527,
+      "step": 21448
+    },
+    {
+      "epoch": 42.898,
+      "grad_norm": 1.3315848112106323,
+      "learning_rate": 2e-05,
+      "loss": 0.05307559,
+      "step": 21449
+    },
+    {
+      "epoch": 42.9,
+      "grad_norm": 1.4399700164794922,
+      "learning_rate": 2e-05,
+      "loss": 0.0532106,
+      "step": 21450
+    },
+    {
+      "epoch": 42.902,
+      "grad_norm": 1.1903473138809204,
+      "learning_rate": 2e-05,
+      "loss": 0.03630826,
+      "step": 21451
+    },
+    {
+      "epoch": 42.904,
+      "grad_norm": 1.3730727434158325,
+      "learning_rate": 2e-05,
+      "loss": 0.04516565,
+      "step": 21452
+    },
+    {
+      "epoch": 42.906,
+      "grad_norm": 1.2712746858596802,
+      "learning_rate": 2e-05,
+      "loss": 0.04588057,
+      "step": 21453
+    },
+    {
+      "epoch": 42.908,
+      "grad_norm": 1.266530990600586,
+      "learning_rate": 2e-05,
+      "loss": 0.05008287,
+      "step": 21454
+    },
+    {
+      "epoch": 42.91,
+      "grad_norm": 1.0627597570419312,
+      "learning_rate": 2e-05,
+      "loss": 0.03308658,
+      "step": 21455
+    },
+    {
+      "epoch": 42.912,
+      "grad_norm": 1.8096920251846313,
+      "learning_rate": 2e-05,
+      "loss": 0.04926098,
+      "step": 21456
+    },
+    {
+      "epoch": 42.914,
+      "grad_norm": 1.3828647136688232,
+      "learning_rate": 2e-05,
+      "loss": 0.05100198,
+      "step": 21457
+    },
+    {
+      "epoch": 42.916,
+      "grad_norm": 1.3538262844085693,
+      "learning_rate": 2e-05,
+      "loss": 0.05305772,
+      "step": 21458
+    },
+    {
+      "epoch": 42.918,
+      "grad_norm": 1.327880620956421,
+      "learning_rate": 2e-05,
+      "loss": 0.05849855,
+      "step": 21459
+    },
+    {
+      "epoch": 42.92,
+      "grad_norm": 2.1985175609588623,
+      "learning_rate": 2e-05,
+      "loss": 0.04045462,
+      "step": 21460
+    },
+    {
+      "epoch": 42.922,
+      "grad_norm": 1.3796802759170532,
+      "learning_rate": 2e-05,
+      "loss": 0.05500578,
+      "step": 21461
+    },
+    {
+      "epoch": 42.924,
+      "grad_norm": 1.4442265033721924,
+      "learning_rate": 2e-05,
+      "loss": 0.06637794,
+      "step": 21462
+    },
+    {
+      "epoch": 42.926,
+      "grad_norm": 1.2436082363128662,
+      "learning_rate": 2e-05,
+      "loss": 0.05103221,
+      "step": 21463
+    },
+    {
+      "epoch": 42.928,
+      "grad_norm": 1.0881454944610596,
+      "learning_rate": 2e-05,
+      "loss": 0.05290727,
+      "step": 21464
+    },
+    {
+      "epoch": 42.93,
+      "grad_norm": 1.2928407192230225,
+      "learning_rate": 2e-05,
+      "loss": 0.06411622,
+      "step": 21465
+    },
+    {
+      "epoch": 42.932,
+      "grad_norm": 3.3758764266967773,
+      "learning_rate": 2e-05,
+      "loss": 0.05194551,
+      "step": 21466
+    },
+    {
+      "epoch": 42.934,
+      "grad_norm": 1.0337151288986206,
+      "learning_rate": 2e-05,
+      "loss": 0.0436775,
+      "step": 21467
+    },
+    {
+      "epoch": 42.936,
+      "grad_norm": 1.7715908288955688,
+      "learning_rate": 2e-05,
+      "loss": 0.05398581,
+      "step": 21468
+    },
+    {
+      "epoch": 42.938,
+      "grad_norm": 1.2396409511566162,
+      "learning_rate": 2e-05,
+      "loss": 0.06144951,
+      "step": 21469
+    },
+    {
+      "epoch": 42.94,
+      "grad_norm": 1.1984061002731323,
+      "learning_rate": 2e-05,
+      "loss": 0.04839371,
+      "step": 21470
+    },
+    {
+      "epoch": 42.942,
+      "grad_norm": 1.130316138267517,
+      "learning_rate": 2e-05,
+      "loss": 0.05355977,
+      "step": 21471
+    },
+    {
+      "epoch": 42.944,
+      "grad_norm": 1.217844009399414,
+      "learning_rate": 2e-05,
+      "loss": 0.04676696,
+      "step": 21472
+    },
+    {
+      "epoch": 42.946,
+      "grad_norm": 1.0561518669128418,
+      "learning_rate": 2e-05,
+      "loss": 0.0412074,
+      "step": 21473
+    },
+    {
+      "epoch": 42.948,
+      "grad_norm": 1.0529133081436157,
+      "learning_rate": 2e-05,
+      "loss": 0.03983943,
+      "step": 21474
+    },
+    {
+      "epoch": 42.95,
+      "grad_norm": 1.1867733001708984,
+      "learning_rate": 2e-05,
+      "loss": 0.05299602,
+      "step": 21475
+    },
+    {
+      "epoch": 42.952,
+      "grad_norm": 1.284942865371704,
+      "learning_rate": 2e-05,
+      "loss": 0.04374795,
+      "step": 21476
+    },
+    {
+      "epoch": 42.954,
+      "grad_norm": 1.425246000289917,
+      "learning_rate": 2e-05,
+      "loss": 0.07214919,
+      "step": 21477
+    },
+    {
+      "epoch": 42.956,
+      "grad_norm": 1.3944209814071655,
+      "learning_rate": 2e-05,
+      "loss": 0.06095017,
+      "step": 21478
+    },
+    {
+      "epoch": 42.958,
+      "grad_norm": 1.2820805311203003,
+      "learning_rate": 2e-05,
+      "loss": 0.039196,
+      "step": 21479
+    },
+    {
+      "epoch": 42.96,
+      "grad_norm": 1.0585222244262695,
+      "learning_rate": 2e-05,
+      "loss": 0.04182017,
+      "step": 21480
+    },
+    {
+      "epoch": 42.962,
+      "grad_norm": 1.012815237045288,
+      "learning_rate": 2e-05,
+      "loss": 0.03696268,
+      "step": 21481
+    },
+    {
+      "epoch": 42.964,
+      "grad_norm": 1.024418592453003,
+      "learning_rate": 2e-05,
+      "loss": 0.03427555,
+      "step": 21482
+    },
+    {
+      "epoch": 42.966,
+      "grad_norm": 1.316025972366333,
+      "learning_rate": 2e-05,
+      "loss": 0.06875788,
+      "step": 21483
+    },
+    {
+      "epoch": 42.968,
+      "grad_norm": 1.1476069688796997,
+      "learning_rate": 2e-05,
+      "loss": 0.05424814,
+      "step": 21484
+    },
+    {
+      "epoch": 42.97,
+      "grad_norm": 1.8176919221878052,
+      "learning_rate": 2e-05,
+      "loss": 0.06173361,
+      "step": 21485
+    },
+    {
+      "epoch": 42.972,
+      "grad_norm": 1.1928600072860718,
+      "learning_rate": 2e-05,
+      "loss": 0.05281177,
+      "step": 21486
+    },
+    {
+      "epoch": 42.974,
+      "grad_norm": 0.9522770047187805,
+      "learning_rate": 2e-05,
+      "loss": 0.03397568,
+      "step": 21487
+    },
+    {
+      "epoch": 42.976,
+      "grad_norm": 1.3486912250518799,
+      "learning_rate": 2e-05,
+      "loss": 0.05722672,
+      "step": 21488
+    },
+    {
+      "epoch": 42.978,
+      "grad_norm": 1.3525702953338623,
+      "learning_rate": 2e-05,
+      "loss": 0.05092058,
+      "step": 21489
+    },
+    {
+      "epoch": 42.98,
+      "grad_norm": 1.5690195560455322,
+      "learning_rate": 2e-05,
+      "loss": 0.0533328,
+      "step": 21490
+    },
+    {
+      "epoch": 42.982,
+      "grad_norm": 1.4370635747909546,
+      "learning_rate": 2e-05,
+      "loss": 0.06423539,
+      "step": 21491
+    },
+    {
+      "epoch": 42.984,
+      "grad_norm": 1.0913952589035034,
+      "learning_rate": 2e-05,
+      "loss": 0.03920151,
+      "step": 21492
+    },
+    {
+      "epoch": 42.986,
+      "grad_norm": 1.426624059677124,
+      "learning_rate": 2e-05,
+      "loss": 0.07058513,
+      "step": 21493
+    },
+    {
+      "epoch": 42.988,
+      "grad_norm": 1.0381181240081787,
+      "learning_rate": 2e-05,
+      "loss": 0.03804436,
+      "step": 21494
+    },
+    {
+      "epoch": 42.99,
+      "grad_norm": 0.9893444776535034,
+      "learning_rate": 2e-05,
+      "loss": 0.03246005,
+      "step": 21495
+    },
+    {
+      "epoch": 42.992,
+      "grad_norm": 1.4209214448928833,
+      "learning_rate": 2e-05,
+      "loss": 0.05637572,
+      "step": 21496
+    },
+    {
+      "epoch": 42.994,
+      "grad_norm": 1.1668082475662231,
+      "learning_rate": 2e-05,
+      "loss": 0.04799629,
+      "step": 21497
+    },
+    {
+      "epoch": 42.996,
+      "grad_norm": 1.103164792060852,
+      "learning_rate": 2e-05,
+      "loss": 0.04127581,
+      "step": 21498
+    },
+    {
+      "epoch": 42.998,
+      "grad_norm": 2.728940725326538,
+      "learning_rate": 2e-05,
+      "loss": 0.05009279,
+      "step": 21499
+    },
+    {
+      "epoch": 43.0,
+      "grad_norm": 1.1477065086364746,
+      "learning_rate": 2e-05,
+      "loss": 0.03524557,
+      "step": 21500
+    },
+    {
+      "epoch": 43.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9860279441117764,
+        "Equal_1": 1.0,
+        "Equal_2": 0.9860279441117764,
+        "Equal_3": 0.9920159680638723,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 0.998003992015968,
+        "LineComparison_3": 0.998003992015968,
+        "Parallel_1": 0.9959919839679359,
+        "Parallel_2": 0.9959919839679359,
+        "Parallel_3": 0.99,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.994,
+        "Perpendicular_3": 0.9018036072144289,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 1.0,
+        "PointLiesOnCircle_3": 0.99,
+        "PointLiesOnLine_1": 0.9939879759519038,
+        "PointLiesOnLine_2": 0.9919839679358717,
+        "PointLiesOnLine_3": 0.9860279441117764
+      },
+      "eval_runtime": 319.7331,
+      "eval_samples_per_second": 32.84,
+      "eval_steps_per_second": 0.657,
+      "step": 21500
+    },
+    {
+      "epoch": 43.002,
+      "grad_norm": 1.0772887468338013,
+      "learning_rate": 2e-05,
+      "loss": 0.04629757,
+      "step": 21501
+    },
+    {
+      "epoch": 43.004,
+      "grad_norm": 1.0623633861541748,
+      "learning_rate": 2e-05,
+      "loss": 0.03715014,
+      "step": 21502
+    },
+    {
+      "epoch": 43.006,
+      "grad_norm": 1.1207759380340576,
+      "learning_rate": 2e-05,
+      "loss": 0.05135489,
+      "step": 21503
+    },
+    {
+      "epoch": 43.008,
+      "grad_norm": 1.6616121530532837,
+      "learning_rate": 2e-05,
+      "loss": 0.04297728,
+      "step": 21504
+    },
+    {
+      "epoch": 43.01,
+      "grad_norm": 1.3650739192962646,
+      "learning_rate": 2e-05,
+      "loss": 0.05856933,
+      "step": 21505
+    },
+    {
+      "epoch": 43.012,
+      "grad_norm": 1.8282989263534546,
+      "learning_rate": 2e-05,
+      "loss": 0.04885706,
+      "step": 21506
+    },
+    {
+      "epoch": 43.014,
+      "grad_norm": 1.361122727394104,
+      "learning_rate": 2e-05,
+      "loss": 0.05703342,
+      "step": 21507
+    },
+    {
+      "epoch": 43.016,
+      "grad_norm": 1.263135313987732,
+      "learning_rate": 2e-05,
+      "loss": 0.04379937,
+      "step": 21508
+    },
+    {
+      "epoch": 43.018,
+      "grad_norm": 1.3780148029327393,
+      "learning_rate": 2e-05,
+      "loss": 0.05682152,
+      "step": 21509
+    },
+    {
+      "epoch": 43.02,
+      "grad_norm": 1.3219780921936035,
+      "learning_rate": 2e-05,
+      "loss": 0.0547793,
+      "step": 21510
+    },
+    {
+      "epoch": 43.022,
+      "grad_norm": 1.2654283046722412,
+      "learning_rate": 2e-05,
+      "loss": 0.047323,
+      "step": 21511
+    },
+    {
+      "epoch": 43.024,
+      "grad_norm": 1.272671103477478,
+      "learning_rate": 2e-05,
+      "loss": 0.04931159,
+      "step": 21512
+    },
+    {
+      "epoch": 43.026,
+      "grad_norm": 1.1386663913726807,
+      "learning_rate": 2e-05,
+      "loss": 0.03507162,
+      "step": 21513
+    },
+    {
+      "epoch": 43.028,
+      "grad_norm": 1.292199730873108,
+      "learning_rate": 2e-05,
+      "loss": 0.04890214,
+      "step": 21514
+    },
+    {
+      "epoch": 43.03,
+      "grad_norm": 1.5617351531982422,
+      "learning_rate": 2e-05,
+      "loss": 0.05029353,
+      "step": 21515
+    },
+    {
+      "epoch": 43.032,
+      "grad_norm": 2.036496639251709,
+      "learning_rate": 2e-05,
+      "loss": 0.07027332,
+      "step": 21516
+    },
+    {
+      "epoch": 43.034,
+      "grad_norm": 0.9926701784133911,
+      "learning_rate": 2e-05,
+      "loss": 0.03244736,
+      "step": 21517
+    },
+    {
+      "epoch": 43.036,
+      "grad_norm": 1.0814783573150635,
+      "learning_rate": 2e-05,
+      "loss": 0.04349055,
+      "step": 21518
+    },
+    {
+      "epoch": 43.038,
+      "grad_norm": 1.1061608791351318,
+      "learning_rate": 2e-05,
+      "loss": 0.04114666,
+      "step": 21519
+    },
+    {
+      "epoch": 43.04,
+      "grad_norm": 1.1178923845291138,
+      "learning_rate": 2e-05,
+      "loss": 0.04589276,
+      "step": 21520
+    },
+    {
+      "epoch": 43.042,
+      "grad_norm": 1.3732867240905762,
+      "learning_rate": 2e-05,
+      "loss": 0.0422243,
+      "step": 21521
+    },
+    {
+      "epoch": 43.044,
+      "grad_norm": 2.6664786338806152,
+      "learning_rate": 2e-05,
+      "loss": 0.06500043,
+      "step": 21522
+    },
+    {
+      "epoch": 43.046,
+      "grad_norm": 1.1350311040878296,
+      "learning_rate": 2e-05,
+      "loss": 0.03851867,
+      "step": 21523
+    },
+    {
+      "epoch": 43.048,
+      "grad_norm": 1.0867100954055786,
+      "learning_rate": 2e-05,
+      "loss": 0.04170727,
+      "step": 21524
+    },
+    {
+      "epoch": 43.05,
+      "grad_norm": 1.3278635740280151,
+      "learning_rate": 2e-05,
+      "loss": 0.05376638,
+      "step": 21525
+    },
+    {
+      "epoch": 43.052,
+      "grad_norm": 1.2718002796173096,
+      "learning_rate": 2e-05,
+      "loss": 0.06238486,
+      "step": 21526
+    },
+    {
+      "epoch": 43.054,
+      "grad_norm": 1.2181652784347534,
+      "learning_rate": 2e-05,
+      "loss": 0.05618663,
+      "step": 21527
+    },
+    {
+      "epoch": 43.056,
+      "grad_norm": 1.2053866386413574,
+      "learning_rate": 2e-05,
+      "loss": 0.05018714,
+      "step": 21528
+    },
+    {
+      "epoch": 43.058,
+      "grad_norm": 1.7344564199447632,
+      "learning_rate": 2e-05,
+      "loss": 0.05830881,
+      "step": 21529
+    },
+    {
+      "epoch": 43.06,
+      "grad_norm": 1.2700293064117432,
+      "learning_rate": 2e-05,
+      "loss": 0.05242781,
+      "step": 21530
+    },
+    {
+      "epoch": 43.062,
+      "grad_norm": 1.3638324737548828,
+      "learning_rate": 2e-05,
+      "loss": 0.05736677,
+      "step": 21531
+    },
+    {
+      "epoch": 43.064,
+      "grad_norm": 1.4970639944076538,
+      "learning_rate": 2e-05,
+      "loss": 0.07066285,
+      "step": 21532
+    },
+    {
+      "epoch": 43.066,
+      "grad_norm": 1.4570978879928589,
+      "learning_rate": 2e-05,
+      "loss": 0.05045886,
+      "step": 21533
+    },
+    {
+      "epoch": 43.068,
+      "grad_norm": 1.0011194944381714,
+      "learning_rate": 2e-05,
+      "loss": 0.04105179,
+      "step": 21534
+    },
+    {
+      "epoch": 43.07,
+      "grad_norm": 1.718483805656433,
+      "learning_rate": 2e-05,
+      "loss": 0.0531066,
+      "step": 21535
+    },
+    {
+      "epoch": 43.072,
+      "grad_norm": 1.9208921194076538,
+      "learning_rate": 2e-05,
+      "loss": 0.07216385,
+      "step": 21536
+    },
+    {
+      "epoch": 43.074,
+      "grad_norm": 1.464358925819397,
+      "learning_rate": 2e-05,
+      "loss": 0.05128159,
+      "step": 21537
+    },
+    {
+      "epoch": 43.076,
+      "grad_norm": 1.617860198020935,
+      "learning_rate": 2e-05,
+      "loss": 0.05280685,
+      "step": 21538
+    },
+    {
+      "epoch": 43.078,
+      "grad_norm": 1.2591980695724487,
+      "learning_rate": 2e-05,
+      "loss": 0.03685856,
+      "step": 21539
+    },
+    {
+      "epoch": 43.08,
+      "grad_norm": 1.155810832977295,
+      "learning_rate": 2e-05,
+      "loss": 0.04095223,
+      "step": 21540
+    },
+    {
+      "epoch": 43.082,
+      "grad_norm": 1.4928919076919556,
+      "learning_rate": 2e-05,
+      "loss": 0.0549767,
+      "step": 21541
+    },
+    {
+      "epoch": 43.084,
+      "grad_norm": 1.7492938041687012,
+      "learning_rate": 2e-05,
+      "loss": 0.06689818,
+      "step": 21542
+    },
+    {
+      "epoch": 43.086,
+      "grad_norm": 1.3056284189224243,
+      "learning_rate": 2e-05,
+      "loss": 0.05678186,
+      "step": 21543
+    },
+    {
+      "epoch": 43.088,
+      "grad_norm": 1.1400055885314941,
+      "learning_rate": 2e-05,
+      "loss": 0.05558588,
+      "step": 21544
+    },
+    {
+      "epoch": 43.09,
+      "grad_norm": 1.2211138010025024,
+      "learning_rate": 2e-05,
+      "loss": 0.0515589,
+      "step": 21545
+    },
+    {
+      "epoch": 43.092,
+      "grad_norm": 1.1707724332809448,
+      "learning_rate": 2e-05,
+      "loss": 0.03701507,
+      "step": 21546
+    },
+    {
+      "epoch": 43.094,
+      "grad_norm": 1.9183954000473022,
+      "learning_rate": 2e-05,
+      "loss": 0.06833592,
+      "step": 21547
+    },
+    {
+      "epoch": 43.096,
+      "grad_norm": 1.1282914876937866,
+      "learning_rate": 2e-05,
+      "loss": 0.04579955,
+      "step": 21548
+    },
+    {
+      "epoch": 43.098,
+      "grad_norm": 1.173416256904602,
+      "learning_rate": 2e-05,
+      "loss": 0.0542321,
+      "step": 21549
+    },
+    {
+      "epoch": 43.1,
+      "grad_norm": 1.2352608442306519,
+      "learning_rate": 2e-05,
+      "loss": 0.05904563,
+      "step": 21550
+    },
+    {
+      "epoch": 43.102,
+      "grad_norm": 1.1427820920944214,
+      "learning_rate": 2e-05,
+      "loss": 0.04212008,
+      "step": 21551
+    },
+    {
+      "epoch": 43.104,
+      "grad_norm": 1.5755908489227295,
+      "learning_rate": 2e-05,
+      "loss": 0.04956923,
+      "step": 21552
+    },
+    {
+      "epoch": 43.106,
+      "grad_norm": 1.5542879104614258,
+      "learning_rate": 2e-05,
+      "loss": 0.06011878,
+      "step": 21553
+    },
+    {
+      "epoch": 43.108,
+      "grad_norm": 1.389430284500122,
+      "learning_rate": 2e-05,
+      "loss": 0.04906828,
+      "step": 21554
+    },
+    {
+      "epoch": 43.11,
+      "grad_norm": 1.4012058973312378,
+      "learning_rate": 2e-05,
+      "loss": 0.06038637,
+      "step": 21555
+    },
+    {
+      "epoch": 43.112,
+      "grad_norm": 3.3362956047058105,
+      "learning_rate": 2e-05,
+      "loss": 0.05289589,
+      "step": 21556
+    },
+    {
+      "epoch": 43.114,
+      "grad_norm": 1.240237832069397,
+      "learning_rate": 2e-05,
+      "loss": 0.04178648,
+      "step": 21557
+    },
+    {
+      "epoch": 43.116,
+      "grad_norm": 1.0733537673950195,
+      "learning_rate": 2e-05,
+      "loss": 0.0356668,
+      "step": 21558
+    },
+    {
+      "epoch": 43.118,
+      "grad_norm": 1.472996711730957,
+      "learning_rate": 2e-05,
+      "loss": 0.04951042,
+      "step": 21559
+    },
+    {
+      "epoch": 43.12,
+      "grad_norm": 1.41386878490448,
+      "learning_rate": 2e-05,
+      "loss": 0.05777404,
+      "step": 21560
+    },
+    {
+      "epoch": 43.122,
+      "grad_norm": 1.279788851737976,
+      "learning_rate": 2e-05,
+      "loss": 0.04153184,
+      "step": 21561
+    },
+    {
+      "epoch": 43.124,
+      "grad_norm": 1.375981092453003,
+      "learning_rate": 2e-05,
+      "loss": 0.05480566,
+      "step": 21562
+    },
+    {
+      "epoch": 43.126,
+      "grad_norm": 1.1689554452896118,
+      "learning_rate": 2e-05,
+      "loss": 0.03462553,
+      "step": 21563
+    },
+    {
+      "epoch": 43.128,
+      "grad_norm": 1.4414396286010742,
+      "learning_rate": 2e-05,
+      "loss": 0.05469351,
+      "step": 21564
+    },
+    {
+      "epoch": 43.13,
+      "grad_norm": 1.1081470251083374,
+      "learning_rate": 2e-05,
+      "loss": 0.0445482,
+      "step": 21565
+    },
+    {
+      "epoch": 43.132,
+      "grad_norm": 1.3130306005477905,
+      "learning_rate": 2e-05,
+      "loss": 0.03380114,
+      "step": 21566
+    },
+    {
+      "epoch": 43.134,
+      "grad_norm": 1.2124284505844116,
+      "learning_rate": 2e-05,
+      "loss": 0.0559901,
+      "step": 21567
+    },
+    {
+      "epoch": 43.136,
+      "grad_norm": 1.755231499671936,
+      "learning_rate": 2e-05,
+      "loss": 0.05166025,
+      "step": 21568
+    },
+    {
+      "epoch": 43.138,
+      "grad_norm": 1.3129682540893555,
+      "learning_rate": 2e-05,
+      "loss": 0.05173158,
+      "step": 21569
+    },
+    {
+      "epoch": 43.14,
+      "grad_norm": 1.3563148975372314,
+      "learning_rate": 2e-05,
+      "loss": 0.05071842,
+      "step": 21570
+    },
+    {
+      "epoch": 43.142,
+      "grad_norm": 1.0146980285644531,
+      "learning_rate": 2e-05,
+      "loss": 0.04583877,
+      "step": 21571
+    },
+    {
+      "epoch": 43.144,
+      "grad_norm": 1.3281503915786743,
+      "learning_rate": 2e-05,
+      "loss": 0.04625765,
+      "step": 21572
+    },
+    {
+      "epoch": 43.146,
+      "grad_norm": 1.9704128503799438,
+      "learning_rate": 2e-05,
+      "loss": 0.04025682,
+      "step": 21573
+    },
+    {
+      "epoch": 43.148,
+      "grad_norm": 1.2888715267181396,
+      "learning_rate": 2e-05,
+      "loss": 0.05024606,
+      "step": 21574
+    },
+    {
+      "epoch": 43.15,
+      "grad_norm": 1.2609823942184448,
+      "learning_rate": 2e-05,
+      "loss": 0.04634705,
+      "step": 21575
+    },
+    {
+      "epoch": 43.152,
+      "grad_norm": 1.2393063306808472,
+      "learning_rate": 2e-05,
+      "loss": 0.05042982,
+      "step": 21576
+    },
+    {
+      "epoch": 43.154,
+      "grad_norm": 1.1345429420471191,
+      "learning_rate": 2e-05,
+      "loss": 0.02828003,
+      "step": 21577
+    },
+    {
+      "epoch": 43.156,
+      "grad_norm": 1.1094892024993896,
+      "learning_rate": 2e-05,
+      "loss": 0.04609562,
+      "step": 21578
+    },
+    {
+      "epoch": 43.158,
+      "grad_norm": 1.19045889377594,
+      "learning_rate": 2e-05,
+      "loss": 0.05094993,
+      "step": 21579
+    },
+    {
+      "epoch": 43.16,
+      "grad_norm": 1.1686400175094604,
+      "learning_rate": 2e-05,
+      "loss": 0.03705244,
+      "step": 21580
+    },
+    {
+      "epoch": 43.162,
+      "grad_norm": 1.0274404287338257,
+      "learning_rate": 2e-05,
+      "loss": 0.03788059,
+      "step": 21581
+    },
+    {
+      "epoch": 43.164,
+      "grad_norm": 0.9964593052864075,
+      "learning_rate": 2e-05,
+      "loss": 0.03574125,
+      "step": 21582
+    },
+    {
+      "epoch": 43.166,
+      "grad_norm": 1.8188575506210327,
+      "learning_rate": 2e-05,
+      "loss": 0.05399974,
+      "step": 21583
+    },
+    {
+      "epoch": 43.168,
+      "grad_norm": 2.3859317302703857,
+      "learning_rate": 2e-05,
+      "loss": 0.04840841,
+      "step": 21584
+    },
+    {
+      "epoch": 43.17,
+      "grad_norm": 1.4382333755493164,
+      "learning_rate": 2e-05,
+      "loss": 0.05152167,
+      "step": 21585
+    },
+    {
+      "epoch": 43.172,
+      "grad_norm": 1.2169090509414673,
+      "learning_rate": 2e-05,
+      "loss": 0.04776974,
+      "step": 21586
+    },
+    {
+      "epoch": 43.174,
+      "grad_norm": 1.3083583116531372,
+      "learning_rate": 2e-05,
+      "loss": 0.05877044,
+      "step": 21587
+    },
+    {
+      "epoch": 43.176,
+      "grad_norm": 1.264892578125,
+      "learning_rate": 2e-05,
+      "loss": 0.04724294,
+      "step": 21588
+    },
+    {
+      "epoch": 43.178,
+      "grad_norm": 1.7333793640136719,
+      "learning_rate": 2e-05,
+      "loss": 0.05662074,
+      "step": 21589
+    },
+    {
+      "epoch": 43.18,
+      "grad_norm": 1.1243252754211426,
+      "learning_rate": 2e-05,
+      "loss": 0.0430276,
+      "step": 21590
+    },
+    {
+      "epoch": 43.182,
+      "grad_norm": 1.2926504611968994,
+      "learning_rate": 2e-05,
+      "loss": 0.04414042,
+      "step": 21591
+    },
+    {
+      "epoch": 43.184,
+      "grad_norm": 1.2931801080703735,
+      "learning_rate": 2e-05,
+      "loss": 0.04339027,
+      "step": 21592
+    },
+    {
+      "epoch": 43.186,
+      "grad_norm": 1.1842244863510132,
+      "learning_rate": 2e-05,
+      "loss": 0.05153984,
+      "step": 21593
+    },
+    {
+      "epoch": 43.188,
+      "grad_norm": 1.176692008972168,
+      "learning_rate": 2e-05,
+      "loss": 0.04030944,
+      "step": 21594
+    },
+    {
+      "epoch": 43.19,
+      "grad_norm": 1.0769304037094116,
+      "learning_rate": 2e-05,
+      "loss": 0.03942397,
+      "step": 21595
+    },
+    {
+      "epoch": 43.192,
+      "grad_norm": 1.4324592351913452,
+      "learning_rate": 2e-05,
+      "loss": 0.04200315,
+      "step": 21596
+    },
+    {
+      "epoch": 43.194,
+      "grad_norm": 1.7846524715423584,
+      "learning_rate": 2e-05,
+      "loss": 0.05106489,
+      "step": 21597
+    },
+    {
+      "epoch": 43.196,
+      "grad_norm": 1.0251342058181763,
+      "learning_rate": 2e-05,
+      "loss": 0.03952106,
+      "step": 21598
+    },
+    {
+      "epoch": 43.198,
+      "grad_norm": 1.1627730131149292,
+      "learning_rate": 2e-05,
+      "loss": 0.02958023,
+      "step": 21599
+    },
+    {
+      "epoch": 43.2,
+      "grad_norm": 1.2952426671981812,
+      "learning_rate": 2e-05,
+      "loss": 0.04737274,
+      "step": 21600
+    },
+    {
+      "epoch": 43.202,
+      "grad_norm": 1.7284358739852905,
+      "learning_rate": 2e-05,
+      "loss": 0.06269697,
+      "step": 21601
+    },
+    {
+      "epoch": 43.204,
+      "grad_norm": 1.3142390251159668,
+      "learning_rate": 2e-05,
+      "loss": 0.0550441,
+      "step": 21602
+    },
+    {
+      "epoch": 43.206,
+      "grad_norm": 2.0693068504333496,
+      "learning_rate": 2e-05,
+      "loss": 0.05673666,
+      "step": 21603
+    },
+    {
+      "epoch": 43.208,
+      "grad_norm": 1.3296486139297485,
+      "learning_rate": 2e-05,
+      "loss": 0.04539488,
+      "step": 21604
+    },
+    {
+      "epoch": 43.21,
+      "grad_norm": 1.2718302011489868,
+      "learning_rate": 2e-05,
+      "loss": 0.04932036,
+      "step": 21605
+    },
+    {
+      "epoch": 43.212,
+      "grad_norm": 1.0521125793457031,
+      "learning_rate": 2e-05,
+      "loss": 0.04189048,
+      "step": 21606
+    },
+    {
+      "epoch": 43.214,
+      "grad_norm": 1.1329081058502197,
+      "learning_rate": 2e-05,
+      "loss": 0.04553405,
+      "step": 21607
+    },
+    {
+      "epoch": 43.216,
+      "grad_norm": 1.156334400177002,
+      "learning_rate": 2e-05,
+      "loss": 0.05069153,
+      "step": 21608
+    },
+    {
+      "epoch": 43.218,
+      "grad_norm": 1.0390546321868896,
+      "learning_rate": 2e-05,
+      "loss": 0.04528834,
+      "step": 21609
+    },
+    {
+      "epoch": 43.22,
+      "grad_norm": 0.9134246110916138,
+      "learning_rate": 2e-05,
+      "loss": 0.02971689,
+      "step": 21610
+    },
+    {
+      "epoch": 43.222,
+      "grad_norm": 1.1092662811279297,
+      "learning_rate": 2e-05,
+      "loss": 0.03292485,
+      "step": 21611
+    },
+    {
+      "epoch": 43.224,
+      "grad_norm": 2.882272958755493,
+      "learning_rate": 2e-05,
+      "loss": 0.05360059,
+      "step": 21612
+    },
+    {
+      "epoch": 43.226,
+      "grad_norm": 1.372489094734192,
+      "learning_rate": 2e-05,
+      "loss": 0.04475276,
+      "step": 21613
+    },
+    {
+      "epoch": 43.228,
+      "grad_norm": 1.0251941680908203,
+      "learning_rate": 2e-05,
+      "loss": 0.04171642,
+      "step": 21614
+    },
+    {
+      "epoch": 43.23,
+      "grad_norm": 1.3642363548278809,
+      "learning_rate": 2e-05,
+      "loss": 0.05339691,
+      "step": 21615
+    },
+    {
+      "epoch": 43.232,
+      "grad_norm": 0.9985992908477783,
+      "learning_rate": 2e-05,
+      "loss": 0.04003511,
+      "step": 21616
+    },
+    {
+      "epoch": 43.234,
+      "grad_norm": 1.0361459255218506,
+      "learning_rate": 2e-05,
+      "loss": 0.04068734,
+      "step": 21617
+    },
+    {
+      "epoch": 43.236,
+      "grad_norm": 1.290665864944458,
+      "learning_rate": 2e-05,
+      "loss": 0.05313302,
+      "step": 21618
+    },
+    {
+      "epoch": 43.238,
+      "grad_norm": 1.1882466077804565,
+      "learning_rate": 2e-05,
+      "loss": 0.04230958,
+      "step": 21619
+    },
+    {
+      "epoch": 43.24,
+      "grad_norm": 1.1811344623565674,
+      "learning_rate": 2e-05,
+      "loss": 0.04467118,
+      "step": 21620
+    },
+    {
+      "epoch": 43.242,
+      "grad_norm": 1.3368701934814453,
+      "learning_rate": 2e-05,
+      "loss": 0.04908449,
+      "step": 21621
+    },
+    {
+      "epoch": 43.244,
+      "grad_norm": 1.071096658706665,
+      "learning_rate": 2e-05,
+      "loss": 0.04370485,
+      "step": 21622
+    },
+    {
+      "epoch": 43.246,
+      "grad_norm": 1.1309154033660889,
+      "learning_rate": 2e-05,
+      "loss": 0.04717417,
+      "step": 21623
+    },
+    {
+      "epoch": 43.248,
+      "grad_norm": 2.024557590484619,
+      "learning_rate": 2e-05,
+      "loss": 0.04132883,
+      "step": 21624
+    },
+    {
+      "epoch": 43.25,
+      "grad_norm": 1.165766954421997,
+      "learning_rate": 2e-05,
+      "loss": 0.0444683,
+      "step": 21625
+    },
+    {
+      "epoch": 43.252,
+      "grad_norm": 1.0271003246307373,
+      "learning_rate": 2e-05,
+      "loss": 0.04147922,
+      "step": 21626
+    },
+    {
+      "epoch": 43.254,
+      "grad_norm": 1.3019037246704102,
+      "learning_rate": 2e-05,
+      "loss": 0.04356762,
+      "step": 21627
+    },
+    {
+      "epoch": 43.256,
+      "grad_norm": 1.1984200477600098,
+      "learning_rate": 2e-05,
+      "loss": 0.04804221,
+      "step": 21628
+    },
+    {
+      "epoch": 43.258,
+      "grad_norm": 0.9510525465011597,
+      "learning_rate": 2e-05,
+      "loss": 0.03916862,
+      "step": 21629
+    },
+    {
+      "epoch": 43.26,
+      "grad_norm": 1.4483345746994019,
+      "learning_rate": 2e-05,
+      "loss": 0.05215236,
+      "step": 21630
+    },
+    {
+      "epoch": 43.262,
+      "grad_norm": 1.4785760641098022,
+      "learning_rate": 2e-05,
+      "loss": 0.05577825,
+      "step": 21631
+    },
+    {
+      "epoch": 43.264,
+      "grad_norm": 2.0617661476135254,
+      "learning_rate": 2e-05,
+      "loss": 0.05145205,
+      "step": 21632
+    },
+    {
+      "epoch": 43.266,
+      "grad_norm": 1.1790543794631958,
+      "learning_rate": 2e-05,
+      "loss": 0.04180574,
+      "step": 21633
+    },
+    {
+      "epoch": 43.268,
+      "grad_norm": 1.2387884855270386,
+      "learning_rate": 2e-05,
+      "loss": 0.04779726,
+      "step": 21634
+    },
+    {
+      "epoch": 43.27,
+      "grad_norm": 1.2821975946426392,
+      "learning_rate": 2e-05,
+      "loss": 0.04490368,
+      "step": 21635
+    },
+    {
+      "epoch": 43.272,
+      "grad_norm": 1.1714630126953125,
+      "learning_rate": 2e-05,
+      "loss": 0.04618052,
+      "step": 21636
+    },
+    {
+      "epoch": 43.274,
+      "grad_norm": 1.2499911785125732,
+      "learning_rate": 2e-05,
+      "loss": 0.04752894,
+      "step": 21637
+    },
+    {
+      "epoch": 43.276,
+      "grad_norm": 1.9510254859924316,
+      "learning_rate": 2e-05,
+      "loss": 0.0413634,
+      "step": 21638
+    },
+    {
+      "epoch": 43.278,
+      "grad_norm": 0.9319188594818115,
+      "learning_rate": 2e-05,
+      "loss": 0.03260102,
+      "step": 21639
+    },
+    {
+      "epoch": 43.28,
+      "grad_norm": 1.2887738943099976,
+      "learning_rate": 2e-05,
+      "loss": 0.05729091,
+      "step": 21640
+    },
+    {
+      "epoch": 43.282,
+      "grad_norm": 0.9247579574584961,
+      "learning_rate": 2e-05,
+      "loss": 0.03868435,
+      "step": 21641
+    },
+    {
+      "epoch": 43.284,
+      "grad_norm": 1.2112103700637817,
+      "learning_rate": 2e-05,
+      "loss": 0.04973643,
+      "step": 21642
+    },
+    {
+      "epoch": 43.286,
+      "grad_norm": 1.4075590372085571,
+      "learning_rate": 2e-05,
+      "loss": 0.0373517,
+      "step": 21643
+    },
+    {
+      "epoch": 43.288,
+      "grad_norm": 1.0551819801330566,
+      "learning_rate": 2e-05,
+      "loss": 0.04350845,
+      "step": 21644
+    },
+    {
+      "epoch": 43.29,
+      "grad_norm": 1.4103049039840698,
+      "learning_rate": 2e-05,
+      "loss": 0.0423544,
+      "step": 21645
+    },
+    {
+      "epoch": 43.292,
+      "grad_norm": 1.1112929582595825,
+      "learning_rate": 2e-05,
+      "loss": 0.03513883,
+      "step": 21646
+    },
+    {
+      "epoch": 43.294,
+      "grad_norm": 0.9542357325553894,
+      "learning_rate": 2e-05,
+      "loss": 0.02830679,
+      "step": 21647
+    },
+    {
+      "epoch": 43.296,
+      "grad_norm": 1.5825624465942383,
+      "learning_rate": 2e-05,
+      "loss": 0.04779094,
+      "step": 21648
+    },
+    {
+      "epoch": 43.298,
+      "grad_norm": 1.1777398586273193,
+      "learning_rate": 2e-05,
+      "loss": 0.0480368,
+      "step": 21649
+    },
+    {
+      "epoch": 43.3,
+      "grad_norm": 1.8794219493865967,
+      "learning_rate": 2e-05,
+      "loss": 0.05965583,
+      "step": 21650
+    },
+    {
+      "epoch": 43.302,
+      "grad_norm": 1.262937307357788,
+      "learning_rate": 2e-05,
+      "loss": 0.06756496,
+      "step": 21651
+    },
+    {
+      "epoch": 43.304,
+      "grad_norm": 2.5823323726654053,
+      "learning_rate": 2e-05,
+      "loss": 0.08038095,
+      "step": 21652
+    },
+    {
+      "epoch": 43.306,
+      "grad_norm": 1.1853344440460205,
+      "learning_rate": 2e-05,
+      "loss": 0.03952831,
+      "step": 21653
+    },
+    {
+      "epoch": 43.308,
+      "grad_norm": 1.780608892440796,
+      "learning_rate": 2e-05,
+      "loss": 0.0625229,
+      "step": 21654
+    },
+    {
+      "epoch": 43.31,
+      "grad_norm": 1.1544749736785889,
+      "learning_rate": 2e-05,
+      "loss": 0.04620733,
+      "step": 21655
+    },
+    {
+      "epoch": 43.312,
+      "grad_norm": 1.3556686639785767,
+      "learning_rate": 2e-05,
+      "loss": 0.04416988,
+      "step": 21656
+    },
+    {
+      "epoch": 43.314,
+      "grad_norm": 1.0415023565292358,
+      "learning_rate": 2e-05,
+      "loss": 0.04206276,
+      "step": 21657
+    },
+    {
+      "epoch": 43.316,
+      "grad_norm": 1.2264755964279175,
+      "learning_rate": 2e-05,
+      "loss": 0.0508853,
+      "step": 21658
+    },
+    {
+      "epoch": 43.318,
+      "grad_norm": 1.3368821144104004,
+      "learning_rate": 2e-05,
+      "loss": 0.05324509,
+      "step": 21659
+    },
+    {
+      "epoch": 43.32,
+      "grad_norm": 1.3012473583221436,
+      "learning_rate": 2e-05,
+      "loss": 0.05128824,
+      "step": 21660
+    },
+    {
+      "epoch": 43.322,
+      "grad_norm": 0.9434711933135986,
+      "learning_rate": 2e-05,
+      "loss": 0.0308494,
+      "step": 21661
+    },
+    {
+      "epoch": 43.324,
+      "grad_norm": 1.6305221319198608,
+      "learning_rate": 2e-05,
+      "loss": 0.05408993,
+      "step": 21662
+    },
+    {
+      "epoch": 43.326,
+      "grad_norm": 1.6955217123031616,
+      "learning_rate": 2e-05,
+      "loss": 0.07358453,
+      "step": 21663
+    },
+    {
+      "epoch": 43.328,
+      "grad_norm": 1.4787384271621704,
+      "learning_rate": 2e-05,
+      "loss": 0.05799458,
+      "step": 21664
+    },
+    {
+      "epoch": 43.33,
+      "grad_norm": 1.2819902896881104,
+      "learning_rate": 2e-05,
+      "loss": 0.040464,
+      "step": 21665
+    },
+    {
+      "epoch": 43.332,
+      "grad_norm": 1.4899117946624756,
+      "learning_rate": 2e-05,
+      "loss": 0.05940196,
+      "step": 21666
+    },
+    {
+      "epoch": 43.334,
+      "grad_norm": 1.644298791885376,
+      "learning_rate": 2e-05,
+      "loss": 0.06026432,
+      "step": 21667
+    },
+    {
+      "epoch": 43.336,
+      "grad_norm": 1.2563204765319824,
+      "learning_rate": 2e-05,
+      "loss": 0.05732097,
+      "step": 21668
+    },
+    {
+      "epoch": 43.338,
+      "grad_norm": 1.1418761014938354,
+      "learning_rate": 2e-05,
+      "loss": 0.04549374,
+      "step": 21669
+    },
+    {
+      "epoch": 43.34,
+      "grad_norm": 1.0015463829040527,
+      "learning_rate": 2e-05,
+      "loss": 0.03093121,
+      "step": 21670
+    },
+    {
+      "epoch": 43.342,
+      "grad_norm": 1.0512152910232544,
+      "learning_rate": 2e-05,
+      "loss": 0.04237778,
+      "step": 21671
+    },
+    {
+      "epoch": 43.344,
+      "grad_norm": 1.1289246082305908,
+      "learning_rate": 2e-05,
+      "loss": 0.04607947,
+      "step": 21672
+    },
+    {
+      "epoch": 43.346,
+      "grad_norm": 1.371304988861084,
+      "learning_rate": 2e-05,
+      "loss": 0.05412474,
+      "step": 21673
+    },
+    {
+      "epoch": 43.348,
+      "grad_norm": 1.2538726329803467,
+      "learning_rate": 2e-05,
+      "loss": 0.0535718,
+      "step": 21674
+    },
+    {
+      "epoch": 43.35,
+      "grad_norm": 1.3319662809371948,
+      "learning_rate": 2e-05,
+      "loss": 0.04659107,
+      "step": 21675
+    },
+    {
+      "epoch": 43.352,
+      "grad_norm": 1.1024374961853027,
+      "learning_rate": 2e-05,
+      "loss": 0.03754035,
+      "step": 21676
+    },
+    {
+      "epoch": 43.354,
+      "grad_norm": 1.2166569232940674,
+      "learning_rate": 2e-05,
+      "loss": 0.06684306,
+      "step": 21677
+    },
+    {
+      "epoch": 43.356,
+      "grad_norm": 2.1136088371276855,
+      "learning_rate": 2e-05,
+      "loss": 0.04331131,
+      "step": 21678
+    },
+    {
+      "epoch": 43.358,
+      "grad_norm": 1.2272017002105713,
+      "learning_rate": 2e-05,
+      "loss": 0.04815719,
+      "step": 21679
+    },
+    {
+      "epoch": 43.36,
+      "grad_norm": 1.0744694471359253,
+      "learning_rate": 2e-05,
+      "loss": 0.04111854,
+      "step": 21680
+    },
+    {
+      "epoch": 43.362,
+      "grad_norm": 1.242724061012268,
+      "learning_rate": 2e-05,
+      "loss": 0.06325316,
+      "step": 21681
+    },
+    {
+      "epoch": 43.364,
+      "grad_norm": 1.985144853591919,
+      "learning_rate": 2e-05,
+      "loss": 0.04359365,
+      "step": 21682
+    },
+    {
+      "epoch": 43.366,
+      "grad_norm": 1.1064767837524414,
+      "learning_rate": 2e-05,
+      "loss": 0.05320355,
+      "step": 21683
+    },
+    {
+      "epoch": 43.368,
+      "grad_norm": 1.189843773841858,
+      "learning_rate": 2e-05,
+      "loss": 0.04457613,
+      "step": 21684
+    },
+    {
+      "epoch": 43.37,
+      "grad_norm": 1.203808069229126,
+      "learning_rate": 2e-05,
+      "loss": 0.05187619,
+      "step": 21685
+    },
+    {
+      "epoch": 43.372,
+      "grad_norm": 1.1101083755493164,
+      "learning_rate": 2e-05,
+      "loss": 0.04831094,
+      "step": 21686
+    },
+    {
+      "epoch": 43.374,
+      "grad_norm": 1.428242564201355,
+      "learning_rate": 2e-05,
+      "loss": 0.06388721,
+      "step": 21687
+    },
+    {
+      "epoch": 43.376,
+      "grad_norm": 1.2503690719604492,
+      "learning_rate": 2e-05,
+      "loss": 0.04847308,
+      "step": 21688
+    },
+    {
+      "epoch": 43.378,
+      "grad_norm": 1.1128005981445312,
+      "learning_rate": 2e-05,
+      "loss": 0.03843355,
+      "step": 21689
+    },
+    {
+      "epoch": 43.38,
+      "grad_norm": 0.9491602778434753,
+      "learning_rate": 2e-05,
+      "loss": 0.0259631,
+      "step": 21690
+    },
+    {
+      "epoch": 43.382,
+      "grad_norm": 1.233364224433899,
+      "learning_rate": 2e-05,
+      "loss": 0.0568186,
+      "step": 21691
+    },
+    {
+      "epoch": 43.384,
+      "grad_norm": 0.9893776774406433,
+      "learning_rate": 2e-05,
+      "loss": 0.04174522,
+      "step": 21692
+    },
+    {
+      "epoch": 43.386,
+      "grad_norm": 1.187612771987915,
+      "learning_rate": 2e-05,
+      "loss": 0.0390992,
+      "step": 21693
+    },
+    {
+      "epoch": 43.388,
+      "grad_norm": 1.3221521377563477,
+      "learning_rate": 2e-05,
+      "loss": 0.05131374,
+      "step": 21694
+    },
+    {
+      "epoch": 43.39,
+      "grad_norm": 1.2869195938110352,
+      "learning_rate": 2e-05,
+      "loss": 0.05698796,
+      "step": 21695
+    },
+    {
+      "epoch": 43.392,
+      "grad_norm": 1.5506631135940552,
+      "learning_rate": 2e-05,
+      "loss": 0.05239779,
+      "step": 21696
+    },
+    {
+      "epoch": 43.394,
+      "grad_norm": 1.1303107738494873,
+      "learning_rate": 2e-05,
+      "loss": 0.0321939,
+      "step": 21697
+    },
+    {
+      "epoch": 43.396,
+      "grad_norm": 1.2023539543151855,
+      "learning_rate": 2e-05,
+      "loss": 0.04435757,
+      "step": 21698
+    },
+    {
+      "epoch": 43.398,
+      "grad_norm": 1.644532322883606,
+      "learning_rate": 2e-05,
+      "loss": 0.05108094,
+      "step": 21699
+    },
+    {
+      "epoch": 43.4,
+      "grad_norm": 0.9367179274559021,
+      "learning_rate": 2e-05,
+      "loss": 0.03235884,
+      "step": 21700
+    },
+    {
+      "epoch": 43.402,
+      "grad_norm": 1.0146784782409668,
+      "learning_rate": 2e-05,
+      "loss": 0.03669849,
+      "step": 21701
+    },
+    {
+      "epoch": 43.404,
+      "grad_norm": 1.1251667737960815,
+      "learning_rate": 2e-05,
+      "loss": 0.04423268,
+      "step": 21702
+    },
+    {
+      "epoch": 43.406,
+      "grad_norm": 1.0853943824768066,
+      "learning_rate": 2e-05,
+      "loss": 0.04164097,
+      "step": 21703
+    },
+    {
+      "epoch": 43.408,
+      "grad_norm": 1.3860262632369995,
+      "learning_rate": 2e-05,
+      "loss": 0.06020261,
+      "step": 21704
+    },
+    {
+      "epoch": 43.41,
+      "grad_norm": 1.621960163116455,
+      "learning_rate": 2e-05,
+      "loss": 0.05325014,
+      "step": 21705
+    },
+    {
+      "epoch": 43.412,
+      "grad_norm": 1.4392348527908325,
+      "learning_rate": 2e-05,
+      "loss": 0.03889326,
+      "step": 21706
+    },
+    {
+      "epoch": 43.414,
+      "grad_norm": 1.0264551639556885,
+      "learning_rate": 2e-05,
+      "loss": 0.03898916,
+      "step": 21707
+    },
+    {
+      "epoch": 43.416,
+      "grad_norm": 1.2684388160705566,
+      "learning_rate": 2e-05,
+      "loss": 0.05191773,
+      "step": 21708
+    },
+    {
+      "epoch": 43.418,
+      "grad_norm": 1.1081558465957642,
+      "learning_rate": 2e-05,
+      "loss": 0.04196195,
+      "step": 21709
+    },
+    {
+      "epoch": 43.42,
+      "grad_norm": 1.1390728950500488,
+      "learning_rate": 2e-05,
+      "loss": 0.03881846,
+      "step": 21710
+    },
+    {
+      "epoch": 43.422,
+      "grad_norm": 1.3561352491378784,
+      "learning_rate": 2e-05,
+      "loss": 0.05348574,
+      "step": 21711
+    },
+    {
+      "epoch": 43.424,
+      "grad_norm": 1.4882853031158447,
+      "learning_rate": 2e-05,
+      "loss": 0.06980874,
+      "step": 21712
+    },
+    {
+      "epoch": 43.426,
+      "grad_norm": 1.37758469581604,
+      "learning_rate": 2e-05,
+      "loss": 0.06751367,
+      "step": 21713
+    },
+    {
+      "epoch": 43.428,
+      "grad_norm": 1.1444541215896606,
+      "learning_rate": 2e-05,
+      "loss": 0.04994627,
+      "step": 21714
+    },
+    {
+      "epoch": 43.43,
+      "grad_norm": 1.2537963390350342,
+      "learning_rate": 2e-05,
+      "loss": 0.05444419,
+      "step": 21715
+    },
+    {
+      "epoch": 43.432,
+      "grad_norm": 1.0901340246200562,
+      "learning_rate": 2e-05,
+      "loss": 0.05521618,
+      "step": 21716
+    },
+    {
+      "epoch": 43.434,
+      "grad_norm": 1.0437510013580322,
+      "learning_rate": 2e-05,
+      "loss": 0.0424891,
+      "step": 21717
+    },
+    {
+      "epoch": 43.436,
+      "grad_norm": 1.073522925376892,
+      "learning_rate": 2e-05,
+      "loss": 0.03688889,
+      "step": 21718
+    },
+    {
+      "epoch": 43.438,
+      "grad_norm": 1.129959225654602,
+      "learning_rate": 2e-05,
+      "loss": 0.04021513,
+      "step": 21719
+    },
+    {
+      "epoch": 43.44,
+      "grad_norm": 1.2731077671051025,
+      "learning_rate": 2e-05,
+      "loss": 0.0577286,
+      "step": 21720
+    },
+    {
+      "epoch": 43.442,
+      "grad_norm": 1.4466447830200195,
+      "learning_rate": 2e-05,
+      "loss": 0.06440642,
+      "step": 21721
+    },
+    {
+      "epoch": 43.444,
+      "grad_norm": 1.0096185207366943,
+      "learning_rate": 2e-05,
+      "loss": 0.03558685,
+      "step": 21722
+    },
+    {
+      "epoch": 43.446,
+      "grad_norm": 1.1729049682617188,
+      "learning_rate": 2e-05,
+      "loss": 0.05121062,
+      "step": 21723
+    },
+    {
+      "epoch": 43.448,
+      "grad_norm": 1.3237818479537964,
+      "learning_rate": 2e-05,
+      "loss": 0.05145048,
+      "step": 21724
+    },
+    {
+      "epoch": 43.45,
+      "grad_norm": 1.4835050106048584,
+      "learning_rate": 2e-05,
+      "loss": 0.04858468,
+      "step": 21725
+    },
+    {
+      "epoch": 43.452,
+      "grad_norm": 1.2140564918518066,
+      "learning_rate": 2e-05,
+      "loss": 0.04449757,
+      "step": 21726
+    },
+    {
+      "epoch": 43.454,
+      "grad_norm": 1.1771492958068848,
+      "learning_rate": 2e-05,
+      "loss": 0.05757561,
+      "step": 21727
+    },
+    {
+      "epoch": 43.456,
+      "grad_norm": 1.0299850702285767,
+      "learning_rate": 2e-05,
+      "loss": 0.03556947,
+      "step": 21728
+    },
+    {
+      "epoch": 43.458,
+      "grad_norm": 1.5381124019622803,
+      "learning_rate": 2e-05,
+      "loss": 0.06413664,
+      "step": 21729
+    },
+    {
+      "epoch": 43.46,
+      "grad_norm": 2.5248165130615234,
+      "learning_rate": 2e-05,
+      "loss": 0.05302243,
+      "step": 21730
+    },
+    {
+      "epoch": 43.462,
+      "grad_norm": 1.3111073970794678,
+      "learning_rate": 2e-05,
+      "loss": 0.04211747,
+      "step": 21731
+    },
+    {
+      "epoch": 43.464,
+      "grad_norm": 1.2731165885925293,
+      "learning_rate": 2e-05,
+      "loss": 0.05004764,
+      "step": 21732
+    },
+    {
+      "epoch": 43.466,
+      "grad_norm": 0.9703748822212219,
+      "learning_rate": 2e-05,
+      "loss": 0.03877125,
+      "step": 21733
+    },
+    {
+      "epoch": 43.468,
+      "grad_norm": 1.021268367767334,
+      "learning_rate": 2e-05,
+      "loss": 0.03762333,
+      "step": 21734
+    },
+    {
+      "epoch": 43.47,
+      "grad_norm": 1.2705427408218384,
+      "learning_rate": 2e-05,
+      "loss": 0.05988788,
+      "step": 21735
+    },
+    {
+      "epoch": 43.472,
+      "grad_norm": 1.2365317344665527,
+      "learning_rate": 2e-05,
+      "loss": 0.05386936,
+      "step": 21736
+    },
+    {
+      "epoch": 43.474,
+      "grad_norm": 1.3400436639785767,
+      "learning_rate": 2e-05,
+      "loss": 0.04573278,
+      "step": 21737
+    },
+    {
+      "epoch": 43.476,
+      "grad_norm": 1.3688085079193115,
+      "learning_rate": 2e-05,
+      "loss": 0.05334489,
+      "step": 21738
+    },
+    {
+      "epoch": 43.478,
+      "grad_norm": 1.086027979850769,
+      "learning_rate": 2e-05,
+      "loss": 0.0416183,
+      "step": 21739
+    },
+    {
+      "epoch": 43.48,
+      "grad_norm": 1.3055415153503418,
+      "learning_rate": 2e-05,
+      "loss": 0.03842257,
+      "step": 21740
+    },
+    {
+      "epoch": 43.482,
+      "grad_norm": 1.111611247062683,
+      "learning_rate": 2e-05,
+      "loss": 0.0508645,
+      "step": 21741
+    },
+    {
+      "epoch": 43.484,
+      "grad_norm": 1.194685935974121,
+      "learning_rate": 2e-05,
+      "loss": 0.03484659,
+      "step": 21742
+    },
+    {
+      "epoch": 43.486,
+      "grad_norm": 1.132163405418396,
+      "learning_rate": 2e-05,
+      "loss": 0.03633229,
+      "step": 21743
+    },
+    {
+      "epoch": 43.488,
+      "grad_norm": 1.3565982580184937,
+      "learning_rate": 2e-05,
+      "loss": 0.04545782,
+      "step": 21744
+    },
+    {
+      "epoch": 43.49,
+      "grad_norm": 1.4622793197631836,
+      "learning_rate": 2e-05,
+      "loss": 0.05259608,
+      "step": 21745
+    },
+    {
+      "epoch": 43.492,
+      "grad_norm": 1.0451035499572754,
+      "learning_rate": 2e-05,
+      "loss": 0.04075598,
+      "step": 21746
+    },
+    {
+      "epoch": 43.494,
+      "grad_norm": 2.7072818279266357,
+      "learning_rate": 2e-05,
+      "loss": 0.06001053,
+      "step": 21747
+    },
+    {
+      "epoch": 43.496,
+      "grad_norm": 1.031333088874817,
+      "learning_rate": 2e-05,
+      "loss": 0.03389198,
+      "step": 21748
+    },
+    {
+      "epoch": 43.498,
+      "grad_norm": 0.9913927316665649,
+      "learning_rate": 2e-05,
+      "loss": 0.03781866,
+      "step": 21749
+    },
+    {
+      "epoch": 43.5,
+      "grad_norm": 1.303167462348938,
+      "learning_rate": 2e-05,
+      "loss": 0.03760751,
+      "step": 21750
+    },
+    {
+      "epoch": 43.502,
+      "grad_norm": 1.3275164365768433,
+      "learning_rate": 2e-05,
+      "loss": 0.04503237,
+      "step": 21751
+    },
+    {
+      "epoch": 43.504,
+      "grad_norm": 1.729905366897583,
+      "learning_rate": 2e-05,
+      "loss": 0.04633639,
+      "step": 21752
+    },
+    {
+      "epoch": 43.506,
+      "grad_norm": 1.1333032846450806,
+      "learning_rate": 2e-05,
+      "loss": 0.05070849,
+      "step": 21753
+    },
+    {
+      "epoch": 43.508,
+      "grad_norm": 1.3890432119369507,
+      "learning_rate": 2e-05,
+      "loss": 0.0517424,
+      "step": 21754
+    },
+    {
+      "epoch": 43.51,
+      "grad_norm": 1.8249478340148926,
+      "learning_rate": 2e-05,
+      "loss": 0.03821604,
+      "step": 21755
+    },
+    {
+      "epoch": 43.512,
+      "grad_norm": 1.1964874267578125,
+      "learning_rate": 2e-05,
+      "loss": 0.04850964,
+      "step": 21756
+    },
+    {
+      "epoch": 43.514,
+      "grad_norm": 1.322390079498291,
+      "learning_rate": 2e-05,
+      "loss": 0.04571023,
+      "step": 21757
+    },
+    {
+      "epoch": 43.516,
+      "grad_norm": 1.1808900833129883,
+      "learning_rate": 2e-05,
+      "loss": 0.05330661,
+      "step": 21758
+    },
+    {
+      "epoch": 43.518,
+      "grad_norm": 1.255068063735962,
+      "learning_rate": 2e-05,
+      "loss": 0.05100637,
+      "step": 21759
+    },
+    {
+      "epoch": 43.52,
+      "grad_norm": 1.2289645671844482,
+      "learning_rate": 2e-05,
+      "loss": 0.04602298,
+      "step": 21760
+    },
+    {
+      "epoch": 43.522,
+      "grad_norm": 1.1310776472091675,
+      "learning_rate": 2e-05,
+      "loss": 0.05270489,
+      "step": 21761
+    },
+    {
+      "epoch": 43.524,
+      "grad_norm": 1.2839418649673462,
+      "learning_rate": 2e-05,
+      "loss": 0.04927288,
+      "step": 21762
+    },
+    {
+      "epoch": 43.526,
+      "grad_norm": 1.0740174055099487,
+      "learning_rate": 2e-05,
+      "loss": 0.03569305,
+      "step": 21763
+    },
+    {
+      "epoch": 43.528,
+      "grad_norm": 1.2840776443481445,
+      "learning_rate": 2e-05,
+      "loss": 0.0497875,
+      "step": 21764
+    },
+    {
+      "epoch": 43.53,
+      "grad_norm": 1.718205213546753,
+      "learning_rate": 2e-05,
+      "loss": 0.05506313,
+      "step": 21765
+    },
+    {
+      "epoch": 43.532,
+      "grad_norm": 1.0014818906784058,
+      "learning_rate": 2e-05,
+      "loss": 0.03927164,
+      "step": 21766
+    },
+    {
+      "epoch": 43.534,
+      "grad_norm": 1.1391730308532715,
+      "learning_rate": 2e-05,
+      "loss": 0.03087741,
+      "step": 21767
+    },
+    {
+      "epoch": 43.536,
+      "grad_norm": 1.2874438762664795,
+      "learning_rate": 2e-05,
+      "loss": 0.05720982,
+      "step": 21768
+    },
+    {
+      "epoch": 43.538,
+      "grad_norm": 1.5402920246124268,
+      "learning_rate": 2e-05,
+      "loss": 0.05315606,
+      "step": 21769
+    },
+    {
+      "epoch": 43.54,
+      "grad_norm": 1.420612096786499,
+      "learning_rate": 2e-05,
+      "loss": 0.04444424,
+      "step": 21770
+    },
+    {
+      "epoch": 43.542,
+      "grad_norm": 0.9867546558380127,
+      "learning_rate": 2e-05,
+      "loss": 0.03263182,
+      "step": 21771
+    },
+    {
+      "epoch": 43.544,
+      "grad_norm": 1.4081987142562866,
+      "learning_rate": 2e-05,
+      "loss": 0.05344656,
+      "step": 21772
+    },
+    {
+      "epoch": 43.546,
+      "grad_norm": 1.277359962463379,
+      "learning_rate": 2e-05,
+      "loss": 0.05032256,
+      "step": 21773
+    },
+    {
+      "epoch": 43.548,
+      "grad_norm": 1.1443043947219849,
+      "learning_rate": 2e-05,
+      "loss": 0.04011863,
+      "step": 21774
+    },
+    {
+      "epoch": 43.55,
+      "grad_norm": 1.1632673740386963,
+      "learning_rate": 2e-05,
+      "loss": 0.03645647,
+      "step": 21775
+    },
+    {
+      "epoch": 43.552,
+      "grad_norm": 1.0224082469940186,
+      "learning_rate": 2e-05,
+      "loss": 0.02952418,
+      "step": 21776
+    },
+    {
+      "epoch": 43.554,
+      "grad_norm": 1.3022348880767822,
+      "learning_rate": 2e-05,
+      "loss": 0.05518476,
+      "step": 21777
+    },
+    {
+      "epoch": 43.556,
+      "grad_norm": 1.1476200819015503,
+      "learning_rate": 2e-05,
+      "loss": 0.03775834,
+      "step": 21778
+    },
+    {
+      "epoch": 43.558,
+      "grad_norm": 1.0009914636611938,
+      "learning_rate": 2e-05,
+      "loss": 0.03371383,
+      "step": 21779
+    },
+    {
+      "epoch": 43.56,
+      "grad_norm": 1.1276997327804565,
+      "learning_rate": 2e-05,
+      "loss": 0.0438572,
+      "step": 21780
+    },
+    {
+      "epoch": 43.562,
+      "grad_norm": 1.9971312284469604,
+      "learning_rate": 2e-05,
+      "loss": 0.05232399,
+      "step": 21781
+    },
+    {
+      "epoch": 43.564,
+      "grad_norm": 1.3514554500579834,
+      "learning_rate": 2e-05,
+      "loss": 0.0524235,
+      "step": 21782
+    },
+    {
+      "epoch": 43.566,
+      "grad_norm": 1.3473267555236816,
+      "learning_rate": 2e-05,
+      "loss": 0.05278003,
+      "step": 21783
+    },
+    {
+      "epoch": 43.568,
+      "grad_norm": 1.1983340978622437,
+      "learning_rate": 2e-05,
+      "loss": 0.05057519,
+      "step": 21784
+    },
+    {
+      "epoch": 43.57,
+      "grad_norm": 1.5899887084960938,
+      "learning_rate": 2e-05,
+      "loss": 0.03519575,
+      "step": 21785
+    },
+    {
+      "epoch": 43.572,
+      "grad_norm": 1.1169793605804443,
+      "learning_rate": 2e-05,
+      "loss": 0.04725049,
+      "step": 21786
+    },
+    {
+      "epoch": 43.574,
+      "grad_norm": 1.2387906312942505,
+      "learning_rate": 2e-05,
+      "loss": 0.03475452,
+      "step": 21787
+    },
+    {
+      "epoch": 43.576,
+      "grad_norm": 2.203519821166992,
+      "learning_rate": 2e-05,
+      "loss": 0.04944497,
+      "step": 21788
+    },
+    {
+      "epoch": 43.578,
+      "grad_norm": 1.1906425952911377,
+      "learning_rate": 2e-05,
+      "loss": 0.04174117,
+      "step": 21789
+    },
+    {
+      "epoch": 43.58,
+      "grad_norm": 1.4967057704925537,
+      "learning_rate": 2e-05,
+      "loss": 0.05349187,
+      "step": 21790
+    },
+    {
+      "epoch": 43.582,
+      "grad_norm": 1.9748947620391846,
+      "learning_rate": 2e-05,
+      "loss": 0.03838755,
+      "step": 21791
+    },
+    {
+      "epoch": 43.584,
+      "grad_norm": 1.223773717880249,
+      "learning_rate": 2e-05,
+      "loss": 0.03617196,
+      "step": 21792
+    },
+    {
+      "epoch": 43.586,
+      "grad_norm": 1.0227309465408325,
+      "learning_rate": 2e-05,
+      "loss": 0.04097284,
+      "step": 21793
+    },
+    {
+      "epoch": 43.588,
+      "grad_norm": 1.2495092153549194,
+      "learning_rate": 2e-05,
+      "loss": 0.05023031,
+      "step": 21794
+    },
+    {
+      "epoch": 43.59,
+      "grad_norm": 1.443373203277588,
+      "learning_rate": 2e-05,
+      "loss": 0.05545416,
+      "step": 21795
+    },
+    {
+      "epoch": 43.592,
+      "grad_norm": 1.3429638147354126,
+      "learning_rate": 2e-05,
+      "loss": 0.04617237,
+      "step": 21796
+    },
+    {
+      "epoch": 43.594,
+      "grad_norm": 1.17878258228302,
+      "learning_rate": 2e-05,
+      "loss": 0.04778713,
+      "step": 21797
+    },
+    {
+      "epoch": 43.596,
+      "grad_norm": 1.4965510368347168,
+      "learning_rate": 2e-05,
+      "loss": 0.07662305,
+      "step": 21798
+    },
+    {
+      "epoch": 43.598,
+      "grad_norm": 1.0231486558914185,
+      "learning_rate": 2e-05,
+      "loss": 0.03974038,
+      "step": 21799
+    },
+    {
+      "epoch": 43.6,
+      "grad_norm": 1.3388340473175049,
+      "learning_rate": 2e-05,
+      "loss": 0.05395049,
+      "step": 21800
+    },
+    {
+      "epoch": 43.602,
+      "grad_norm": 1.15225088596344,
+      "learning_rate": 2e-05,
+      "loss": 0.05144007,
+      "step": 21801
+    },
+    {
+      "epoch": 43.604,
+      "grad_norm": 1.2700670957565308,
+      "learning_rate": 2e-05,
+      "loss": 0.05085467,
+      "step": 21802
+    },
+    {
+      "epoch": 43.606,
+      "grad_norm": 1.434041976928711,
+      "learning_rate": 2e-05,
+      "loss": 0.05072808,
+      "step": 21803
+    },
+    {
+      "epoch": 43.608,
+      "grad_norm": 1.410075306892395,
+      "learning_rate": 2e-05,
+      "loss": 0.04461277,
+      "step": 21804
+    },
+    {
+      "epoch": 43.61,
+      "grad_norm": 1.0994086265563965,
+      "learning_rate": 2e-05,
+      "loss": 0.03965208,
+      "step": 21805
+    },
+    {
+      "epoch": 43.612,
+      "grad_norm": 1.2478328943252563,
+      "learning_rate": 2e-05,
+      "loss": 0.05292454,
+      "step": 21806
+    },
+    {
+      "epoch": 43.614,
+      "grad_norm": 1.4595047235488892,
+      "learning_rate": 2e-05,
+      "loss": 0.05401939,
+      "step": 21807
+    },
+    {
+      "epoch": 43.616,
+      "grad_norm": 1.2957178354263306,
+      "learning_rate": 2e-05,
+      "loss": 0.05286802,
+      "step": 21808
+    },
+    {
+      "epoch": 43.618,
+      "grad_norm": 1.0460758209228516,
+      "learning_rate": 2e-05,
+      "loss": 0.03856381,
+      "step": 21809
+    },
+    {
+      "epoch": 43.62,
+      "grad_norm": 1.4046540260314941,
+      "learning_rate": 2e-05,
+      "loss": 0.04930651,
+      "step": 21810
+    },
+    {
+      "epoch": 43.622,
+      "grad_norm": 0.9765205383300781,
+      "learning_rate": 2e-05,
+      "loss": 0.03233945,
+      "step": 21811
+    },
+    {
+      "epoch": 43.624,
+      "grad_norm": 1.4802238941192627,
+      "learning_rate": 2e-05,
+      "loss": 0.05282279,
+      "step": 21812
+    },
+    {
+      "epoch": 43.626,
+      "grad_norm": 1.6739566326141357,
+      "learning_rate": 2e-05,
+      "loss": 0.04126877,
+      "step": 21813
+    },
+    {
+      "epoch": 43.628,
+      "grad_norm": 1.0595064163208008,
+      "learning_rate": 2e-05,
+      "loss": 0.04565667,
+      "step": 21814
+    },
+    {
+      "epoch": 43.63,
+      "grad_norm": 1.4827672243118286,
+      "learning_rate": 2e-05,
+      "loss": 0.03800133,
+      "step": 21815
+    },
+    {
+      "epoch": 43.632,
+      "grad_norm": 1.0282008647918701,
+      "learning_rate": 2e-05,
+      "loss": 0.03047697,
+      "step": 21816
+    },
+    {
+      "epoch": 43.634,
+      "grad_norm": 2.9951870441436768,
+      "learning_rate": 2e-05,
+      "loss": 0.03744018,
+      "step": 21817
+    },
+    {
+      "epoch": 43.636,
+      "grad_norm": 1.112250566482544,
+      "learning_rate": 2e-05,
+      "loss": 0.04401822,
+      "step": 21818
+    },
+    {
+      "epoch": 43.638,
+      "grad_norm": 1.140373706817627,
+      "learning_rate": 2e-05,
+      "loss": 0.0461672,
+      "step": 21819
+    },
+    {
+      "epoch": 43.64,
+      "grad_norm": 1.1561338901519775,
+      "learning_rate": 2e-05,
+      "loss": 0.0520758,
+      "step": 21820
+    },
+    {
+      "epoch": 43.642,
+      "grad_norm": 1.1361647844314575,
+      "learning_rate": 2e-05,
+      "loss": 0.04934377,
+      "step": 21821
+    },
+    {
+      "epoch": 43.644,
+      "grad_norm": 0.7384290099143982,
+      "learning_rate": 2e-05,
+      "loss": 0.02234074,
+      "step": 21822
+    },
+    {
+      "epoch": 43.646,
+      "grad_norm": 1.0414352416992188,
+      "learning_rate": 2e-05,
+      "loss": 0.03685948,
+      "step": 21823
+    },
+    {
+      "epoch": 43.648,
+      "grad_norm": 1.0054208040237427,
+      "learning_rate": 2e-05,
+      "loss": 0.03611515,
+      "step": 21824
+    },
+    {
+      "epoch": 43.65,
+      "grad_norm": 1.6434670686721802,
+      "learning_rate": 2e-05,
+      "loss": 0.04983237,
+      "step": 21825
+    },
+    {
+      "epoch": 43.652,
+      "grad_norm": 1.1644682884216309,
+      "learning_rate": 2e-05,
+      "loss": 0.04886763,
+      "step": 21826
+    },
+    {
+      "epoch": 43.654,
+      "grad_norm": 2.3917458057403564,
+      "learning_rate": 2e-05,
+      "loss": 0.07453853,
+      "step": 21827
+    },
+    {
+      "epoch": 43.656,
+      "grad_norm": 1.2234466075897217,
+      "learning_rate": 2e-05,
+      "loss": 0.05833396,
+      "step": 21828
+    },
+    {
+      "epoch": 43.658,
+      "grad_norm": 1.1793771982192993,
+      "learning_rate": 2e-05,
+      "loss": 0.04483267,
+      "step": 21829
+    },
+    {
+      "epoch": 43.66,
+      "grad_norm": 1.5220239162445068,
+      "learning_rate": 2e-05,
+      "loss": 0.06673879,
+      "step": 21830
+    },
+    {
+      "epoch": 43.662,
+      "grad_norm": 1.2569732666015625,
+      "learning_rate": 2e-05,
+      "loss": 0.04199339,
+      "step": 21831
+    },
+    {
+      "epoch": 43.664,
+      "grad_norm": 1.400726079940796,
+      "learning_rate": 2e-05,
+      "loss": 0.04842927,
+      "step": 21832
+    },
+    {
+      "epoch": 43.666,
+      "grad_norm": 1.587497591972351,
+      "learning_rate": 2e-05,
+      "loss": 0.06091474,
+      "step": 21833
+    },
+    {
+      "epoch": 43.668,
+      "grad_norm": 2.467782735824585,
+      "learning_rate": 2e-05,
+      "loss": 0.057216,
+      "step": 21834
+    },
+    {
+      "epoch": 43.67,
+      "grad_norm": 1.1979477405548096,
+      "learning_rate": 2e-05,
+      "loss": 0.05099071,
+      "step": 21835
+    },
+    {
+      "epoch": 43.672,
+      "grad_norm": 1.6900266408920288,
+      "learning_rate": 2e-05,
+      "loss": 0.04967137,
+      "step": 21836
+    },
+    {
+      "epoch": 43.674,
+      "grad_norm": 1.1722626686096191,
+      "learning_rate": 2e-05,
+      "loss": 0.0484486,
+      "step": 21837
+    },
+    {
+      "epoch": 43.676,
+      "grad_norm": 1.4999947547912598,
+      "learning_rate": 2e-05,
+      "loss": 0.06908549,
+      "step": 21838
+    },
+    {
+      "epoch": 43.678,
+      "grad_norm": 1.2642544507980347,
+      "learning_rate": 2e-05,
+      "loss": 0.04293026,
+      "step": 21839
+    },
+    {
+      "epoch": 43.68,
+      "grad_norm": 1.0956248044967651,
+      "learning_rate": 2e-05,
+      "loss": 0.03712745,
+      "step": 21840
+    },
+    {
+      "epoch": 43.682,
+      "grad_norm": 2.061314344406128,
+      "learning_rate": 2e-05,
+      "loss": 0.05671034,
+      "step": 21841
+    },
+    {
+      "epoch": 43.684,
+      "grad_norm": 1.2066631317138672,
+      "learning_rate": 2e-05,
+      "loss": 0.05251576,
+      "step": 21842
+    },
+    {
+      "epoch": 43.686,
+      "grad_norm": 1.1805452108383179,
+      "learning_rate": 2e-05,
+      "loss": 0.05300939,
+      "step": 21843
+    },
+    {
+      "epoch": 43.688,
+      "grad_norm": 1.5431995391845703,
+      "learning_rate": 2e-05,
+      "loss": 0.06416997,
+      "step": 21844
+    },
+    {
+      "epoch": 43.69,
+      "grad_norm": 0.9500445127487183,
+      "learning_rate": 2e-05,
+      "loss": 0.02867823,
+      "step": 21845
+    },
+    {
+      "epoch": 43.692,
+      "grad_norm": 1.0574393272399902,
+      "learning_rate": 2e-05,
+      "loss": 0.03581215,
+      "step": 21846
+    },
+    {
+      "epoch": 43.694,
+      "grad_norm": 1.081818699836731,
+      "learning_rate": 2e-05,
+      "loss": 0.03437539,
+      "step": 21847
+    },
+    {
+      "epoch": 43.696,
+      "grad_norm": 1.1467571258544922,
+      "learning_rate": 2e-05,
+      "loss": 0.04538688,
+      "step": 21848
+    },
+    {
+      "epoch": 43.698,
+      "grad_norm": 1.7992892265319824,
+      "learning_rate": 2e-05,
+      "loss": 0.05461973,
+      "step": 21849
+    },
+    {
+      "epoch": 43.7,
+      "grad_norm": 1.0957931280136108,
+      "learning_rate": 2e-05,
+      "loss": 0.03794325,
+      "step": 21850
+    },
+    {
+      "epoch": 43.702,
+      "grad_norm": 1.1432968378067017,
+      "learning_rate": 2e-05,
+      "loss": 0.04415257,
+      "step": 21851
+    },
+    {
+      "epoch": 43.704,
+      "grad_norm": 1.2000794410705566,
+      "learning_rate": 2e-05,
+      "loss": 0.03692243,
+      "step": 21852
+    },
+    {
+      "epoch": 43.706,
+      "grad_norm": 2.0741629600524902,
+      "learning_rate": 2e-05,
+      "loss": 0.05918703,
+      "step": 21853
+    },
+    {
+      "epoch": 43.708,
+      "grad_norm": 5.487260341644287,
+      "learning_rate": 2e-05,
+      "loss": 0.07302981,
+      "step": 21854
+    },
+    {
+      "epoch": 43.71,
+      "grad_norm": 1.12295401096344,
+      "learning_rate": 2e-05,
+      "loss": 0.03448126,
+      "step": 21855
+    },
+    {
+      "epoch": 43.712,
+      "grad_norm": 1.3459410667419434,
+      "learning_rate": 2e-05,
+      "loss": 0.05415928,
+      "step": 21856
+    },
+    {
+      "epoch": 43.714,
+      "grad_norm": 2.604322910308838,
+      "learning_rate": 2e-05,
+      "loss": 0.0712513,
+      "step": 21857
+    },
+    {
+      "epoch": 43.716,
+      "grad_norm": 1.216117024421692,
+      "learning_rate": 2e-05,
+      "loss": 0.04590901,
+      "step": 21858
+    },
+    {
+      "epoch": 43.718,
+      "grad_norm": 1.3675901889801025,
+      "learning_rate": 2e-05,
+      "loss": 0.04854389,
+      "step": 21859
+    },
+    {
+      "epoch": 43.72,
+      "grad_norm": 2.20420241355896,
+      "learning_rate": 2e-05,
+      "loss": 0.05691512,
+      "step": 21860
+    },
+    {
+      "epoch": 43.722,
+      "grad_norm": 1.0691967010498047,
+      "learning_rate": 2e-05,
+      "loss": 0.04248612,
+      "step": 21861
+    },
+    {
+      "epoch": 43.724,
+      "grad_norm": 1.1441617012023926,
+      "learning_rate": 2e-05,
+      "loss": 0.04443025,
+      "step": 21862
+    },
+    {
+      "epoch": 43.726,
+      "grad_norm": 1.215785264968872,
+      "learning_rate": 2e-05,
+      "loss": 0.04109699,
+      "step": 21863
+    },
+    {
+      "epoch": 43.728,
+      "grad_norm": 1.1096992492675781,
+      "learning_rate": 2e-05,
+      "loss": 0.04342065,
+      "step": 21864
+    },
+    {
+      "epoch": 43.73,
+      "grad_norm": 1.2204557657241821,
+      "learning_rate": 2e-05,
+      "loss": 0.04820478,
+      "step": 21865
+    },
+    {
+      "epoch": 43.732,
+      "grad_norm": 1.8152563571929932,
+      "learning_rate": 2e-05,
+      "loss": 0.04331843,
+      "step": 21866
+    },
+    {
+      "epoch": 43.734,
+      "grad_norm": 1.228575587272644,
+      "learning_rate": 2e-05,
+      "loss": 0.0453392,
+      "step": 21867
+    },
+    {
+      "epoch": 43.736,
+      "grad_norm": 1.2545289993286133,
+      "learning_rate": 2e-05,
+      "loss": 0.05211987,
+      "step": 21868
+    },
+    {
+      "epoch": 43.738,
+      "grad_norm": 1.061532974243164,
+      "learning_rate": 2e-05,
+      "loss": 0.03736347,
+      "step": 21869
+    },
+    {
+      "epoch": 43.74,
+      "grad_norm": 1.4098767042160034,
+      "learning_rate": 2e-05,
+      "loss": 0.04484908,
+      "step": 21870
+    },
+    {
+      "epoch": 43.742,
+      "grad_norm": 1.2383657693862915,
+      "learning_rate": 2e-05,
+      "loss": 0.05023118,
+      "step": 21871
+    },
+    {
+      "epoch": 43.744,
+      "grad_norm": 1.0549107789993286,
+      "learning_rate": 2e-05,
+      "loss": 0.03575772,
+      "step": 21872
+    },
+    {
+      "epoch": 43.746,
+      "grad_norm": 1.266374945640564,
+      "learning_rate": 2e-05,
+      "loss": 0.05691652,
+      "step": 21873
+    },
+    {
+      "epoch": 43.748,
+      "grad_norm": 1.2940677404403687,
+      "learning_rate": 2e-05,
+      "loss": 0.05207343,
+      "step": 21874
+    },
+    {
+      "epoch": 43.75,
+      "grad_norm": 1.380552887916565,
+      "learning_rate": 2e-05,
+      "loss": 0.03950589,
+      "step": 21875
+    },
+    {
+      "epoch": 43.752,
+      "grad_norm": 1.1680132150650024,
+      "learning_rate": 2e-05,
+      "loss": 0.0435213,
+      "step": 21876
+    },
+    {
+      "epoch": 43.754,
+      "grad_norm": 2.5544819831848145,
+      "learning_rate": 2e-05,
+      "loss": 0.0586822,
+      "step": 21877
+    },
+    {
+      "epoch": 43.756,
+      "grad_norm": 1.1925801038742065,
+      "learning_rate": 2e-05,
+      "loss": 0.0523452,
+      "step": 21878
+    },
+    {
+      "epoch": 43.758,
+      "grad_norm": 1.2226167917251587,
+      "learning_rate": 2e-05,
+      "loss": 0.04687249,
+      "step": 21879
+    },
+    {
+      "epoch": 43.76,
+      "grad_norm": 1.0646800994873047,
+      "learning_rate": 2e-05,
+      "loss": 0.04498032,
+      "step": 21880
+    },
+    {
+      "epoch": 43.762,
+      "grad_norm": 1.3412402868270874,
+      "learning_rate": 2e-05,
+      "loss": 0.05164725,
+      "step": 21881
+    },
+    {
+      "epoch": 43.764,
+      "grad_norm": 1.0574461221694946,
+      "learning_rate": 2e-05,
+      "loss": 0.0448842,
+      "step": 21882
+    },
+    {
+      "epoch": 43.766,
+      "grad_norm": 1.0488017797470093,
+      "learning_rate": 2e-05,
+      "loss": 0.02794564,
+      "step": 21883
+    },
+    {
+      "epoch": 43.768,
+      "grad_norm": 1.2221869230270386,
+      "learning_rate": 2e-05,
+      "loss": 0.05013753,
+      "step": 21884
+    },
+    {
+      "epoch": 43.77,
+      "grad_norm": 1.2905614376068115,
+      "learning_rate": 2e-05,
+      "loss": 0.04898647,
+      "step": 21885
+    },
+    {
+      "epoch": 43.772,
+      "grad_norm": 1.3449629545211792,
+      "learning_rate": 2e-05,
+      "loss": 0.05343453,
+      "step": 21886
+    },
+    {
+      "epoch": 43.774,
+      "grad_norm": 1.0985403060913086,
+      "learning_rate": 2e-05,
+      "loss": 0.04567765,
+      "step": 21887
+    },
+    {
+      "epoch": 43.776,
+      "grad_norm": 1.7854421138763428,
+      "learning_rate": 2e-05,
+      "loss": 0.05000356,
+      "step": 21888
+    },
+    {
+      "epoch": 43.778,
+      "grad_norm": 0.9867087602615356,
+      "learning_rate": 2e-05,
+      "loss": 0.03435191,
+      "step": 21889
+    },
+    {
+      "epoch": 43.78,
+      "grad_norm": 1.1816110610961914,
+      "learning_rate": 2e-05,
+      "loss": 0.04493886,
+      "step": 21890
+    },
+    {
+      "epoch": 43.782,
+      "grad_norm": 1.8454886674880981,
+      "learning_rate": 2e-05,
+      "loss": 0.07082605,
+      "step": 21891
+    },
+    {
+      "epoch": 43.784,
+      "grad_norm": 1.6479289531707764,
+      "learning_rate": 2e-05,
+      "loss": 0.05434608,
+      "step": 21892
+    },
+    {
+      "epoch": 43.786,
+      "grad_norm": 1.2260222434997559,
+      "learning_rate": 2e-05,
+      "loss": 0.0397054,
+      "step": 21893
+    },
+    {
+      "epoch": 43.788,
+      "grad_norm": 1.0151551961898804,
+      "learning_rate": 2e-05,
+      "loss": 0.03879777,
+      "step": 21894
+    },
+    {
+      "epoch": 43.79,
+      "grad_norm": 1.233112096786499,
+      "learning_rate": 2e-05,
+      "loss": 0.04530869,
+      "step": 21895
+    },
+    {
+      "epoch": 43.792,
+      "grad_norm": 1.4653815031051636,
+      "learning_rate": 2e-05,
+      "loss": 0.06113261,
+      "step": 21896
+    },
+    {
+      "epoch": 43.794,
+      "grad_norm": 1.832763671875,
+      "learning_rate": 2e-05,
+      "loss": 0.06107483,
+      "step": 21897
+    },
+    {
+      "epoch": 43.796,
+      "grad_norm": 1.1519736051559448,
+      "learning_rate": 2e-05,
+      "loss": 0.04647954,
+      "step": 21898
+    },
+    {
+      "epoch": 43.798,
+      "grad_norm": 1.4427329301834106,
+      "learning_rate": 2e-05,
+      "loss": 0.0451367,
+      "step": 21899
+    },
+    {
+      "epoch": 43.8,
+      "grad_norm": 1.0031224489212036,
+      "learning_rate": 2e-05,
+      "loss": 0.04659247,
+      "step": 21900
+    },
+    {
+      "epoch": 43.802,
+      "grad_norm": 1.3513147830963135,
+      "learning_rate": 2e-05,
+      "loss": 0.03858761,
+      "step": 21901
+    },
+    {
+      "epoch": 43.804,
+      "grad_norm": 1.4595290422439575,
+      "learning_rate": 2e-05,
+      "loss": 0.04975832,
+      "step": 21902
+    },
+    {
+      "epoch": 43.806,
+      "grad_norm": 1.4404231309890747,
+      "learning_rate": 2e-05,
+      "loss": 0.04388293,
+      "step": 21903
+    },
+    {
+      "epoch": 43.808,
+      "grad_norm": 1.1193671226501465,
+      "learning_rate": 2e-05,
+      "loss": 0.0433252,
+      "step": 21904
+    },
+    {
+      "epoch": 43.81,
+      "grad_norm": 1.7537304162979126,
+      "learning_rate": 2e-05,
+      "loss": 0.05322967,
+      "step": 21905
+    },
+    {
+      "epoch": 43.812,
+      "grad_norm": 1.0735844373703003,
+      "learning_rate": 2e-05,
+      "loss": 0.03970418,
+      "step": 21906
+    },
+    {
+      "epoch": 43.814,
+      "grad_norm": 1.533126711845398,
+      "learning_rate": 2e-05,
+      "loss": 0.05987874,
+      "step": 21907
+    },
+    {
+      "epoch": 43.816,
+      "grad_norm": 1.0501418113708496,
+      "learning_rate": 2e-05,
+      "loss": 0.03239897,
+      "step": 21908
+    },
+    {
+      "epoch": 43.818,
+      "grad_norm": 1.326298475265503,
+      "learning_rate": 2e-05,
+      "loss": 0.05782353,
+      "step": 21909
+    },
+    {
+      "epoch": 43.82,
+      "grad_norm": 1.2999595403671265,
+      "learning_rate": 2e-05,
+      "loss": 0.04503709,
+      "step": 21910
+    },
+    {
+      "epoch": 43.822,
+      "grad_norm": 1.109490156173706,
+      "learning_rate": 2e-05,
+      "loss": 0.0462186,
+      "step": 21911
+    },
+    {
+      "epoch": 43.824,
+      "grad_norm": 1.1068756580352783,
+      "learning_rate": 2e-05,
+      "loss": 0.03144252,
+      "step": 21912
+    },
+    {
+      "epoch": 43.826,
+      "grad_norm": 1.1656792163848877,
+      "learning_rate": 2e-05,
+      "loss": 0.04489972,
+      "step": 21913
+    },
+    {
+      "epoch": 43.828,
+      "grad_norm": 1.3886311054229736,
+      "learning_rate": 2e-05,
+      "loss": 0.04568454,
+      "step": 21914
+    },
+    {
+      "epoch": 43.83,
+      "grad_norm": 1.2280189990997314,
+      "learning_rate": 2e-05,
+      "loss": 0.04975533,
+      "step": 21915
+    },
+    {
+      "epoch": 43.832,
+      "grad_norm": 1.2794169187545776,
+      "learning_rate": 2e-05,
+      "loss": 0.04802956,
+      "step": 21916
+    },
+    {
+      "epoch": 43.834,
+      "grad_norm": 1.431793451309204,
+      "learning_rate": 2e-05,
+      "loss": 0.0503118,
+      "step": 21917
+    },
+    {
+      "epoch": 43.836,
+      "grad_norm": 1.1943049430847168,
+      "learning_rate": 2e-05,
+      "loss": 0.05117747,
+      "step": 21918
+    },
+    {
+      "epoch": 43.838,
+      "grad_norm": 1.1528602838516235,
+      "learning_rate": 2e-05,
+      "loss": 0.03906956,
+      "step": 21919
+    },
+    {
+      "epoch": 43.84,
+      "grad_norm": 1.3231438398361206,
+      "learning_rate": 2e-05,
+      "loss": 0.04882731,
+      "step": 21920
+    },
+    {
+      "epoch": 43.842,
+      "grad_norm": 1.6413450241088867,
+      "learning_rate": 2e-05,
+      "loss": 0.04453898,
+      "step": 21921
+    },
+    {
+      "epoch": 43.844,
+      "grad_norm": 1.188307285308838,
+      "learning_rate": 2e-05,
+      "loss": 0.04830351,
+      "step": 21922
+    },
+    {
+      "epoch": 43.846,
+      "grad_norm": 1.5874241590499878,
+      "learning_rate": 2e-05,
+      "loss": 0.05982009,
+      "step": 21923
+    },
+    {
+      "epoch": 43.848,
+      "grad_norm": 0.9717968702316284,
+      "learning_rate": 2e-05,
+      "loss": 0.03369679,
+      "step": 21924
+    },
+    {
+      "epoch": 43.85,
+      "grad_norm": 1.2073335647583008,
+      "learning_rate": 2e-05,
+      "loss": 0.03507089,
+      "step": 21925
+    },
+    {
+      "epoch": 43.852,
+      "grad_norm": 1.133551001548767,
+      "learning_rate": 2e-05,
+      "loss": 0.04241706,
+      "step": 21926
+    },
+    {
+      "epoch": 43.854,
+      "grad_norm": 1.2795426845550537,
+      "learning_rate": 2e-05,
+      "loss": 0.04569482,
+      "step": 21927
+    },
+    {
+      "epoch": 43.856,
+      "grad_norm": 1.216707706451416,
+      "learning_rate": 2e-05,
+      "loss": 0.05699688,
+      "step": 21928
+    },
+    {
+      "epoch": 43.858,
+      "grad_norm": 1.9533652067184448,
+      "learning_rate": 2e-05,
+      "loss": 0.07641211,
+      "step": 21929
+    },
+    {
+      "epoch": 43.86,
+      "grad_norm": 1.0922882556915283,
+      "learning_rate": 2e-05,
+      "loss": 0.02667849,
+      "step": 21930
+    },
+    {
+      "epoch": 43.862,
+      "grad_norm": 1.0828759670257568,
+      "learning_rate": 2e-05,
+      "loss": 0.04916801,
+      "step": 21931
+    },
+    {
+      "epoch": 43.864,
+      "grad_norm": 1.2230288982391357,
+      "learning_rate": 2e-05,
+      "loss": 0.05529765,
+      "step": 21932
+    },
+    {
+      "epoch": 43.866,
+      "grad_norm": 1.1177533864974976,
+      "learning_rate": 2e-05,
+      "loss": 0.03495992,
+      "step": 21933
+    },
+    {
+      "epoch": 43.868,
+      "grad_norm": 1.0803815126419067,
+      "learning_rate": 2e-05,
+      "loss": 0.04231985,
+      "step": 21934
+    },
+    {
+      "epoch": 43.87,
+      "grad_norm": 1.170892357826233,
+      "learning_rate": 2e-05,
+      "loss": 0.04216111,
+      "step": 21935
+    },
+    {
+      "epoch": 43.872,
+      "grad_norm": 1.6169553995132446,
+      "learning_rate": 2e-05,
+      "loss": 0.04968107,
+      "step": 21936
+    },
+    {
+      "epoch": 43.874,
+      "grad_norm": 1.2540957927703857,
+      "learning_rate": 2e-05,
+      "loss": 0.05158932,
+      "step": 21937
+    },
+    {
+      "epoch": 43.876,
+      "grad_norm": 0.9766530394554138,
+      "learning_rate": 2e-05,
+      "loss": 0.02565751,
+      "step": 21938
+    },
+    {
+      "epoch": 43.878,
+      "grad_norm": 1.1864181756973267,
+      "learning_rate": 2e-05,
+      "loss": 0.04187727,
+      "step": 21939
+    },
+    {
+      "epoch": 43.88,
+      "grad_norm": 1.157779335975647,
+      "learning_rate": 2e-05,
+      "loss": 0.04422162,
+      "step": 21940
+    },
+    {
+      "epoch": 43.882,
+      "grad_norm": 3.2718393802642822,
+      "learning_rate": 2e-05,
+      "loss": 0.03786396,
+      "step": 21941
+    },
+    {
+      "epoch": 43.884,
+      "grad_norm": 1.7684119939804077,
+      "learning_rate": 2e-05,
+      "loss": 0.04769549,
+      "step": 21942
+    },
+    {
+      "epoch": 43.886,
+      "grad_norm": 1.5492713451385498,
+      "learning_rate": 2e-05,
+      "loss": 0.06170796,
+      "step": 21943
+    },
+    {
+      "epoch": 43.888,
+      "grad_norm": 1.0928599834442139,
+      "learning_rate": 2e-05,
+      "loss": 0.03716299,
+      "step": 21944
+    },
+    {
+      "epoch": 43.89,
+      "grad_norm": 2.3191888332366943,
+      "learning_rate": 2e-05,
+      "loss": 0.06007298,
+      "step": 21945
+    },
+    {
+      "epoch": 43.892,
+      "grad_norm": 1.7292722463607788,
+      "learning_rate": 2e-05,
+      "loss": 0.04911011,
+      "step": 21946
+    },
+    {
+      "epoch": 43.894,
+      "grad_norm": 1.0400488376617432,
+      "learning_rate": 2e-05,
+      "loss": 0.03599721,
+      "step": 21947
+    },
+    {
+      "epoch": 43.896,
+      "grad_norm": 1.2575953006744385,
+      "learning_rate": 2e-05,
+      "loss": 0.05165,
+      "step": 21948
+    },
+    {
+      "epoch": 43.898,
+      "grad_norm": 1.1358708143234253,
+      "learning_rate": 2e-05,
+      "loss": 0.04962153,
+      "step": 21949
+    },
+    {
+      "epoch": 43.9,
+      "grad_norm": 1.5051313638687134,
+      "learning_rate": 2e-05,
+      "loss": 0.04658557,
+      "step": 21950
+    },
+    {
+      "epoch": 43.902,
+      "grad_norm": 0.9277594685554504,
+      "learning_rate": 2e-05,
+      "loss": 0.03316977,
+      "step": 21951
+    },
+    {
+      "epoch": 43.904,
+      "grad_norm": 1.3635467290878296,
+      "learning_rate": 2e-05,
+      "loss": 0.05430625,
+      "step": 21952
+    },
+    {
+      "epoch": 43.906,
+      "grad_norm": 1.391068458557129,
+      "learning_rate": 2e-05,
+      "loss": 0.06159748,
+      "step": 21953
+    },
+    {
+      "epoch": 43.908,
+      "grad_norm": 1.538338303565979,
+      "learning_rate": 2e-05,
+      "loss": 0.04379988,
+      "step": 21954
+    },
+    {
+      "epoch": 43.91,
+      "grad_norm": 3.397902250289917,
+      "learning_rate": 2e-05,
+      "loss": 0.06411252,
+      "step": 21955
+    },
+    {
+      "epoch": 43.912,
+      "grad_norm": 1.0770752429962158,
+      "learning_rate": 2e-05,
+      "loss": 0.04424539,
+      "step": 21956
+    },
+    {
+      "epoch": 43.914,
+      "grad_norm": 1.3942207098007202,
+      "learning_rate": 2e-05,
+      "loss": 0.04902345,
+      "step": 21957
+    },
+    {
+      "epoch": 43.916,
+      "grad_norm": 1.1789013147354126,
+      "learning_rate": 2e-05,
+      "loss": 0.04915366,
+      "step": 21958
+    },
+    {
+      "epoch": 43.918,
+      "grad_norm": 1.31106698513031,
+      "learning_rate": 2e-05,
+      "loss": 0.03723363,
+      "step": 21959
+    },
+    {
+      "epoch": 43.92,
+      "grad_norm": 1.3015276193618774,
+      "learning_rate": 2e-05,
+      "loss": 0.04564174,
+      "step": 21960
+    },
+    {
+      "epoch": 43.922,
+      "grad_norm": 1.144248127937317,
+      "learning_rate": 2e-05,
+      "loss": 0.0414278,
+      "step": 21961
+    },
+    {
+      "epoch": 43.924,
+      "grad_norm": 1.1458982229232788,
+      "learning_rate": 2e-05,
+      "loss": 0.04920912,
+      "step": 21962
+    },
+    {
+      "epoch": 43.926,
+      "grad_norm": 1.6706799268722534,
+      "learning_rate": 2e-05,
+      "loss": 0.04606791,
+      "step": 21963
+    },
+    {
+      "epoch": 43.928,
+      "grad_norm": 1.1837114095687866,
+      "learning_rate": 2e-05,
+      "loss": 0.04852568,
+      "step": 21964
+    },
+    {
+      "epoch": 43.93,
+      "grad_norm": 1.614950180053711,
+      "learning_rate": 2e-05,
+      "loss": 0.05900152,
+      "step": 21965
+    },
+    {
+      "epoch": 43.932,
+      "grad_norm": 1.168605923652649,
+      "learning_rate": 2e-05,
+      "loss": 0.04590261,
+      "step": 21966
+    },
+    {
+      "epoch": 43.934,
+      "grad_norm": 1.91372811794281,
+      "learning_rate": 2e-05,
+      "loss": 0.05605849,
+      "step": 21967
+    },
+    {
+      "epoch": 43.936,
+      "grad_norm": 1.2598594427108765,
+      "learning_rate": 2e-05,
+      "loss": 0.05659298,
+      "step": 21968
+    },
+    {
+      "epoch": 43.938,
+      "grad_norm": 1.135242223739624,
+      "learning_rate": 2e-05,
+      "loss": 0.03806643,
+      "step": 21969
+    },
+    {
+      "epoch": 43.94,
+      "grad_norm": 1.2540684938430786,
+      "learning_rate": 2e-05,
+      "loss": 0.04728854,
+      "step": 21970
+    },
+    {
+      "epoch": 43.942,
+      "grad_norm": 1.2394566535949707,
+      "learning_rate": 2e-05,
+      "loss": 0.05657537,
+      "step": 21971
+    },
+    {
+      "epoch": 43.944,
+      "grad_norm": 1.1225388050079346,
+      "learning_rate": 2e-05,
+      "loss": 0.05487181,
+      "step": 21972
+    },
+    {
+      "epoch": 43.946,
+      "grad_norm": 1.3079673051834106,
+      "learning_rate": 2e-05,
+      "loss": 0.04643994,
+      "step": 21973
+    },
+    {
+      "epoch": 43.948,
+      "grad_norm": 1.1278849840164185,
+      "learning_rate": 2e-05,
+      "loss": 0.05071158,
+      "step": 21974
+    },
+    {
+      "epoch": 43.95,
+      "grad_norm": 1.311438798904419,
+      "learning_rate": 2e-05,
+      "loss": 0.0565824,
+      "step": 21975
+    },
+    {
+      "epoch": 43.952,
+      "grad_norm": 1.0105366706848145,
+      "learning_rate": 2e-05,
+      "loss": 0.04423777,
+      "step": 21976
+    },
+    {
+      "epoch": 43.954,
+      "grad_norm": 5.449398040771484,
+      "learning_rate": 2e-05,
+      "loss": 0.06142551,
+      "step": 21977
+    },
+    {
+      "epoch": 43.956,
+      "grad_norm": 1.5247281789779663,
+      "learning_rate": 2e-05,
+      "loss": 0.05841498,
+      "step": 21978
+    },
+    {
+      "epoch": 43.958,
+      "grad_norm": 1.2645142078399658,
+      "learning_rate": 2e-05,
+      "loss": 0.05982171,
+      "step": 21979
+    },
+    {
+      "epoch": 43.96,
+      "grad_norm": 1.3923887014389038,
+      "learning_rate": 2e-05,
+      "loss": 0.05867479,
+      "step": 21980
+    },
+    {
+      "epoch": 43.962,
+      "grad_norm": 1.2104207277297974,
+      "learning_rate": 2e-05,
+      "loss": 0.05177038,
+      "step": 21981
+    },
+    {
+      "epoch": 43.964,
+      "grad_norm": 1.2273800373077393,
+      "learning_rate": 2e-05,
+      "loss": 0.03514304,
+      "step": 21982
+    },
+    {
+      "epoch": 43.966,
+      "grad_norm": 1.2372844219207764,
+      "learning_rate": 2e-05,
+      "loss": 0.05396766,
+      "step": 21983
+    },
+    {
+      "epoch": 43.968,
+      "grad_norm": 1.1264572143554688,
+      "learning_rate": 2e-05,
+      "loss": 0.03575965,
+      "step": 21984
+    },
+    {
+      "epoch": 43.97,
+      "grad_norm": 1.2060368061065674,
+      "learning_rate": 2e-05,
+      "loss": 0.05158073,
+      "step": 21985
+    },
+    {
+      "epoch": 43.972,
+      "grad_norm": 1.7094749212265015,
+      "learning_rate": 2e-05,
+      "loss": 0.06007314,
+      "step": 21986
+    },
+    {
+      "epoch": 43.974,
+      "grad_norm": 1.0798571109771729,
+      "learning_rate": 2e-05,
+      "loss": 0.05095109,
+      "step": 21987
+    },
+    {
+      "epoch": 43.976,
+      "grad_norm": 1.1857688426971436,
+      "learning_rate": 2e-05,
+      "loss": 0.04817619,
+      "step": 21988
+    },
+    {
+      "epoch": 43.978,
+      "grad_norm": 1.956149697303772,
+      "learning_rate": 2e-05,
+      "loss": 0.05255031,
+      "step": 21989
+    },
+    {
+      "epoch": 43.98,
+      "grad_norm": 1.298830270767212,
+      "learning_rate": 2e-05,
+      "loss": 0.0518288,
+      "step": 21990
+    },
+    {
+      "epoch": 43.982,
+      "grad_norm": 2.401399612426758,
+      "learning_rate": 2e-05,
+      "loss": 0.0452053,
+      "step": 21991
+    },
+    {
+      "epoch": 43.984,
+      "grad_norm": 1.1325953006744385,
+      "learning_rate": 2e-05,
+      "loss": 0.039924,
+      "step": 21992
+    },
+    {
+      "epoch": 43.986,
+      "grad_norm": 1.7556862831115723,
+      "learning_rate": 2e-05,
+      "loss": 0.052302,
+      "step": 21993
+    },
+    {
+      "epoch": 43.988,
+      "grad_norm": 1.3267320394515991,
+      "learning_rate": 2e-05,
+      "loss": 0.05184503,
+      "step": 21994
+    },
+    {
+      "epoch": 43.99,
+      "grad_norm": 1.1495281457901,
+      "learning_rate": 2e-05,
+      "loss": 0.04765497,
+      "step": 21995
+    },
+    {
+      "epoch": 43.992,
+      "grad_norm": 1.6266371011734009,
+      "learning_rate": 2e-05,
+      "loss": 0.05227334,
+      "step": 21996
+    },
+    {
+      "epoch": 43.994,
+      "grad_norm": 0.914303719997406,
+      "learning_rate": 2e-05,
+      "loss": 0.02800883,
+      "step": 21997
+    },
+    {
+      "epoch": 43.996,
+      "grad_norm": 1.2601372003555298,
+      "learning_rate": 2e-05,
+      "loss": 0.04211642,
+      "step": 21998
+    },
+    {
+      "epoch": 43.998,
+      "grad_norm": 1.1143946647644043,
+      "learning_rate": 2e-05,
+      "loss": 0.0415227,
+      "step": 21999
+    },
+    {
+      "epoch": 44.0,
+      "grad_norm": 1.461881160736084,
+      "learning_rate": 2e-05,
+      "loss": 0.06111758,
+      "step": 22000
+    },
+    {
+      "epoch": 44.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9860279441117764,
+        "Equal_1": 1.0,
+        "Equal_2": 0.9920159680638723,
+        "Equal_3": 0.9900199600798403,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.998003992015968,
+        "Parallel_1": 0.9899799599198397,
+        "Parallel_2": 0.9979959919839679,
+        "Parallel_3": 0.992,
+        "Perpendicular_1": 1.0,
+        "Perpendicular_2": 0.992,
+        "Perpendicular_3": 0.8937875751503006,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 1.0,
+        "PointLiesOnCircle_3": 0.99,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9939879759519038,
+        "PointLiesOnLine_3": 0.9840319361277445
+      },
+      "eval_runtime": 319.306,
+      "eval_samples_per_second": 32.884,
+      "eval_steps_per_second": 0.658,
+      "step": 22000
+    },
+    {
+      "epoch": 44.002,
+      "grad_norm": 1.1951003074645996,
+      "learning_rate": 2e-05,
+      "loss": 0.04387069,
+      "step": 22001
+    },
+    {
+      "epoch": 44.004,
+      "grad_norm": 1.2349199056625366,
+      "learning_rate": 2e-05,
+      "loss": 0.04709039,
+      "step": 22002
+    },
+    {
+      "epoch": 44.006,
+      "grad_norm": 1.0489506721496582,
+      "learning_rate": 2e-05,
+      "loss": 0.03470477,
+      "step": 22003
+    },
+    {
+      "epoch": 44.008,
+      "grad_norm": 1.1739320755004883,
+      "learning_rate": 2e-05,
+      "loss": 0.03277641,
+      "step": 22004
+    },
+    {
+      "epoch": 44.01,
+      "grad_norm": 2.5618724822998047,
+      "learning_rate": 2e-05,
+      "loss": 0.0545634,
+      "step": 22005
+    },
+    {
+      "epoch": 44.012,
+      "grad_norm": 1.349908471107483,
+      "learning_rate": 2e-05,
+      "loss": 0.04546019,
+      "step": 22006
+    },
+    {
+      "epoch": 44.014,
+      "grad_norm": 1.1877539157867432,
+      "learning_rate": 2e-05,
+      "loss": 0.05321749,
+      "step": 22007
+    },
+    {
+      "epoch": 44.016,
+      "grad_norm": 1.0178004503250122,
+      "learning_rate": 2e-05,
+      "loss": 0.02961883,
+      "step": 22008
+    },
+    {
+      "epoch": 44.018,
+      "grad_norm": 1.0790722370147705,
+      "learning_rate": 2e-05,
+      "loss": 0.04679262,
+      "step": 22009
+    },
+    {
+      "epoch": 44.02,
+      "grad_norm": 1.1426124572753906,
+      "learning_rate": 2e-05,
+      "loss": 0.06072292,
+      "step": 22010
+    },
+    {
+      "epoch": 44.022,
+      "grad_norm": 1.529797077178955,
+      "learning_rate": 2e-05,
+      "loss": 0.05173718,
+      "step": 22011
+    },
+    {
+      "epoch": 44.024,
+      "grad_norm": 1.4179189205169678,
+      "learning_rate": 2e-05,
+      "loss": 0.04669178,
+      "step": 22012
+    },
+    {
+      "epoch": 44.026,
+      "grad_norm": 1.4494482278823853,
+      "learning_rate": 2e-05,
+      "loss": 0.04122728,
+      "step": 22013
+    },
+    {
+      "epoch": 44.028,
+      "grad_norm": 1.4146836996078491,
+      "learning_rate": 2e-05,
+      "loss": 0.05055712,
+      "step": 22014
+    },
+    {
+      "epoch": 44.03,
+      "grad_norm": 1.372752070426941,
+      "learning_rate": 2e-05,
+      "loss": 0.05814591,
+      "step": 22015
+    },
+    {
+      "epoch": 44.032,
+      "grad_norm": 1.9753098487854004,
+      "learning_rate": 2e-05,
+      "loss": 0.0562949,
+      "step": 22016
+    },
+    {
+      "epoch": 44.034,
+      "grad_norm": 0.9869195222854614,
+      "learning_rate": 2e-05,
+      "loss": 0.04519263,
+      "step": 22017
+    },
+    {
+      "epoch": 44.036,
+      "grad_norm": 1.016710638999939,
+      "learning_rate": 2e-05,
+      "loss": 0.03520773,
+      "step": 22018
+    },
+    {
+      "epoch": 44.038,
+      "grad_norm": 1.2054697275161743,
+      "learning_rate": 2e-05,
+      "loss": 0.05319132,
+      "step": 22019
+    },
+    {
+      "epoch": 44.04,
+      "grad_norm": 1.1351009607315063,
+      "learning_rate": 2e-05,
+      "loss": 0.03978164,
+      "step": 22020
+    },
+    {
+      "epoch": 44.042,
+      "grad_norm": 1.0979878902435303,
+      "learning_rate": 2e-05,
+      "loss": 0.03140669,
+      "step": 22021
+    },
+    {
+      "epoch": 44.044,
+      "grad_norm": 1.446884036064148,
+      "learning_rate": 2e-05,
+      "loss": 0.05437474,
+      "step": 22022
+    },
+    {
+      "epoch": 44.046,
+      "grad_norm": 1.3147832155227661,
+      "learning_rate": 2e-05,
+      "loss": 0.05483336,
+      "step": 22023
+    },
+    {
+      "epoch": 44.048,
+      "grad_norm": 1.4436129331588745,
+      "learning_rate": 2e-05,
+      "loss": 0.05707975,
+      "step": 22024
+    },
+    {
+      "epoch": 44.05,
+      "grad_norm": 1.2296398878097534,
+      "learning_rate": 2e-05,
+      "loss": 0.05186768,
+      "step": 22025
+    },
+    {
+      "epoch": 44.052,
+      "grad_norm": 1.414892554283142,
+      "learning_rate": 2e-05,
+      "loss": 0.04964254,
+      "step": 22026
+    },
+    {
+      "epoch": 44.054,
+      "grad_norm": 0.9986278414726257,
+      "learning_rate": 2e-05,
+      "loss": 0.04020878,
+      "step": 22027
+    },
+    {
+      "epoch": 44.056,
+      "grad_norm": 1.0219697952270508,
+      "learning_rate": 2e-05,
+      "loss": 0.03773836,
+      "step": 22028
+    },
+    {
+      "epoch": 44.058,
+      "grad_norm": 0.9788745045661926,
+      "learning_rate": 2e-05,
+      "loss": 0.03778582,
+      "step": 22029
+    },
+    {
+      "epoch": 44.06,
+      "grad_norm": 1.209914207458496,
+      "learning_rate": 2e-05,
+      "loss": 0.03958928,
+      "step": 22030
+    },
+    {
+      "epoch": 44.062,
+      "grad_norm": 1.2830939292907715,
+      "learning_rate": 2e-05,
+      "loss": 0.04328164,
+      "step": 22031
+    },
+    {
+      "epoch": 44.064,
+      "grad_norm": 1.635453462600708,
+      "learning_rate": 2e-05,
+      "loss": 0.05318267,
+      "step": 22032
+    },
+    {
+      "epoch": 44.066,
+      "grad_norm": 0.9193920493125916,
+      "learning_rate": 2e-05,
+      "loss": 0.02856221,
+      "step": 22033
+    },
+    {
+      "epoch": 44.068,
+      "grad_norm": 1.4379791021347046,
+      "learning_rate": 2e-05,
+      "loss": 0.05032479,
+      "step": 22034
+    },
+    {
+      "epoch": 44.07,
+      "grad_norm": 2.1090707778930664,
+      "learning_rate": 2e-05,
+      "loss": 0.08148749,
+      "step": 22035
+    },
+    {
+      "epoch": 44.072,
+      "grad_norm": 1.4268964529037476,
+      "learning_rate": 2e-05,
+      "loss": 0.03891208,
+      "step": 22036
+    },
+    {
+      "epoch": 44.074,
+      "grad_norm": 1.3950912952423096,
+      "learning_rate": 2e-05,
+      "loss": 0.03417912,
+      "step": 22037
+    },
+    {
+      "epoch": 44.076,
+      "grad_norm": 1.4302080869674683,
+      "learning_rate": 2e-05,
+      "loss": 0.04385927,
+      "step": 22038
+    },
+    {
+      "epoch": 44.078,
+      "grad_norm": 1.038596272468567,
+      "learning_rate": 2e-05,
+      "loss": 0.03837661,
+      "step": 22039
+    },
+    {
+      "epoch": 44.08,
+      "grad_norm": 2.1667323112487793,
+      "learning_rate": 2e-05,
+      "loss": 0.06082205,
+      "step": 22040
+    },
+    {
+      "epoch": 44.082,
+      "grad_norm": 1.3383917808532715,
+      "learning_rate": 2e-05,
+      "loss": 0.05682547,
+      "step": 22041
+    },
+    {
+      "epoch": 44.084,
+      "grad_norm": 1.3785979747772217,
+      "learning_rate": 2e-05,
+      "loss": 0.05159343,
+      "step": 22042
+    },
+    {
+      "epoch": 44.086,
+      "grad_norm": 1.1491683721542358,
+      "learning_rate": 2e-05,
+      "loss": 0.04531821,
+      "step": 22043
+    },
+    {
+      "epoch": 44.088,
+      "grad_norm": 1.113952398300171,
+      "learning_rate": 2e-05,
+      "loss": 0.04688843,
+      "step": 22044
+    },
+    {
+      "epoch": 44.09,
+      "grad_norm": 1.2150129079818726,
+      "learning_rate": 2e-05,
+      "loss": 0.03948733,
+      "step": 22045
+    },
+    {
+      "epoch": 44.092,
+      "grad_norm": 1.161972999572754,
+      "learning_rate": 2e-05,
+      "loss": 0.03694011,
+      "step": 22046
+    },
+    {
+      "epoch": 44.094,
+      "grad_norm": 1.6014569997787476,
+      "learning_rate": 2e-05,
+      "loss": 0.06034542,
+      "step": 22047
+    },
+    {
+      "epoch": 44.096,
+      "grad_norm": 1.49050772190094,
+      "learning_rate": 2e-05,
+      "loss": 0.05083546,
+      "step": 22048
+    },
+    {
+      "epoch": 44.098,
+      "grad_norm": 1.4574774503707886,
+      "learning_rate": 2e-05,
+      "loss": 0.0556547,
+      "step": 22049
+    },
+    {
+      "epoch": 44.1,
+      "grad_norm": 1.13857102394104,
+      "learning_rate": 2e-05,
+      "loss": 0.04332735,
+      "step": 22050
+    },
+    {
+      "epoch": 44.102,
+      "grad_norm": 1.2549856901168823,
+      "learning_rate": 2e-05,
+      "loss": 0.05079667,
+      "step": 22051
+    },
+    {
+      "epoch": 44.104,
+      "grad_norm": 1.2218992710113525,
+      "learning_rate": 2e-05,
+      "loss": 0.05055429,
+      "step": 22052
+    },
+    {
+      "epoch": 44.106,
+      "grad_norm": 1.3079739809036255,
+      "learning_rate": 2e-05,
+      "loss": 0.05005455,
+      "step": 22053
+    },
+    {
+      "epoch": 44.108,
+      "grad_norm": 1.148532748222351,
+      "learning_rate": 2e-05,
+      "loss": 0.04688092,
+      "step": 22054
+    },
+    {
+      "epoch": 44.11,
+      "grad_norm": 1.397567629814148,
+      "learning_rate": 2e-05,
+      "loss": 0.0634838,
+      "step": 22055
+    },
+    {
+      "epoch": 44.112,
+      "grad_norm": 1.166181206703186,
+      "learning_rate": 2e-05,
+      "loss": 0.0324477,
+      "step": 22056
+    },
+    {
+      "epoch": 44.114,
+      "grad_norm": 1.310949444770813,
+      "learning_rate": 2e-05,
+      "loss": 0.04780035,
+      "step": 22057
+    },
+    {
+      "epoch": 44.116,
+      "grad_norm": 1.1098144054412842,
+      "learning_rate": 2e-05,
+      "loss": 0.03849977,
+      "step": 22058
+    },
+    {
+      "epoch": 44.118,
+      "grad_norm": 1.5903892517089844,
+      "learning_rate": 2e-05,
+      "loss": 0.05317472,
+      "step": 22059
+    },
+    {
+      "epoch": 44.12,
+      "grad_norm": 1.1845721006393433,
+      "learning_rate": 2e-05,
+      "loss": 0.03562384,
+      "step": 22060
+    },
+    {
+      "epoch": 44.122,
+      "grad_norm": 1.2428687810897827,
+      "learning_rate": 2e-05,
+      "loss": 0.05235748,
+      "step": 22061
+    },
+    {
+      "epoch": 44.124,
+      "grad_norm": 1.2767870426177979,
+      "learning_rate": 2e-05,
+      "loss": 0.05173537,
+      "step": 22062
+    },
+    {
+      "epoch": 44.126,
+      "grad_norm": 1.3087096214294434,
+      "learning_rate": 2e-05,
+      "loss": 0.05469508,
+      "step": 22063
+    },
+    {
+      "epoch": 44.128,
+      "grad_norm": 1.2753592729568481,
+      "learning_rate": 2e-05,
+      "loss": 0.05185987,
+      "step": 22064
+    },
+    {
+      "epoch": 44.13,
+      "grad_norm": 1.3028455972671509,
+      "learning_rate": 2e-05,
+      "loss": 0.05058619,
+      "step": 22065
+    },
+    {
+      "epoch": 44.132,
+      "grad_norm": 1.2384607791900635,
+      "learning_rate": 2e-05,
+      "loss": 0.04439488,
+      "step": 22066
+    },
+    {
+      "epoch": 44.134,
+      "grad_norm": 1.494373083114624,
+      "learning_rate": 2e-05,
+      "loss": 0.06326284,
+      "step": 22067
+    },
+    {
+      "epoch": 44.136,
+      "grad_norm": 4.574721813201904,
+      "learning_rate": 2e-05,
+      "loss": 0.0534743,
+      "step": 22068
+    },
+    {
+      "epoch": 44.138,
+      "grad_norm": 1.5381698608398438,
+      "learning_rate": 2e-05,
+      "loss": 0.06085927,
+      "step": 22069
+    },
+    {
+      "epoch": 44.14,
+      "grad_norm": 1.221481442451477,
+      "learning_rate": 2e-05,
+      "loss": 0.04375359,
+      "step": 22070
+    },
+    {
+      "epoch": 44.142,
+      "grad_norm": 1.3349063396453857,
+      "learning_rate": 2e-05,
+      "loss": 0.04402313,
+      "step": 22071
+    },
+    {
+      "epoch": 44.144,
+      "grad_norm": 1.3375403881072998,
+      "learning_rate": 2e-05,
+      "loss": 0.05106883,
+      "step": 22072
+    },
+    {
+      "epoch": 44.146,
+      "grad_norm": 1.1496771574020386,
+      "learning_rate": 2e-05,
+      "loss": 0.04525649,
+      "step": 22073
+    },
+    {
+      "epoch": 44.148,
+      "grad_norm": 1.2824701070785522,
+      "learning_rate": 2e-05,
+      "loss": 0.04934424,
+      "step": 22074
+    },
+    {
+      "epoch": 44.15,
+      "grad_norm": 1.1325626373291016,
+      "learning_rate": 2e-05,
+      "loss": 0.0284467,
+      "step": 22075
+    },
+    {
+      "epoch": 44.152,
+      "grad_norm": 1.3874157667160034,
+      "learning_rate": 2e-05,
+      "loss": 0.04672045,
+      "step": 22076
+    },
+    {
+      "epoch": 44.154,
+      "grad_norm": 3.2082982063293457,
+      "learning_rate": 2e-05,
+      "loss": 0.0601771,
+      "step": 22077
+    },
+    {
+      "epoch": 44.156,
+      "grad_norm": 2.588113784790039,
+      "learning_rate": 2e-05,
+      "loss": 0.07006887,
+      "step": 22078
+    },
+    {
+      "epoch": 44.158,
+      "grad_norm": 1.3412790298461914,
+      "learning_rate": 2e-05,
+      "loss": 0.04630299,
+      "step": 22079
+    },
+    {
+      "epoch": 44.16,
+      "grad_norm": 1.2379393577575684,
+      "learning_rate": 2e-05,
+      "loss": 0.04339472,
+      "step": 22080
+    },
+    {
+      "epoch": 44.162,
+      "grad_norm": 1.1797010898590088,
+      "learning_rate": 2e-05,
+      "loss": 0.0429916,
+      "step": 22081
+    },
+    {
+      "epoch": 44.164,
+      "grad_norm": 3.3057401180267334,
+      "learning_rate": 2e-05,
+      "loss": 0.04493307,
+      "step": 22082
+    },
+    {
+      "epoch": 44.166,
+      "grad_norm": 1.1744298934936523,
+      "learning_rate": 2e-05,
+      "loss": 0.03810238,
+      "step": 22083
+    },
+    {
+      "epoch": 44.168,
+      "grad_norm": 1.669509768486023,
+      "learning_rate": 2e-05,
+      "loss": 0.06496652,
+      "step": 22084
+    },
+    {
+      "epoch": 44.17,
+      "grad_norm": 1.5125724077224731,
+      "learning_rate": 2e-05,
+      "loss": 0.04929143,
+      "step": 22085
+    },
+    {
+      "epoch": 44.172,
+      "grad_norm": 1.4067198038101196,
+      "learning_rate": 2e-05,
+      "loss": 0.05213205,
+      "step": 22086
+    },
+    {
+      "epoch": 44.174,
+      "grad_norm": 1.189210057258606,
+      "learning_rate": 2e-05,
+      "loss": 0.04036959,
+      "step": 22087
+    },
+    {
+      "epoch": 44.176,
+      "grad_norm": 1.154815673828125,
+      "learning_rate": 2e-05,
+      "loss": 0.04020704,
+      "step": 22088
+    },
+    {
+      "epoch": 44.178,
+      "grad_norm": 1.233259677886963,
+      "learning_rate": 2e-05,
+      "loss": 0.05813593,
+      "step": 22089
+    },
+    {
+      "epoch": 44.18,
+      "grad_norm": 1.2222100496292114,
+      "learning_rate": 2e-05,
+      "loss": 0.04773624,
+      "step": 22090
+    },
+    {
+      "epoch": 44.182,
+      "grad_norm": 1.3617769479751587,
+      "learning_rate": 2e-05,
+      "loss": 0.04983572,
+      "step": 22091
+    },
+    {
+      "epoch": 44.184,
+      "grad_norm": 1.1668729782104492,
+      "learning_rate": 2e-05,
+      "loss": 0.04586762,
+      "step": 22092
+    },
+    {
+      "epoch": 44.186,
+      "grad_norm": 1.3036890029907227,
+      "learning_rate": 2e-05,
+      "loss": 0.05323526,
+      "step": 22093
+    },
+    {
+      "epoch": 44.188,
+      "grad_norm": 3.2073822021484375,
+      "learning_rate": 2e-05,
+      "loss": 0.04346972,
+      "step": 22094
+    },
+    {
+      "epoch": 44.19,
+      "grad_norm": 1.1499048471450806,
+      "learning_rate": 2e-05,
+      "loss": 0.05191308,
+      "step": 22095
+    },
+    {
+      "epoch": 44.192,
+      "grad_norm": 1.2350811958312988,
+      "learning_rate": 2e-05,
+      "loss": 0.04665541,
+      "step": 22096
+    },
+    {
+      "epoch": 44.194,
+      "grad_norm": 1.1341971158981323,
+      "learning_rate": 2e-05,
+      "loss": 0.03508368,
+      "step": 22097
+    },
+    {
+      "epoch": 44.196,
+      "grad_norm": 1.3536897897720337,
+      "learning_rate": 2e-05,
+      "loss": 0.04077377,
+      "step": 22098
+    },
+    {
+      "epoch": 44.198,
+      "grad_norm": 1.0278751850128174,
+      "learning_rate": 2e-05,
+      "loss": 0.03572151,
+      "step": 22099
+    },
+    {
+      "epoch": 44.2,
+      "grad_norm": 1.1718095541000366,
+      "learning_rate": 2e-05,
+      "loss": 0.04011806,
+      "step": 22100
+    },
+    {
+      "epoch": 44.202,
+      "grad_norm": 1.184627652168274,
+      "learning_rate": 2e-05,
+      "loss": 0.04278344,
+      "step": 22101
+    },
+    {
+      "epoch": 44.204,
+      "grad_norm": 1.8265912532806396,
+      "learning_rate": 2e-05,
+      "loss": 0.04747808,
+      "step": 22102
+    },
+    {
+      "epoch": 44.206,
+      "grad_norm": 1.1171393394470215,
+      "learning_rate": 2e-05,
+      "loss": 0.04222995,
+      "step": 22103
+    },
+    {
+      "epoch": 44.208,
+      "grad_norm": 1.0637423992156982,
+      "learning_rate": 2e-05,
+      "loss": 0.04015394,
+      "step": 22104
+    },
+    {
+      "epoch": 44.21,
+      "grad_norm": 1.0722296237945557,
+      "learning_rate": 2e-05,
+      "loss": 0.04177207,
+      "step": 22105
+    },
+    {
+      "epoch": 44.212,
+      "grad_norm": 1.208267092704773,
+      "learning_rate": 2e-05,
+      "loss": 0.04094924,
+      "step": 22106
+    },
+    {
+      "epoch": 44.214,
+      "grad_norm": 2.877411365509033,
+      "learning_rate": 2e-05,
+      "loss": 0.05537339,
+      "step": 22107
+    },
+    {
+      "epoch": 44.216,
+      "grad_norm": 3.210109233856201,
+      "learning_rate": 2e-05,
+      "loss": 0.06678712,
+      "step": 22108
+    },
+    {
+      "epoch": 44.218,
+      "grad_norm": 1.6754549741744995,
+      "learning_rate": 2e-05,
+      "loss": 0.05157789,
+      "step": 22109
+    },
+    {
+      "epoch": 44.22,
+      "grad_norm": 1.1392402648925781,
+      "learning_rate": 2e-05,
+      "loss": 0.04350193,
+      "step": 22110
+    },
+    {
+      "epoch": 44.222,
+      "grad_norm": 1.8385316133499146,
+      "learning_rate": 2e-05,
+      "loss": 0.03925511,
+      "step": 22111
+    },
+    {
+      "epoch": 44.224,
+      "grad_norm": 1.453076958656311,
+      "learning_rate": 2e-05,
+      "loss": 0.05029003,
+      "step": 22112
+    },
+    {
+      "epoch": 44.226,
+      "grad_norm": 1.1068811416625977,
+      "learning_rate": 2e-05,
+      "loss": 0.05229628,
+      "step": 22113
+    },
+    {
+      "epoch": 44.228,
+      "grad_norm": 1.3013540506362915,
+      "learning_rate": 2e-05,
+      "loss": 0.05385014,
+      "step": 22114
+    },
+    {
+      "epoch": 44.23,
+      "grad_norm": 1.258337140083313,
+      "learning_rate": 2e-05,
+      "loss": 0.04877231,
+      "step": 22115
+    },
+    {
+      "epoch": 44.232,
+      "grad_norm": 1.5271419286727905,
+      "learning_rate": 2e-05,
+      "loss": 0.0496937,
+      "step": 22116
+    },
+    {
+      "epoch": 44.234,
+      "grad_norm": 1.1020830869674683,
+      "learning_rate": 2e-05,
+      "loss": 0.0386355,
+      "step": 22117
+    },
+    {
+      "epoch": 44.236,
+      "grad_norm": 0.9935838580131531,
+      "learning_rate": 2e-05,
+      "loss": 0.02613112,
+      "step": 22118
+    },
+    {
+      "epoch": 44.238,
+      "grad_norm": 1.5669888257980347,
+      "learning_rate": 2e-05,
+      "loss": 0.05612132,
+      "step": 22119
+    },
+    {
+      "epoch": 44.24,
+      "grad_norm": 1.1000120639801025,
+      "learning_rate": 2e-05,
+      "loss": 0.041154,
+      "step": 22120
+    },
+    {
+      "epoch": 44.242,
+      "grad_norm": 1.1130436658859253,
+      "learning_rate": 2e-05,
+      "loss": 0.03398903,
+      "step": 22121
+    },
+    {
+      "epoch": 44.244,
+      "grad_norm": 1.3796424865722656,
+      "learning_rate": 2e-05,
+      "loss": 0.05808487,
+      "step": 22122
+    },
+    {
+      "epoch": 44.246,
+      "grad_norm": 1.7627161741256714,
+      "learning_rate": 2e-05,
+      "loss": 0.05050261,
+      "step": 22123
+    },
+    {
+      "epoch": 44.248,
+      "grad_norm": 1.2039992809295654,
+      "learning_rate": 2e-05,
+      "loss": 0.0452828,
+      "step": 22124
+    },
+    {
+      "epoch": 44.25,
+      "grad_norm": 1.6339401006698608,
+      "learning_rate": 2e-05,
+      "loss": 0.07191573,
+      "step": 22125
+    },
+    {
+      "epoch": 44.252,
+      "grad_norm": 1.870229959487915,
+      "learning_rate": 2e-05,
+      "loss": 0.04964413,
+      "step": 22126
+    },
+    {
+      "epoch": 44.254,
+      "grad_norm": 1.4508622884750366,
+      "learning_rate": 2e-05,
+      "loss": 0.05981822,
+      "step": 22127
+    },
+    {
+      "epoch": 44.256,
+      "grad_norm": 1.3774610757827759,
+      "learning_rate": 2e-05,
+      "loss": 0.04445232,
+      "step": 22128
+    },
+    {
+      "epoch": 44.258,
+      "grad_norm": 1.1424156427383423,
+      "learning_rate": 2e-05,
+      "loss": 0.04557105,
+      "step": 22129
+    },
+    {
+      "epoch": 44.26,
+      "grad_norm": 1.0394034385681152,
+      "learning_rate": 2e-05,
+      "loss": 0.0336099,
+      "step": 22130
+    },
+    {
+      "epoch": 44.262,
+      "grad_norm": 1.2787925004959106,
+      "learning_rate": 2e-05,
+      "loss": 0.04919507,
+      "step": 22131
+    },
+    {
+      "epoch": 44.264,
+      "grad_norm": 0.9578052759170532,
+      "learning_rate": 2e-05,
+      "loss": 0.03036173,
+      "step": 22132
+    },
+    {
+      "epoch": 44.266,
+      "grad_norm": 1.0824671983718872,
+      "learning_rate": 2e-05,
+      "loss": 0.04263349,
+      "step": 22133
+    },
+    {
+      "epoch": 44.268,
+      "grad_norm": 1.273948311805725,
+      "learning_rate": 2e-05,
+      "loss": 0.03732668,
+      "step": 22134
+    },
+    {
+      "epoch": 44.27,
+      "grad_norm": 1.1770116090774536,
+      "learning_rate": 2e-05,
+      "loss": 0.04749018,
+      "step": 22135
+    },
+    {
+      "epoch": 44.272,
+      "grad_norm": 1.3313310146331787,
+      "learning_rate": 2e-05,
+      "loss": 0.04976882,
+      "step": 22136
+    },
+    {
+      "epoch": 44.274,
+      "grad_norm": 2.539590358734131,
+      "learning_rate": 2e-05,
+      "loss": 0.04594205,
+      "step": 22137
+    },
+    {
+      "epoch": 44.276,
+      "grad_norm": 1.6731163263320923,
+      "learning_rate": 2e-05,
+      "loss": 0.06411099,
+      "step": 22138
+    },
+    {
+      "epoch": 44.278,
+      "grad_norm": 1.0700182914733887,
+      "learning_rate": 2e-05,
+      "loss": 0.03831624,
+      "step": 22139
+    },
+    {
+      "epoch": 44.28,
+      "grad_norm": 1.190203309059143,
+      "learning_rate": 2e-05,
+      "loss": 0.04036564,
+      "step": 22140
+    },
+    {
+      "epoch": 44.282,
+      "grad_norm": 3.227332830429077,
+      "learning_rate": 2e-05,
+      "loss": 0.06097856,
+      "step": 22141
+    },
+    {
+      "epoch": 44.284,
+      "grad_norm": 1.274222493171692,
+      "learning_rate": 2e-05,
+      "loss": 0.04809301,
+      "step": 22142
+    },
+    {
+      "epoch": 44.286,
+      "grad_norm": 1.2631959915161133,
+      "learning_rate": 2e-05,
+      "loss": 0.0570578,
+      "step": 22143
+    },
+    {
+      "epoch": 44.288,
+      "grad_norm": 1.4169923067092896,
+      "learning_rate": 2e-05,
+      "loss": 0.05213282,
+      "step": 22144
+    },
+    {
+      "epoch": 44.29,
+      "grad_norm": 1.2868964672088623,
+      "learning_rate": 2e-05,
+      "loss": 0.05694466,
+      "step": 22145
+    },
+    {
+      "epoch": 44.292,
+      "grad_norm": 1.0576223134994507,
+      "learning_rate": 2e-05,
+      "loss": 0.04163115,
+      "step": 22146
+    },
+    {
+      "epoch": 44.294,
+      "grad_norm": 2.8826777935028076,
+      "learning_rate": 2e-05,
+      "loss": 0.05488094,
+      "step": 22147
+    },
+    {
+      "epoch": 44.296,
+      "grad_norm": 1.1898715496063232,
+      "learning_rate": 2e-05,
+      "loss": 0.05081913,
+      "step": 22148
+    },
+    {
+      "epoch": 44.298,
+      "grad_norm": 1.2379268407821655,
+      "learning_rate": 2e-05,
+      "loss": 0.05061274,
+      "step": 22149
+    },
+    {
+      "epoch": 44.3,
+      "grad_norm": 1.355218529701233,
+      "learning_rate": 2e-05,
+      "loss": 0.05804717,
+      "step": 22150
+    },
+    {
+      "epoch": 44.302,
+      "grad_norm": 1.3628958463668823,
+      "learning_rate": 2e-05,
+      "loss": 0.04630186,
+      "step": 22151
+    },
+    {
+      "epoch": 44.304,
+      "grad_norm": 1.064475178718567,
+      "learning_rate": 2e-05,
+      "loss": 0.0363181,
+      "step": 22152
+    },
+    {
+      "epoch": 44.306,
+      "grad_norm": 1.3461925983428955,
+      "learning_rate": 2e-05,
+      "loss": 0.03945461,
+      "step": 22153
+    },
+    {
+      "epoch": 44.308,
+      "grad_norm": 1.2384178638458252,
+      "learning_rate": 2e-05,
+      "loss": 0.03994835,
+      "step": 22154
+    },
+    {
+      "epoch": 44.31,
+      "grad_norm": 1.2434017658233643,
+      "learning_rate": 2e-05,
+      "loss": 0.05772359,
+      "step": 22155
+    },
+    {
+      "epoch": 44.312,
+      "grad_norm": 1.719817876815796,
+      "learning_rate": 2e-05,
+      "loss": 0.04721334,
+      "step": 22156
+    },
+    {
+      "epoch": 44.314,
+      "grad_norm": 1.0811158418655396,
+      "learning_rate": 2e-05,
+      "loss": 0.05218527,
+      "step": 22157
+    },
+    {
+      "epoch": 44.316,
+      "grad_norm": 1.283691644668579,
+      "learning_rate": 2e-05,
+      "loss": 0.05287713,
+      "step": 22158
+    },
+    {
+      "epoch": 44.318,
+      "grad_norm": 1.1990177631378174,
+      "learning_rate": 2e-05,
+      "loss": 0.04951636,
+      "step": 22159
+    },
+    {
+      "epoch": 44.32,
+      "grad_norm": 2.536614418029785,
+      "learning_rate": 2e-05,
+      "loss": 0.04560195,
+      "step": 22160
+    },
+    {
+      "epoch": 44.322,
+      "grad_norm": 1.0507843494415283,
+      "learning_rate": 2e-05,
+      "loss": 0.03920343,
+      "step": 22161
+    },
+    {
+      "epoch": 44.324,
+      "grad_norm": 1.4071505069732666,
+      "learning_rate": 2e-05,
+      "loss": 0.05998628,
+      "step": 22162
+    },
+    {
+      "epoch": 44.326,
+      "grad_norm": 1.6779695749282837,
+      "learning_rate": 2e-05,
+      "loss": 0.05833928,
+      "step": 22163
+    },
+    {
+      "epoch": 44.328,
+      "grad_norm": 1.0807090997695923,
+      "learning_rate": 2e-05,
+      "loss": 0.04173718,
+      "step": 22164
+    },
+    {
+      "epoch": 44.33,
+      "grad_norm": 1.7728112936019897,
+      "learning_rate": 2e-05,
+      "loss": 0.05153607,
+      "step": 22165
+    },
+    {
+      "epoch": 44.332,
+      "grad_norm": 1.454960584640503,
+      "learning_rate": 2e-05,
+      "loss": 0.05345749,
+      "step": 22166
+    },
+    {
+      "epoch": 44.334,
+      "grad_norm": 1.146530032157898,
+      "learning_rate": 2e-05,
+      "loss": 0.05334022,
+      "step": 22167
+    },
+    {
+      "epoch": 44.336,
+      "grad_norm": 1.7574084997177124,
+      "learning_rate": 2e-05,
+      "loss": 0.03719595,
+      "step": 22168
+    },
+    {
+      "epoch": 44.338,
+      "grad_norm": 1.4816433191299438,
+      "learning_rate": 2e-05,
+      "loss": 0.07536916,
+      "step": 22169
+    },
+    {
+      "epoch": 44.34,
+      "grad_norm": 1.3355191946029663,
+      "learning_rate": 2e-05,
+      "loss": 0.04837946,
+      "step": 22170
+    },
+    {
+      "epoch": 44.342,
+      "grad_norm": 1.2563939094543457,
+      "learning_rate": 2e-05,
+      "loss": 0.04800646,
+      "step": 22171
+    },
+    {
+      "epoch": 44.344,
+      "grad_norm": 1.0994011163711548,
+      "learning_rate": 2e-05,
+      "loss": 0.03690449,
+      "step": 22172
+    },
+    {
+      "epoch": 44.346,
+      "grad_norm": 1.6743828058242798,
+      "learning_rate": 2e-05,
+      "loss": 0.04229587,
+      "step": 22173
+    },
+    {
+      "epoch": 44.348,
+      "grad_norm": 1.1569823026657104,
+      "learning_rate": 2e-05,
+      "loss": 0.06016696,
+      "step": 22174
+    },
+    {
+      "epoch": 44.35,
+      "grad_norm": 1.8598023653030396,
+      "learning_rate": 2e-05,
+      "loss": 0.06339575,
+      "step": 22175
+    },
+    {
+      "epoch": 44.352,
+      "grad_norm": 1.2611335515975952,
+      "learning_rate": 2e-05,
+      "loss": 0.05222043,
+      "step": 22176
+    },
+    {
+      "epoch": 44.354,
+      "grad_norm": 0.9472651481628418,
+      "learning_rate": 2e-05,
+      "loss": 0.03059671,
+      "step": 22177
+    },
+    {
+      "epoch": 44.356,
+      "grad_norm": 1.521095633506775,
+      "learning_rate": 2e-05,
+      "loss": 0.05748924,
+      "step": 22178
+    },
+    {
+      "epoch": 44.358,
+      "grad_norm": 1.19306218624115,
+      "learning_rate": 2e-05,
+      "loss": 0.04892451,
+      "step": 22179
+    },
+    {
+      "epoch": 44.36,
+      "grad_norm": 1.12234628200531,
+      "learning_rate": 2e-05,
+      "loss": 0.0413163,
+      "step": 22180
+    },
+    {
+      "epoch": 44.362,
+      "grad_norm": 1.408831000328064,
+      "learning_rate": 2e-05,
+      "loss": 0.04378711,
+      "step": 22181
+    },
+    {
+      "epoch": 44.364,
+      "grad_norm": 1.1348880529403687,
+      "learning_rate": 2e-05,
+      "loss": 0.04664857,
+      "step": 22182
+    },
+    {
+      "epoch": 44.366,
+      "grad_norm": 1.254754662513733,
+      "learning_rate": 2e-05,
+      "loss": 0.03291727,
+      "step": 22183
+    },
+    {
+      "epoch": 44.368,
+      "grad_norm": 2.050689458847046,
+      "learning_rate": 2e-05,
+      "loss": 0.06108903,
+      "step": 22184
+    },
+    {
+      "epoch": 44.37,
+      "grad_norm": 1.1247540712356567,
+      "learning_rate": 2e-05,
+      "loss": 0.04102144,
+      "step": 22185
+    },
+    {
+      "epoch": 44.372,
+      "grad_norm": 1.2489968538284302,
+      "learning_rate": 2e-05,
+      "loss": 0.04336529,
+      "step": 22186
+    },
+    {
+      "epoch": 44.374,
+      "grad_norm": 1.2883358001708984,
+      "learning_rate": 2e-05,
+      "loss": 0.04382469,
+      "step": 22187
+    },
+    {
+      "epoch": 44.376,
+      "grad_norm": 1.5653047561645508,
+      "learning_rate": 2e-05,
+      "loss": 0.05455358,
+      "step": 22188
+    },
+    {
+      "epoch": 44.378,
+      "grad_norm": 1.1407490968704224,
+      "learning_rate": 2e-05,
+      "loss": 0.04385475,
+      "step": 22189
+    },
+    {
+      "epoch": 44.38,
+      "grad_norm": 0.9423237442970276,
+      "learning_rate": 2e-05,
+      "loss": 0.03583224,
+      "step": 22190
+    },
+    {
+      "epoch": 44.382,
+      "grad_norm": 1.5692898035049438,
+      "learning_rate": 2e-05,
+      "loss": 0.06093635,
+      "step": 22191
+    },
+    {
+      "epoch": 44.384,
+      "grad_norm": 1.2798019647598267,
+      "learning_rate": 2e-05,
+      "loss": 0.04287341,
+      "step": 22192
+    },
+    {
+      "epoch": 44.386,
+      "grad_norm": 2.308946132659912,
+      "learning_rate": 2e-05,
+      "loss": 0.0538884,
+      "step": 22193
+    },
+    {
+      "epoch": 44.388,
+      "grad_norm": 1.9377248287200928,
+      "learning_rate": 2e-05,
+      "loss": 0.06390283,
+      "step": 22194
+    },
+    {
+      "epoch": 44.39,
+      "grad_norm": 1.157041072845459,
+      "learning_rate": 2e-05,
+      "loss": 0.04512091,
+      "step": 22195
+    },
+    {
+      "epoch": 44.392,
+      "grad_norm": 1.2009409666061401,
+      "learning_rate": 2e-05,
+      "loss": 0.04480921,
+      "step": 22196
+    },
+    {
+      "epoch": 44.394,
+      "grad_norm": 1.0632206201553345,
+      "learning_rate": 2e-05,
+      "loss": 0.04720378,
+      "step": 22197
+    },
+    {
+      "epoch": 44.396,
+      "grad_norm": 2.500654935836792,
+      "learning_rate": 2e-05,
+      "loss": 0.07204752,
+      "step": 22198
+    },
+    {
+      "epoch": 44.398,
+      "grad_norm": 2.257668972015381,
+      "learning_rate": 2e-05,
+      "loss": 0.06277246,
+      "step": 22199
+    },
+    {
+      "epoch": 44.4,
+      "grad_norm": 1.4456931352615356,
+      "learning_rate": 2e-05,
+      "loss": 0.06364234,
+      "step": 22200
+    },
+    {
+      "epoch": 44.402,
+      "grad_norm": 1.1703506708145142,
+      "learning_rate": 2e-05,
+      "loss": 0.04526365,
+      "step": 22201
+    },
+    {
+      "epoch": 44.404,
+      "grad_norm": 0.9013957381248474,
+      "learning_rate": 2e-05,
+      "loss": 0.02716382,
+      "step": 22202
+    },
+    {
+      "epoch": 44.406,
+      "grad_norm": 1.0054999589920044,
+      "learning_rate": 2e-05,
+      "loss": 0.04023641,
+      "step": 22203
+    },
+    {
+      "epoch": 44.408,
+      "grad_norm": 1.7909125089645386,
+      "learning_rate": 2e-05,
+      "loss": 0.05618212,
+      "step": 22204
+    },
+    {
+      "epoch": 44.41,
+      "grad_norm": 1.3415671586990356,
+      "learning_rate": 2e-05,
+      "loss": 0.04464972,
+      "step": 22205
+    },
+    {
+      "epoch": 44.412,
+      "grad_norm": 1.8720018863677979,
+      "learning_rate": 2e-05,
+      "loss": 0.06667548,
+      "step": 22206
+    },
+    {
+      "epoch": 44.414,
+      "grad_norm": 1.0374834537506104,
+      "learning_rate": 2e-05,
+      "loss": 0.03846564,
+      "step": 22207
+    },
+    {
+      "epoch": 44.416,
+      "grad_norm": 1.2723255157470703,
+      "learning_rate": 2e-05,
+      "loss": 0.05234695,
+      "step": 22208
+    },
+    {
+      "epoch": 44.418,
+      "grad_norm": 1.134573221206665,
+      "learning_rate": 2e-05,
+      "loss": 0.04465498,
+      "step": 22209
+    },
+    {
+      "epoch": 44.42,
+      "grad_norm": 1.5191528797149658,
+      "learning_rate": 2e-05,
+      "loss": 0.0526737,
+      "step": 22210
+    },
+    {
+      "epoch": 44.422,
+      "grad_norm": 1.506665825843811,
+      "learning_rate": 2e-05,
+      "loss": 0.06554738,
+      "step": 22211
+    },
+    {
+      "epoch": 44.424,
+      "grad_norm": 1.1496007442474365,
+      "learning_rate": 2e-05,
+      "loss": 0.05143321,
+      "step": 22212
+    },
+    {
+      "epoch": 44.426,
+      "grad_norm": 1.3427084684371948,
+      "learning_rate": 2e-05,
+      "loss": 0.05444918,
+      "step": 22213
+    },
+    {
+      "epoch": 44.428,
+      "grad_norm": 1.3482060432434082,
+      "learning_rate": 2e-05,
+      "loss": 0.04946334,
+      "step": 22214
+    },
+    {
+      "epoch": 44.43,
+      "grad_norm": 2.4248390197753906,
+      "learning_rate": 2e-05,
+      "loss": 0.06155992,
+      "step": 22215
+    },
+    {
+      "epoch": 44.432,
+      "grad_norm": 1.5861310958862305,
+      "learning_rate": 2e-05,
+      "loss": 0.0607713,
+      "step": 22216
+    },
+    {
+      "epoch": 44.434,
+      "grad_norm": 1.7854304313659668,
+      "learning_rate": 2e-05,
+      "loss": 0.05506391,
+      "step": 22217
+    },
+    {
+      "epoch": 44.436,
+      "grad_norm": 1.5562728643417358,
+      "learning_rate": 2e-05,
+      "loss": 0.05874497,
+      "step": 22218
+    },
+    {
+      "epoch": 44.438,
+      "grad_norm": 1.5572162866592407,
+      "learning_rate": 2e-05,
+      "loss": 0.04674251,
+      "step": 22219
+    },
+    {
+      "epoch": 44.44,
+      "grad_norm": 2.5241782665252686,
+      "learning_rate": 2e-05,
+      "loss": 0.05468413,
+      "step": 22220
+    },
+    {
+      "epoch": 44.442,
+      "grad_norm": 1.4192159175872803,
+      "learning_rate": 2e-05,
+      "loss": 0.05214172,
+      "step": 22221
+    },
+    {
+      "epoch": 44.444,
+      "grad_norm": 1.1719669103622437,
+      "learning_rate": 2e-05,
+      "loss": 0.0375664,
+      "step": 22222
+    },
+    {
+      "epoch": 44.446,
+      "grad_norm": 1.4639357328414917,
+      "learning_rate": 2e-05,
+      "loss": 0.06297034,
+      "step": 22223
+    },
+    {
+      "epoch": 44.448,
+      "grad_norm": 1.3242526054382324,
+      "learning_rate": 2e-05,
+      "loss": 0.05253812,
+      "step": 22224
+    },
+    {
+      "epoch": 44.45,
+      "grad_norm": 1.2953046560287476,
+      "learning_rate": 2e-05,
+      "loss": 0.05739931,
+      "step": 22225
+    },
+    {
+      "epoch": 44.452,
+      "grad_norm": 2.824225425720215,
+      "learning_rate": 2e-05,
+      "loss": 0.05475694,
+      "step": 22226
+    },
+    {
+      "epoch": 44.454,
+      "grad_norm": 2.2453224658966064,
+      "learning_rate": 2e-05,
+      "loss": 0.04598641,
+      "step": 22227
+    },
+    {
+      "epoch": 44.456,
+      "grad_norm": 1.2624878883361816,
+      "learning_rate": 2e-05,
+      "loss": 0.05295527,
+      "step": 22228
+    },
+    {
+      "epoch": 44.458,
+      "grad_norm": 1.4258654117584229,
+      "learning_rate": 2e-05,
+      "loss": 0.05860892,
+      "step": 22229
+    },
+    {
+      "epoch": 44.46,
+      "grad_norm": 1.2828129529953003,
+      "learning_rate": 2e-05,
+      "loss": 0.04799199,
+      "step": 22230
+    },
+    {
+      "epoch": 44.462,
+      "grad_norm": 1.1467986106872559,
+      "learning_rate": 2e-05,
+      "loss": 0.05302534,
+      "step": 22231
+    },
+    {
+      "epoch": 44.464,
+      "grad_norm": 1.229738473892212,
+      "learning_rate": 2e-05,
+      "loss": 0.04447974,
+      "step": 22232
+    },
+    {
+      "epoch": 44.466,
+      "grad_norm": 1.252163052558899,
+      "learning_rate": 2e-05,
+      "loss": 0.04261618,
+      "step": 22233
+    },
+    {
+      "epoch": 44.468,
+      "grad_norm": 1.4980531930923462,
+      "learning_rate": 2e-05,
+      "loss": 0.04990222,
+      "step": 22234
+    },
+    {
+      "epoch": 44.47,
+      "grad_norm": 0.8647037148475647,
+      "learning_rate": 2e-05,
+      "loss": 0.03477815,
+      "step": 22235
+    },
+    {
+      "epoch": 44.472,
+      "grad_norm": 1.2490501403808594,
+      "learning_rate": 2e-05,
+      "loss": 0.05593645,
+      "step": 22236
+    },
+    {
+      "epoch": 44.474,
+      "grad_norm": 2.7238969802856445,
+      "learning_rate": 2e-05,
+      "loss": 0.07153691,
+      "step": 22237
+    },
+    {
+      "epoch": 44.476,
+      "grad_norm": 1.0918712615966797,
+      "learning_rate": 2e-05,
+      "loss": 0.03975195,
+      "step": 22238
+    },
+    {
+      "epoch": 44.478,
+      "grad_norm": 1.2695947885513306,
+      "learning_rate": 2e-05,
+      "loss": 0.0505439,
+      "step": 22239
+    },
+    {
+      "epoch": 44.48,
+      "grad_norm": 1.305942177772522,
+      "learning_rate": 2e-05,
+      "loss": 0.05622169,
+      "step": 22240
+    },
+    {
+      "epoch": 44.482,
+      "grad_norm": 1.5338377952575684,
+      "learning_rate": 2e-05,
+      "loss": 0.06709421,
+      "step": 22241
+    },
+    {
+      "epoch": 44.484,
+      "grad_norm": 1.2748602628707886,
+      "learning_rate": 2e-05,
+      "loss": 0.05205786,
+      "step": 22242
+    },
+    {
+      "epoch": 44.486,
+      "grad_norm": 1.1404742002487183,
+      "learning_rate": 2e-05,
+      "loss": 0.04140478,
+      "step": 22243
+    },
+    {
+      "epoch": 44.488,
+      "grad_norm": 1.2886909246444702,
+      "learning_rate": 2e-05,
+      "loss": 0.05009973,
+      "step": 22244
+    },
+    {
+      "epoch": 44.49,
+      "grad_norm": 1.3291938304901123,
+      "learning_rate": 2e-05,
+      "loss": 0.03881418,
+      "step": 22245
+    },
+    {
+      "epoch": 44.492,
+      "grad_norm": 1.743362307548523,
+      "learning_rate": 2e-05,
+      "loss": 0.06562784,
+      "step": 22246
+    },
+    {
+      "epoch": 44.494,
+      "grad_norm": 1.1744399070739746,
+      "learning_rate": 2e-05,
+      "loss": 0.03672204,
+      "step": 22247
+    },
+    {
+      "epoch": 44.496,
+      "grad_norm": 1.9760868549346924,
+      "learning_rate": 2e-05,
+      "loss": 0.04297481,
+      "step": 22248
+    },
+    {
+      "epoch": 44.498,
+      "grad_norm": 1.2893195152282715,
+      "learning_rate": 2e-05,
+      "loss": 0.0419421,
+      "step": 22249
+    },
+    {
+      "epoch": 44.5,
+      "grad_norm": 1.3587701320648193,
+      "learning_rate": 2e-05,
+      "loss": 0.03800445,
+      "step": 22250
+    },
+    {
+      "epoch": 44.502,
+      "grad_norm": 1.7126094102859497,
+      "learning_rate": 2e-05,
+      "loss": 0.04739122,
+      "step": 22251
+    },
+    {
+      "epoch": 44.504,
+      "grad_norm": 1.3062655925750732,
+      "learning_rate": 2e-05,
+      "loss": 0.0504301,
+      "step": 22252
+    },
+    {
+      "epoch": 44.506,
+      "grad_norm": 1.2278339862823486,
+      "learning_rate": 2e-05,
+      "loss": 0.05388809,
+      "step": 22253
+    },
+    {
+      "epoch": 44.508,
+      "grad_norm": 1.3173941373825073,
+      "learning_rate": 2e-05,
+      "loss": 0.03916169,
+      "step": 22254
+    },
+    {
+      "epoch": 44.51,
+      "grad_norm": 1.1290494203567505,
+      "learning_rate": 2e-05,
+      "loss": 0.04224993,
+      "step": 22255
+    },
+    {
+      "epoch": 44.512,
+      "grad_norm": 1.289548397064209,
+      "learning_rate": 2e-05,
+      "loss": 0.05193936,
+      "step": 22256
+    },
+    {
+      "epoch": 44.514,
+      "grad_norm": 1.2119817733764648,
+      "learning_rate": 2e-05,
+      "loss": 0.04464715,
+      "step": 22257
+    },
+    {
+      "epoch": 44.516,
+      "grad_norm": 1.1168314218521118,
+      "learning_rate": 2e-05,
+      "loss": 0.05150532,
+      "step": 22258
+    },
+    {
+      "epoch": 44.518,
+      "grad_norm": 1.120448112487793,
+      "learning_rate": 2e-05,
+      "loss": 0.04792247,
+      "step": 22259
+    },
+    {
+      "epoch": 44.52,
+      "grad_norm": 1.3556797504425049,
+      "learning_rate": 2e-05,
+      "loss": 0.05941795,
+      "step": 22260
+    },
+    {
+      "epoch": 44.522,
+      "grad_norm": 1.177809715270996,
+      "learning_rate": 2e-05,
+      "loss": 0.04000936,
+      "step": 22261
+    },
+    {
+      "epoch": 44.524,
+      "grad_norm": 1.567018747329712,
+      "learning_rate": 2e-05,
+      "loss": 0.04125568,
+      "step": 22262
+    },
+    {
+      "epoch": 44.526,
+      "grad_norm": 1.170555830001831,
+      "learning_rate": 2e-05,
+      "loss": 0.03684732,
+      "step": 22263
+    },
+    {
+      "epoch": 44.528,
+      "grad_norm": 1.2405741214752197,
+      "learning_rate": 2e-05,
+      "loss": 0.04239763,
+      "step": 22264
+    },
+    {
+      "epoch": 44.53,
+      "grad_norm": 1.0276159048080444,
+      "learning_rate": 2e-05,
+      "loss": 0.0344376,
+      "step": 22265
+    },
+    {
+      "epoch": 44.532,
+      "grad_norm": 1.2948598861694336,
+      "learning_rate": 2e-05,
+      "loss": 0.05724182,
+      "step": 22266
+    },
+    {
+      "epoch": 44.534,
+      "grad_norm": 1.2418267726898193,
+      "learning_rate": 2e-05,
+      "loss": 0.05686314,
+      "step": 22267
+    },
+    {
+      "epoch": 44.536,
+      "grad_norm": 1.2589337825775146,
+      "learning_rate": 2e-05,
+      "loss": 0.04628392,
+      "step": 22268
+    },
+    {
+      "epoch": 44.538,
+      "grad_norm": 1.343007206916809,
+      "learning_rate": 2e-05,
+      "loss": 0.06061236,
+      "step": 22269
+    },
+    {
+      "epoch": 44.54,
+      "grad_norm": 1.2776858806610107,
+      "learning_rate": 2e-05,
+      "loss": 0.0425747,
+      "step": 22270
+    },
+    {
+      "epoch": 44.542,
+      "grad_norm": 1.1150517463684082,
+      "learning_rate": 2e-05,
+      "loss": 0.03600602,
+      "step": 22271
+    },
+    {
+      "epoch": 44.544,
+      "grad_norm": 1.9361085891723633,
+      "learning_rate": 2e-05,
+      "loss": 0.04623732,
+      "step": 22272
+    },
+    {
+      "epoch": 44.546,
+      "grad_norm": 1.4865039587020874,
+      "learning_rate": 2e-05,
+      "loss": 0.04633596,
+      "step": 22273
+    },
+    {
+      "epoch": 44.548,
+      "grad_norm": 1.1509850025177002,
+      "learning_rate": 2e-05,
+      "loss": 0.03759123,
+      "step": 22274
+    },
+    {
+      "epoch": 44.55,
+      "grad_norm": 1.38397216796875,
+      "learning_rate": 2e-05,
+      "loss": 0.05729149,
+      "step": 22275
+    },
+    {
+      "epoch": 44.552,
+      "grad_norm": 1.1179465055465698,
+      "learning_rate": 2e-05,
+      "loss": 0.03808307,
+      "step": 22276
+    },
+    {
+      "epoch": 44.554,
+      "grad_norm": 2.435750961303711,
+      "learning_rate": 2e-05,
+      "loss": 0.02759157,
+      "step": 22277
+    },
+    {
+      "epoch": 44.556,
+      "grad_norm": 1.195075273513794,
+      "learning_rate": 2e-05,
+      "loss": 0.03713936,
+      "step": 22278
+    },
+    {
+      "epoch": 44.558,
+      "grad_norm": 1.1682522296905518,
+      "learning_rate": 2e-05,
+      "loss": 0.04102589,
+      "step": 22279
+    },
+    {
+      "epoch": 44.56,
+      "grad_norm": 1.4527888298034668,
+      "learning_rate": 2e-05,
+      "loss": 0.06393797,
+      "step": 22280
+    },
+    {
+      "epoch": 44.562,
+      "grad_norm": 1.5062603950500488,
+      "learning_rate": 2e-05,
+      "loss": 0.07412426,
+      "step": 22281
+    },
+    {
+      "epoch": 44.564,
+      "grad_norm": 1.2519084215164185,
+      "learning_rate": 2e-05,
+      "loss": 0.04865194,
+      "step": 22282
+    },
+    {
+      "epoch": 44.566,
+      "grad_norm": 1.912338137626648,
+      "learning_rate": 2e-05,
+      "loss": 0.0646146,
+      "step": 22283
+    },
+    {
+      "epoch": 44.568,
+      "grad_norm": 1.1479963064193726,
+      "learning_rate": 2e-05,
+      "loss": 0.04428418,
+      "step": 22284
+    },
+    {
+      "epoch": 44.57,
+      "grad_norm": 1.3839733600616455,
+      "learning_rate": 2e-05,
+      "loss": 0.03918359,
+      "step": 22285
+    },
+    {
+      "epoch": 44.572,
+      "grad_norm": 1.479257345199585,
+      "learning_rate": 2e-05,
+      "loss": 0.06053314,
+      "step": 22286
+    },
+    {
+      "epoch": 44.574,
+      "grad_norm": 2.5502092838287354,
+      "learning_rate": 2e-05,
+      "loss": 0.07071359,
+      "step": 22287
+    },
+    {
+      "epoch": 44.576,
+      "grad_norm": 1.1668108701705933,
+      "learning_rate": 2e-05,
+      "loss": 0.02986339,
+      "step": 22288
+    },
+    {
+      "epoch": 44.578,
+      "grad_norm": 1.1986876726150513,
+      "learning_rate": 2e-05,
+      "loss": 0.05737267,
+      "step": 22289
+    },
+    {
+      "epoch": 44.58,
+      "grad_norm": 1.2797632217407227,
+      "learning_rate": 2e-05,
+      "loss": 0.05133417,
+      "step": 22290
+    },
+    {
+      "epoch": 44.582,
+      "grad_norm": 1.1211811304092407,
+      "learning_rate": 2e-05,
+      "loss": 0.04330084,
+      "step": 22291
+    },
+    {
+      "epoch": 44.584,
+      "grad_norm": 1.906973123550415,
+      "learning_rate": 2e-05,
+      "loss": 0.05037066,
+      "step": 22292
+    },
+    {
+      "epoch": 44.586,
+      "grad_norm": 1.128289818763733,
+      "learning_rate": 2e-05,
+      "loss": 0.03821924,
+      "step": 22293
+    },
+    {
+      "epoch": 44.588,
+      "grad_norm": 2.022193670272827,
+      "learning_rate": 2e-05,
+      "loss": 0.06020941,
+      "step": 22294
+    },
+    {
+      "epoch": 44.59,
+      "grad_norm": 1.26398503780365,
+      "learning_rate": 2e-05,
+      "loss": 0.04926804,
+      "step": 22295
+    },
+    {
+      "epoch": 44.592,
+      "grad_norm": 1.2855454683303833,
+      "learning_rate": 2e-05,
+      "loss": 0.05323242,
+      "step": 22296
+    },
+    {
+      "epoch": 44.594,
+      "grad_norm": 1.0625572204589844,
+      "learning_rate": 2e-05,
+      "loss": 0.03464797,
+      "step": 22297
+    },
+    {
+      "epoch": 44.596,
+      "grad_norm": 1.904098391532898,
+      "learning_rate": 2e-05,
+      "loss": 0.07098804,
+      "step": 22298
+    },
+    {
+      "epoch": 44.598,
+      "grad_norm": 1.219960331916809,
+      "learning_rate": 2e-05,
+      "loss": 0.04161097,
+      "step": 22299
+    },
+    {
+      "epoch": 44.6,
+      "grad_norm": 1.6203744411468506,
+      "learning_rate": 2e-05,
+      "loss": 0.0535806,
+      "step": 22300
+    },
+    {
+      "epoch": 44.602,
+      "grad_norm": 1.138048529624939,
+      "learning_rate": 2e-05,
+      "loss": 0.04342931,
+      "step": 22301
+    },
+    {
+      "epoch": 44.604,
+      "grad_norm": 1.2476096153259277,
+      "learning_rate": 2e-05,
+      "loss": 0.0449223,
+      "step": 22302
+    },
+    {
+      "epoch": 44.606,
+      "grad_norm": 2.6473846435546875,
+      "learning_rate": 2e-05,
+      "loss": 0.04433635,
+      "step": 22303
+    },
+    {
+      "epoch": 44.608,
+      "grad_norm": 1.2410426139831543,
+      "learning_rate": 2e-05,
+      "loss": 0.05194189,
+      "step": 22304
+    },
+    {
+      "epoch": 44.61,
+      "grad_norm": 1.4751123189926147,
+      "learning_rate": 2e-05,
+      "loss": 0.04747447,
+      "step": 22305
+    },
+    {
+      "epoch": 44.612,
+      "grad_norm": 1.1794182062149048,
+      "learning_rate": 2e-05,
+      "loss": 0.04860191,
+      "step": 22306
+    },
+    {
+      "epoch": 44.614,
+      "grad_norm": 0.9958474636077881,
+      "learning_rate": 2e-05,
+      "loss": 0.03155885,
+      "step": 22307
+    },
+    {
+      "epoch": 44.616,
+      "grad_norm": 1.1580699682235718,
+      "learning_rate": 2e-05,
+      "loss": 0.0375239,
+      "step": 22308
+    },
+    {
+      "epoch": 44.618,
+      "grad_norm": 1.1620581150054932,
+      "learning_rate": 2e-05,
+      "loss": 0.05345889,
+      "step": 22309
+    },
+    {
+      "epoch": 44.62,
+      "grad_norm": 1.1996610164642334,
+      "learning_rate": 2e-05,
+      "loss": 0.05827244,
+      "step": 22310
+    },
+    {
+      "epoch": 44.622,
+      "grad_norm": 1.862815260887146,
+      "learning_rate": 2e-05,
+      "loss": 0.04650091,
+      "step": 22311
+    },
+    {
+      "epoch": 44.624,
+      "grad_norm": 1.2667579650878906,
+      "learning_rate": 2e-05,
+      "loss": 0.0441907,
+      "step": 22312
+    },
+    {
+      "epoch": 44.626,
+      "grad_norm": 1.3493322134017944,
+      "learning_rate": 2e-05,
+      "loss": 0.05513781,
+      "step": 22313
+    },
+    {
+      "epoch": 44.628,
+      "grad_norm": 1.2868801355361938,
+      "learning_rate": 2e-05,
+      "loss": 0.05301888,
+      "step": 22314
+    },
+    {
+      "epoch": 44.63,
+      "grad_norm": 1.0114258527755737,
+      "learning_rate": 2e-05,
+      "loss": 0.04150274,
+      "step": 22315
+    },
+    {
+      "epoch": 44.632,
+      "grad_norm": 1.1847352981567383,
+      "learning_rate": 2e-05,
+      "loss": 0.04998568,
+      "step": 22316
+    },
+    {
+      "epoch": 44.634,
+      "grad_norm": 1.4043679237365723,
+      "learning_rate": 2e-05,
+      "loss": 0.04988774,
+      "step": 22317
+    },
+    {
+      "epoch": 44.636,
+      "grad_norm": 2.041414260864258,
+      "learning_rate": 2e-05,
+      "loss": 0.05052672,
+      "step": 22318
+    },
+    {
+      "epoch": 44.638,
+      "grad_norm": 2.7315549850463867,
+      "learning_rate": 2e-05,
+      "loss": 0.05014642,
+      "step": 22319
+    },
+    {
+      "epoch": 44.64,
+      "grad_norm": 1.9132074117660522,
+      "learning_rate": 2e-05,
+      "loss": 0.04718821,
+      "step": 22320
+    },
+    {
+      "epoch": 44.642,
+      "grad_norm": 1.4658045768737793,
+      "learning_rate": 2e-05,
+      "loss": 0.04350495,
+      "step": 22321
+    },
+    {
+      "epoch": 44.644,
+      "grad_norm": 1.4384865760803223,
+      "learning_rate": 2e-05,
+      "loss": 0.04404975,
+      "step": 22322
+    },
+    {
+      "epoch": 44.646,
+      "grad_norm": 1.4497653245925903,
+      "learning_rate": 2e-05,
+      "loss": 0.03452429,
+      "step": 22323
+    },
+    {
+      "epoch": 44.648,
+      "grad_norm": 1.3256078958511353,
+      "learning_rate": 2e-05,
+      "loss": 0.040176,
+      "step": 22324
+    },
+    {
+      "epoch": 44.65,
+      "grad_norm": 1.1343733072280884,
+      "learning_rate": 2e-05,
+      "loss": 0.03543989,
+      "step": 22325
+    },
+    {
+      "epoch": 44.652,
+      "grad_norm": 1.8382833003997803,
+      "learning_rate": 2e-05,
+      "loss": 0.05061209,
+      "step": 22326
+    },
+    {
+      "epoch": 44.654,
+      "grad_norm": 1.3723702430725098,
+      "learning_rate": 2e-05,
+      "loss": 0.04715881,
+      "step": 22327
+    },
+    {
+      "epoch": 44.656,
+      "grad_norm": 1.6342358589172363,
+      "learning_rate": 2e-05,
+      "loss": 0.05269512,
+      "step": 22328
+    },
+    {
+      "epoch": 44.658,
+      "grad_norm": 1.3743045330047607,
+      "learning_rate": 2e-05,
+      "loss": 0.04504273,
+      "step": 22329
+    },
+    {
+      "epoch": 44.66,
+      "grad_norm": 3.2658958435058594,
+      "learning_rate": 2e-05,
+      "loss": 0.05143912,
+      "step": 22330
+    },
+    {
+      "epoch": 44.662,
+      "grad_norm": 2.2157278060913086,
+      "learning_rate": 2e-05,
+      "loss": 0.05146747,
+      "step": 22331
+    },
+    {
+      "epoch": 44.664,
+      "grad_norm": 1.1454540491104126,
+      "learning_rate": 2e-05,
+      "loss": 0.04466172,
+      "step": 22332
+    },
+    {
+      "epoch": 44.666,
+      "grad_norm": 1.2237814664840698,
+      "learning_rate": 2e-05,
+      "loss": 0.05431961,
+      "step": 22333
+    },
+    {
+      "epoch": 44.668,
+      "grad_norm": 1.3983421325683594,
+      "learning_rate": 2e-05,
+      "loss": 0.03210323,
+      "step": 22334
+    },
+    {
+      "epoch": 44.67,
+      "grad_norm": 1.5525994300842285,
+      "learning_rate": 2e-05,
+      "loss": 0.06114915,
+      "step": 22335
+    },
+    {
+      "epoch": 44.672,
+      "grad_norm": 1.054101586341858,
+      "learning_rate": 2e-05,
+      "loss": 0.04592917,
+      "step": 22336
+    },
+    {
+      "epoch": 44.674,
+      "grad_norm": 1.3390069007873535,
+      "learning_rate": 2e-05,
+      "loss": 0.04546981,
+      "step": 22337
+    },
+    {
+      "epoch": 44.676,
+      "grad_norm": 1.3477840423583984,
+      "learning_rate": 2e-05,
+      "loss": 0.05513988,
+      "step": 22338
+    },
+    {
+      "epoch": 44.678,
+      "grad_norm": 1.9911638498306274,
+      "learning_rate": 2e-05,
+      "loss": 0.04411037,
+      "step": 22339
+    },
+    {
+      "epoch": 44.68,
+      "grad_norm": 1.5710328817367554,
+      "learning_rate": 2e-05,
+      "loss": 0.0609922,
+      "step": 22340
+    },
+    {
+      "epoch": 44.682,
+      "grad_norm": 1.218172550201416,
+      "learning_rate": 2e-05,
+      "loss": 0.05370003,
+      "step": 22341
+    },
+    {
+      "epoch": 44.684,
+      "grad_norm": 1.1348539590835571,
+      "learning_rate": 2e-05,
+      "loss": 0.0526722,
+      "step": 22342
+    },
+    {
+      "epoch": 44.686,
+      "grad_norm": 1.610276460647583,
+      "learning_rate": 2e-05,
+      "loss": 0.070089,
+      "step": 22343
+    },
+    {
+      "epoch": 44.688,
+      "grad_norm": 1.2752341032028198,
+      "learning_rate": 2e-05,
+      "loss": 0.04979321,
+      "step": 22344
+    },
+    {
+      "epoch": 44.69,
+      "grad_norm": 1.252686619758606,
+      "learning_rate": 2e-05,
+      "loss": 0.04036684,
+      "step": 22345
+    },
+    {
+      "epoch": 44.692,
+      "grad_norm": 1.023863673210144,
+      "learning_rate": 2e-05,
+      "loss": 0.02940369,
+      "step": 22346
+    },
+    {
+      "epoch": 44.694,
+      "grad_norm": 0.9316999316215515,
+      "learning_rate": 2e-05,
+      "loss": 0.03571773,
+      "step": 22347
+    },
+    {
+      "epoch": 44.696,
+      "grad_norm": 1.4448881149291992,
+      "learning_rate": 2e-05,
+      "loss": 0.06954427,
+      "step": 22348
+    },
+    {
+      "epoch": 44.698,
+      "grad_norm": 1.6992919445037842,
+      "learning_rate": 2e-05,
+      "loss": 0.04004553,
+      "step": 22349
+    },
+    {
+      "epoch": 44.7,
+      "grad_norm": 1.3580870628356934,
+      "learning_rate": 2e-05,
+      "loss": 0.04999983,
+      "step": 22350
+    },
+    {
+      "epoch": 44.702,
+      "grad_norm": 1.1199650764465332,
+      "learning_rate": 2e-05,
+      "loss": 0.04584752,
+      "step": 22351
+    },
+    {
+      "epoch": 44.704,
+      "grad_norm": 1.46734619140625,
+      "learning_rate": 2e-05,
+      "loss": 0.04816905,
+      "step": 22352
+    },
+    {
+      "epoch": 44.706,
+      "grad_norm": 1.2117671966552734,
+      "learning_rate": 2e-05,
+      "loss": 0.03972312,
+      "step": 22353
+    },
+    {
+      "epoch": 44.708,
+      "grad_norm": 1.2888193130493164,
+      "learning_rate": 2e-05,
+      "loss": 0.03945049,
+      "step": 22354
+    },
+    {
+      "epoch": 44.71,
+      "grad_norm": 1.204562783241272,
+      "learning_rate": 2e-05,
+      "loss": 0.04760649,
+      "step": 22355
+    },
+    {
+      "epoch": 44.712,
+      "grad_norm": 1.2271558046340942,
+      "learning_rate": 2e-05,
+      "loss": 0.04851292,
+      "step": 22356
+    },
+    {
+      "epoch": 44.714,
+      "grad_norm": 1.2417253255844116,
+      "learning_rate": 2e-05,
+      "loss": 0.05184454,
+      "step": 22357
+    },
+    {
+      "epoch": 44.716,
+      "grad_norm": 1.2915658950805664,
+      "learning_rate": 2e-05,
+      "loss": 0.0653154,
+      "step": 22358
+    },
+    {
+      "epoch": 44.718,
+      "grad_norm": 1.126898169517517,
+      "learning_rate": 2e-05,
+      "loss": 0.04059651,
+      "step": 22359
+    },
+    {
+      "epoch": 44.72,
+      "grad_norm": 1.2646050453186035,
+      "learning_rate": 2e-05,
+      "loss": 0.0460304,
+      "step": 22360
+    },
+    {
+      "epoch": 44.722,
+      "grad_norm": 1.2595983743667603,
+      "learning_rate": 2e-05,
+      "loss": 0.0446999,
+      "step": 22361
+    },
+    {
+      "epoch": 44.724,
+      "grad_norm": 1.0106483697891235,
+      "learning_rate": 2e-05,
+      "loss": 0.03425622,
+      "step": 22362
+    },
+    {
+      "epoch": 44.726,
+      "grad_norm": 1.0983266830444336,
+      "learning_rate": 2e-05,
+      "loss": 0.04440197,
+      "step": 22363
+    },
+    {
+      "epoch": 44.728,
+      "grad_norm": 1.0528661012649536,
+      "learning_rate": 2e-05,
+      "loss": 0.03264795,
+      "step": 22364
+    },
+    {
+      "epoch": 44.73,
+      "grad_norm": 1.485556721687317,
+      "learning_rate": 2e-05,
+      "loss": 0.04851057,
+      "step": 22365
+    },
+    {
+      "epoch": 44.732,
+      "grad_norm": 1.0484004020690918,
+      "learning_rate": 2e-05,
+      "loss": 0.03785881,
+      "step": 22366
+    },
+    {
+      "epoch": 44.734,
+      "grad_norm": 1.0544923543930054,
+      "learning_rate": 2e-05,
+      "loss": 0.04710227,
+      "step": 22367
+    },
+    {
+      "epoch": 44.736,
+      "grad_norm": 1.6099900007247925,
+      "learning_rate": 2e-05,
+      "loss": 0.0503059,
+      "step": 22368
+    },
+    {
+      "epoch": 44.738,
+      "grad_norm": 1.0800563097000122,
+      "learning_rate": 2e-05,
+      "loss": 0.03581601,
+      "step": 22369
+    },
+    {
+      "epoch": 44.74,
+      "grad_norm": 1.2417271137237549,
+      "learning_rate": 2e-05,
+      "loss": 0.05445302,
+      "step": 22370
+    },
+    {
+      "epoch": 44.742,
+      "grad_norm": 1.3776066303253174,
+      "learning_rate": 2e-05,
+      "loss": 0.05917332,
+      "step": 22371
+    },
+    {
+      "epoch": 44.744,
+      "grad_norm": 6.051969051361084,
+      "learning_rate": 2e-05,
+      "loss": 0.03186266,
+      "step": 22372
+    },
+    {
+      "epoch": 44.746,
+      "grad_norm": 1.193009614944458,
+      "learning_rate": 2e-05,
+      "loss": 0.04203957,
+      "step": 22373
+    },
+    {
+      "epoch": 44.748,
+      "grad_norm": 1.1934806108474731,
+      "learning_rate": 2e-05,
+      "loss": 0.05293836,
+      "step": 22374
+    },
+    {
+      "epoch": 44.75,
+      "grad_norm": 1.2843340635299683,
+      "learning_rate": 2e-05,
+      "loss": 0.05309843,
+      "step": 22375
+    },
+    {
+      "epoch": 44.752,
+      "grad_norm": 1.2187143564224243,
+      "learning_rate": 2e-05,
+      "loss": 0.05113508,
+      "step": 22376
+    },
+    {
+      "epoch": 44.754,
+      "grad_norm": 1.5703728199005127,
+      "learning_rate": 2e-05,
+      "loss": 0.06165824,
+      "step": 22377
+    },
+    {
+      "epoch": 44.756,
+      "grad_norm": 1.9462801218032837,
+      "learning_rate": 2e-05,
+      "loss": 0.06039472,
+      "step": 22378
+    },
+    {
+      "epoch": 44.758,
+      "grad_norm": 1.2913825511932373,
+      "learning_rate": 2e-05,
+      "loss": 0.04901668,
+      "step": 22379
+    },
+    {
+      "epoch": 44.76,
+      "grad_norm": 1.484749436378479,
+      "learning_rate": 2e-05,
+      "loss": 0.05032424,
+      "step": 22380
+    },
+    {
+      "epoch": 44.762,
+      "grad_norm": 1.2912980318069458,
+      "learning_rate": 2e-05,
+      "loss": 0.05285494,
+      "step": 22381
+    },
+    {
+      "epoch": 44.764,
+      "grad_norm": 1.471806526184082,
+      "learning_rate": 2e-05,
+      "loss": 0.05935736,
+      "step": 22382
+    },
+    {
+      "epoch": 44.766,
+      "grad_norm": 1.4022643566131592,
+      "learning_rate": 2e-05,
+      "loss": 0.06408676,
+      "step": 22383
+    },
+    {
+      "epoch": 44.768,
+      "grad_norm": 1.3989965915679932,
+      "learning_rate": 2e-05,
+      "loss": 0.0560586,
+      "step": 22384
+    },
+    {
+      "epoch": 44.77,
+      "grad_norm": 1.7269079685211182,
+      "learning_rate": 2e-05,
+      "loss": 0.0490585,
+      "step": 22385
+    },
+    {
+      "epoch": 44.772,
+      "grad_norm": 2.0063679218292236,
+      "learning_rate": 2e-05,
+      "loss": 0.06107655,
+      "step": 22386
+    },
+    {
+      "epoch": 44.774,
+      "grad_norm": 1.0884474515914917,
+      "learning_rate": 2e-05,
+      "loss": 0.04211147,
+      "step": 22387
+    },
+    {
+      "epoch": 44.776,
+      "grad_norm": 1.1166996955871582,
+      "learning_rate": 2e-05,
+      "loss": 0.04605059,
+      "step": 22388
+    },
+    {
+      "epoch": 44.778,
+      "grad_norm": 1.3494291305541992,
+      "learning_rate": 2e-05,
+      "loss": 0.05825819,
+      "step": 22389
+    },
+    {
+      "epoch": 44.78,
+      "grad_norm": 1.4325668811798096,
+      "learning_rate": 2e-05,
+      "loss": 0.0605996,
+      "step": 22390
+    },
+    {
+      "epoch": 44.782,
+      "grad_norm": 1.1748969554901123,
+      "learning_rate": 2e-05,
+      "loss": 0.03148333,
+      "step": 22391
+    },
+    {
+      "epoch": 44.784,
+      "grad_norm": 1.8170502185821533,
+      "learning_rate": 2e-05,
+      "loss": 0.06581346,
+      "step": 22392
+    },
+    {
+      "epoch": 44.786,
+      "grad_norm": 1.3329365253448486,
+      "learning_rate": 2e-05,
+      "loss": 0.05404956,
+      "step": 22393
+    },
+    {
+      "epoch": 44.788,
+      "grad_norm": 1.3301454782485962,
+      "learning_rate": 2e-05,
+      "loss": 0.04755008,
+      "step": 22394
+    },
+    {
+      "epoch": 44.79,
+      "grad_norm": 1.2659374475479126,
+      "learning_rate": 2e-05,
+      "loss": 0.04017854,
+      "step": 22395
+    },
+    {
+      "epoch": 44.792,
+      "grad_norm": 1.4887542724609375,
+      "learning_rate": 2e-05,
+      "loss": 0.04862245,
+      "step": 22396
+    },
+    {
+      "epoch": 44.794,
+      "grad_norm": 1.8280566930770874,
+      "learning_rate": 2e-05,
+      "loss": 0.05217675,
+      "step": 22397
+    },
+    {
+      "epoch": 44.796,
+      "grad_norm": 1.307554841041565,
+      "learning_rate": 2e-05,
+      "loss": 0.051239,
+      "step": 22398
+    },
+    {
+      "epoch": 44.798,
+      "grad_norm": 1.391467571258545,
+      "learning_rate": 2e-05,
+      "loss": 0.041818,
+      "step": 22399
+    },
+    {
+      "epoch": 44.8,
+      "grad_norm": 1.2929948568344116,
+      "learning_rate": 2e-05,
+      "loss": 0.03885982,
+      "step": 22400
+    },
+    {
+      "epoch": 44.802,
+      "grad_norm": 0.8741863965988159,
+      "learning_rate": 2e-05,
+      "loss": 0.02871921,
+      "step": 22401
+    },
+    {
+      "epoch": 44.804,
+      "grad_norm": 1.1634330749511719,
+      "learning_rate": 2e-05,
+      "loss": 0.05327924,
+      "step": 22402
+    },
+    {
+      "epoch": 44.806,
+      "grad_norm": 1.2257028818130493,
+      "learning_rate": 2e-05,
+      "loss": 0.04897387,
+      "step": 22403
+    },
+    {
+      "epoch": 44.808,
+      "grad_norm": 1.1557226181030273,
+      "learning_rate": 2e-05,
+      "loss": 0.04159442,
+      "step": 22404
+    },
+    {
+      "epoch": 44.81,
+      "grad_norm": 2.137556791305542,
+      "learning_rate": 2e-05,
+      "loss": 0.05608717,
+      "step": 22405
+    },
+    {
+      "epoch": 44.812,
+      "grad_norm": 1.224327802658081,
+      "learning_rate": 2e-05,
+      "loss": 0.04210414,
+      "step": 22406
+    },
+    {
+      "epoch": 44.814,
+      "grad_norm": 1.342980146408081,
+      "learning_rate": 2e-05,
+      "loss": 0.06127755,
+      "step": 22407
+    },
+    {
+      "epoch": 44.816,
+      "grad_norm": 1.5353951454162598,
+      "learning_rate": 2e-05,
+      "loss": 0.06302527,
+      "step": 22408
+    },
+    {
+      "epoch": 44.818,
+      "grad_norm": 1.3694101572036743,
+      "learning_rate": 2e-05,
+      "loss": 0.04164361,
+      "step": 22409
+    },
+    {
+      "epoch": 44.82,
+      "grad_norm": 1.6121137142181396,
+      "learning_rate": 2e-05,
+      "loss": 0.05497941,
+      "step": 22410
+    },
+    {
+      "epoch": 44.822,
+      "grad_norm": 1.375234842300415,
+      "learning_rate": 2e-05,
+      "loss": 0.05874093,
+      "step": 22411
+    },
+    {
+      "epoch": 44.824,
+      "grad_norm": 1.243156909942627,
+      "learning_rate": 2e-05,
+      "loss": 0.05009591,
+      "step": 22412
+    },
+    {
+      "epoch": 44.826,
+      "grad_norm": 1.365031123161316,
+      "learning_rate": 2e-05,
+      "loss": 0.04909578,
+      "step": 22413
+    },
+    {
+      "epoch": 44.828,
+      "grad_norm": 0.9832285046577454,
+      "learning_rate": 2e-05,
+      "loss": 0.0287962,
+      "step": 22414
+    },
+    {
+      "epoch": 44.83,
+      "grad_norm": 1.187656044960022,
+      "learning_rate": 2e-05,
+      "loss": 0.04923886,
+      "step": 22415
+    },
+    {
+      "epoch": 44.832,
+      "grad_norm": 1.2411973476409912,
+      "learning_rate": 2e-05,
+      "loss": 0.05921061,
+      "step": 22416
+    },
+    {
+      "epoch": 44.834,
+      "grad_norm": 1.7300316095352173,
+      "learning_rate": 2e-05,
+      "loss": 0.0390442,
+      "step": 22417
+    },
+    {
+      "epoch": 44.836,
+      "grad_norm": 1.0223551988601685,
+      "learning_rate": 2e-05,
+      "loss": 0.03893992,
+      "step": 22418
+    },
+    {
+      "epoch": 44.838,
+      "grad_norm": 1.9662129878997803,
+      "learning_rate": 2e-05,
+      "loss": 0.03231378,
+      "step": 22419
+    },
+    {
+      "epoch": 44.84,
+      "grad_norm": 1.0947517156600952,
+      "learning_rate": 2e-05,
+      "loss": 0.03922854,
+      "step": 22420
+    },
+    {
+      "epoch": 44.842,
+      "grad_norm": 1.6704392433166504,
+      "learning_rate": 2e-05,
+      "loss": 0.04813308,
+      "step": 22421
+    },
+    {
+      "epoch": 44.844,
+      "grad_norm": 1.2903733253479004,
+      "learning_rate": 2e-05,
+      "loss": 0.04857474,
+      "step": 22422
+    },
+    {
+      "epoch": 44.846,
+      "grad_norm": 1.159749984741211,
+      "learning_rate": 2e-05,
+      "loss": 0.04078946,
+      "step": 22423
+    },
+    {
+      "epoch": 44.848,
+      "grad_norm": 1.364219307899475,
+      "learning_rate": 2e-05,
+      "loss": 0.04483482,
+      "step": 22424
+    },
+    {
+      "epoch": 44.85,
+      "grad_norm": 1.2046773433685303,
+      "learning_rate": 2e-05,
+      "loss": 0.05179583,
+      "step": 22425
+    },
+    {
+      "epoch": 44.852,
+      "grad_norm": 1.419978380203247,
+      "learning_rate": 2e-05,
+      "loss": 0.0563489,
+      "step": 22426
+    },
+    {
+      "epoch": 44.854,
+      "grad_norm": 1.4151443243026733,
+      "learning_rate": 2e-05,
+      "loss": 0.05637079,
+      "step": 22427
+    },
+    {
+      "epoch": 44.856,
+      "grad_norm": 1.2609163522720337,
+      "learning_rate": 2e-05,
+      "loss": 0.05354507,
+      "step": 22428
+    },
+    {
+      "epoch": 44.858,
+      "grad_norm": 1.2031396627426147,
+      "learning_rate": 2e-05,
+      "loss": 0.04809187,
+      "step": 22429
+    },
+    {
+      "epoch": 44.86,
+      "grad_norm": 1.2825238704681396,
+      "learning_rate": 2e-05,
+      "loss": 0.05121071,
+      "step": 22430
+    },
+    {
+      "epoch": 44.862,
+      "grad_norm": 1.0715117454528809,
+      "learning_rate": 2e-05,
+      "loss": 0.03546477,
+      "step": 22431
+    },
+    {
+      "epoch": 44.864,
+      "grad_norm": 1.407547116279602,
+      "learning_rate": 2e-05,
+      "loss": 0.05819602,
+      "step": 22432
+    },
+    {
+      "epoch": 44.866,
+      "grad_norm": 1.705998182296753,
+      "learning_rate": 2e-05,
+      "loss": 0.05690518,
+      "step": 22433
+    },
+    {
+      "epoch": 44.868,
+      "grad_norm": 1.3436830043792725,
+      "learning_rate": 2e-05,
+      "loss": 0.04754122,
+      "step": 22434
+    },
+    {
+      "epoch": 44.87,
+      "grad_norm": 1.3858736753463745,
+      "learning_rate": 2e-05,
+      "loss": 0.06503775,
+      "step": 22435
+    },
+    {
+      "epoch": 44.872,
+      "grad_norm": 1.100571632385254,
+      "learning_rate": 2e-05,
+      "loss": 0.04838911,
+      "step": 22436
+    },
+    {
+      "epoch": 44.874,
+      "grad_norm": 1.277199387550354,
+      "learning_rate": 2e-05,
+      "loss": 0.0342557,
+      "step": 22437
+    },
+    {
+      "epoch": 44.876,
+      "grad_norm": 1.04167640209198,
+      "learning_rate": 2e-05,
+      "loss": 0.03733697,
+      "step": 22438
+    },
+    {
+      "epoch": 44.878,
+      "grad_norm": 1.1401691436767578,
+      "learning_rate": 2e-05,
+      "loss": 0.04081557,
+      "step": 22439
+    },
+    {
+      "epoch": 44.88,
+      "grad_norm": 1.1583184003829956,
+      "learning_rate": 2e-05,
+      "loss": 0.04236028,
+      "step": 22440
+    },
+    {
+      "epoch": 44.882,
+      "grad_norm": 1.224388837814331,
+      "learning_rate": 2e-05,
+      "loss": 0.04657396,
+      "step": 22441
+    },
+    {
+      "epoch": 44.884,
+      "grad_norm": 1.3904510736465454,
+      "learning_rate": 2e-05,
+      "loss": 0.05583341,
+      "step": 22442
+    },
+    {
+      "epoch": 44.886,
+      "grad_norm": 1.3104639053344727,
+      "learning_rate": 2e-05,
+      "loss": 0.04422941,
+      "step": 22443
+    },
+    {
+      "epoch": 44.888,
+      "grad_norm": 1.4339663982391357,
+      "learning_rate": 2e-05,
+      "loss": 0.05825556,
+      "step": 22444
+    },
+    {
+      "epoch": 44.89,
+      "grad_norm": 2.071310520172119,
+      "learning_rate": 2e-05,
+      "loss": 0.04446717,
+      "step": 22445
+    },
+    {
+      "epoch": 44.892,
+      "grad_norm": 1.579842448234558,
+      "learning_rate": 2e-05,
+      "loss": 0.06342889,
+      "step": 22446
+    },
+    {
+      "epoch": 44.894,
+      "grad_norm": 1.1288223266601562,
+      "learning_rate": 2e-05,
+      "loss": 0.04180889,
+      "step": 22447
+    },
+    {
+      "epoch": 44.896,
+      "grad_norm": 0.9220855832099915,
+      "learning_rate": 2e-05,
+      "loss": 0.03075635,
+      "step": 22448
+    },
+    {
+      "epoch": 44.898,
+      "grad_norm": 1.1610661745071411,
+      "learning_rate": 2e-05,
+      "loss": 0.04427385,
+      "step": 22449
+    },
+    {
+      "epoch": 44.9,
+      "grad_norm": 1.0175820589065552,
+      "learning_rate": 2e-05,
+      "loss": 0.03336541,
+      "step": 22450
+    },
+    {
+      "epoch": 44.902,
+      "grad_norm": 1.2916738986968994,
+      "learning_rate": 2e-05,
+      "loss": 0.04845192,
+      "step": 22451
+    },
+    {
+      "epoch": 44.904,
+      "grad_norm": 1.0840328931808472,
+      "learning_rate": 2e-05,
+      "loss": 0.03961993,
+      "step": 22452
+    },
+    {
+      "epoch": 44.906,
+      "grad_norm": 1.168332815170288,
+      "learning_rate": 2e-05,
+      "loss": 0.04420353,
+      "step": 22453
+    },
+    {
+      "epoch": 44.908,
+      "grad_norm": 1.2330690622329712,
+      "learning_rate": 2e-05,
+      "loss": 0.046922,
+      "step": 22454
+    },
+    {
+      "epoch": 44.91,
+      "grad_norm": 1.777421236038208,
+      "learning_rate": 2e-05,
+      "loss": 0.06229702,
+      "step": 22455
+    },
+    {
+      "epoch": 44.912,
+      "grad_norm": 1.3001389503479004,
+      "learning_rate": 2e-05,
+      "loss": 0.05773286,
+      "step": 22456
+    },
+    {
+      "epoch": 44.914,
+      "grad_norm": 1.3984302282333374,
+      "learning_rate": 2e-05,
+      "loss": 0.06564485,
+      "step": 22457
+    },
+    {
+      "epoch": 44.916,
+      "grad_norm": 1.200603723526001,
+      "learning_rate": 2e-05,
+      "loss": 0.04599864,
+      "step": 22458
+    },
+    {
+      "epoch": 44.918,
+      "grad_norm": 1.0556275844573975,
+      "learning_rate": 2e-05,
+      "loss": 0.03204244,
+      "step": 22459
+    },
+    {
+      "epoch": 44.92,
+      "grad_norm": 1.3520619869232178,
+      "learning_rate": 2e-05,
+      "loss": 0.05546412,
+      "step": 22460
+    },
+    {
+      "epoch": 44.922,
+      "grad_norm": 1.3469730615615845,
+      "learning_rate": 2e-05,
+      "loss": 0.04959213,
+      "step": 22461
+    },
+    {
+      "epoch": 44.924,
+      "grad_norm": 1.6626015901565552,
+      "learning_rate": 2e-05,
+      "loss": 0.06908607,
+      "step": 22462
+    },
+    {
+      "epoch": 44.926,
+      "grad_norm": 1.079907774925232,
+      "learning_rate": 2e-05,
+      "loss": 0.03825868,
+      "step": 22463
+    },
+    {
+      "epoch": 44.928,
+      "grad_norm": 1.4360984563827515,
+      "learning_rate": 2e-05,
+      "loss": 0.06819391,
+      "step": 22464
+    },
+    {
+      "epoch": 44.93,
+      "grad_norm": 1.118546485900879,
+      "learning_rate": 2e-05,
+      "loss": 0.03266671,
+      "step": 22465
+    },
+    {
+      "epoch": 44.932,
+      "grad_norm": 1.7453476190567017,
+      "learning_rate": 2e-05,
+      "loss": 0.06611872,
+      "step": 22466
+    },
+    {
+      "epoch": 44.934,
+      "grad_norm": 1.403899073600769,
+      "learning_rate": 2e-05,
+      "loss": 0.05418711,
+      "step": 22467
+    },
+    {
+      "epoch": 44.936,
+      "grad_norm": 4.624436378479004,
+      "learning_rate": 2e-05,
+      "loss": 0.07199771,
+      "step": 22468
+    },
+    {
+      "epoch": 44.938,
+      "grad_norm": 1.1568007469177246,
+      "learning_rate": 2e-05,
+      "loss": 0.05059115,
+      "step": 22469
+    },
+    {
+      "epoch": 44.94,
+      "grad_norm": 1.137229561805725,
+      "learning_rate": 2e-05,
+      "loss": 0.03843639,
+      "step": 22470
+    },
+    {
+      "epoch": 44.942,
+      "grad_norm": 1.4798380136489868,
+      "learning_rate": 2e-05,
+      "loss": 0.04668789,
+      "step": 22471
+    },
+    {
+      "epoch": 44.944,
+      "grad_norm": 1.986149549484253,
+      "learning_rate": 2e-05,
+      "loss": 0.03478555,
+      "step": 22472
+    },
+    {
+      "epoch": 44.946,
+      "grad_norm": 1.0867761373519897,
+      "learning_rate": 2e-05,
+      "loss": 0.03376482,
+      "step": 22473
+    },
+    {
+      "epoch": 44.948,
+      "grad_norm": 1.3647146224975586,
+      "learning_rate": 2e-05,
+      "loss": 0.06791535,
+      "step": 22474
+    },
+    {
+      "epoch": 44.95,
+      "grad_norm": 1.3892532587051392,
+      "learning_rate": 2e-05,
+      "loss": 0.05850136,
+      "step": 22475
+    },
+    {
+      "epoch": 44.952,
+      "grad_norm": 1.2366359233856201,
+      "learning_rate": 2e-05,
+      "loss": 0.04139677,
+      "step": 22476
+    },
+    {
+      "epoch": 44.954,
+      "grad_norm": 2.012018918991089,
+      "learning_rate": 2e-05,
+      "loss": 0.05484272,
+      "step": 22477
+    },
+    {
+      "epoch": 44.956,
+      "grad_norm": 1.2772232294082642,
+      "learning_rate": 2e-05,
+      "loss": 0.04107056,
+      "step": 22478
+    },
+    {
+      "epoch": 44.958,
+      "grad_norm": 1.0646635293960571,
+      "learning_rate": 2e-05,
+      "loss": 0.04107633,
+      "step": 22479
+    },
+    {
+      "epoch": 44.96,
+      "grad_norm": 3.412080764770508,
+      "learning_rate": 2e-05,
+      "loss": 0.06652329,
+      "step": 22480
+    },
+    {
+      "epoch": 44.962,
+      "grad_norm": 1.39540696144104,
+      "learning_rate": 2e-05,
+      "loss": 0.04058058,
+      "step": 22481
+    },
+    {
+      "epoch": 44.964,
+      "grad_norm": 1.372216820716858,
+      "learning_rate": 2e-05,
+      "loss": 0.05258819,
+      "step": 22482
+    },
+    {
+      "epoch": 44.966,
+      "grad_norm": 1.1605424880981445,
+      "learning_rate": 2e-05,
+      "loss": 0.05386285,
+      "step": 22483
+    },
+    {
+      "epoch": 44.968,
+      "grad_norm": 1.0500538349151611,
+      "learning_rate": 2e-05,
+      "loss": 0.03223278,
+      "step": 22484
+    },
+    {
+      "epoch": 44.97,
+      "grad_norm": 1.171493411064148,
+      "learning_rate": 2e-05,
+      "loss": 0.0475014,
+      "step": 22485
+    },
+    {
+      "epoch": 44.972,
+      "grad_norm": 1.1889984607696533,
+      "learning_rate": 2e-05,
+      "loss": 0.03335759,
+      "step": 22486
+    },
+    {
+      "epoch": 44.974,
+      "grad_norm": 1.488581895828247,
+      "learning_rate": 2e-05,
+      "loss": 0.04842,
+      "step": 22487
+    },
+    {
+      "epoch": 44.976,
+      "grad_norm": 1.2926756143569946,
+      "learning_rate": 2e-05,
+      "loss": 0.04585892,
+      "step": 22488
+    },
+    {
+      "epoch": 44.978,
+      "grad_norm": 1.0679482221603394,
+      "learning_rate": 2e-05,
+      "loss": 0.03685507,
+      "step": 22489
+    },
+    {
+      "epoch": 44.98,
+      "grad_norm": 1.2621574401855469,
+      "learning_rate": 2e-05,
+      "loss": 0.05050556,
+      "step": 22490
+    },
+    {
+      "epoch": 44.982,
+      "grad_norm": 1.104034185409546,
+      "learning_rate": 2e-05,
+      "loss": 0.03997359,
+      "step": 22491
+    },
+    {
+      "epoch": 44.984,
+      "grad_norm": 1.1348416805267334,
+      "learning_rate": 2e-05,
+      "loss": 0.03981521,
+      "step": 22492
+    },
+    {
+      "epoch": 44.986,
+      "grad_norm": 1.5478214025497437,
+      "learning_rate": 2e-05,
+      "loss": 0.06456232,
+      "step": 22493
+    },
+    {
+      "epoch": 44.988,
+      "grad_norm": 1.2740987539291382,
+      "learning_rate": 2e-05,
+      "loss": 0.04914555,
+      "step": 22494
+    },
+    {
+      "epoch": 44.99,
+      "grad_norm": 1.046753168106079,
+      "learning_rate": 2e-05,
+      "loss": 0.04628645,
+      "step": 22495
+    },
+    {
+      "epoch": 44.992,
+      "grad_norm": 2.204502820968628,
+      "learning_rate": 2e-05,
+      "loss": 0.04956776,
+      "step": 22496
+    },
+    {
+      "epoch": 44.994,
+      "grad_norm": 1.2961649894714355,
+      "learning_rate": 2e-05,
+      "loss": 0.04877903,
+      "step": 22497
+    },
+    {
+      "epoch": 44.996,
+      "grad_norm": 1.6191555261611938,
+      "learning_rate": 2e-05,
+      "loss": 0.05243221,
+      "step": 22498
+    },
+    {
+      "epoch": 44.998,
+      "grad_norm": 1.054588794708252,
+      "learning_rate": 2e-05,
+      "loss": 0.04243375,
+      "step": 22499
+    },
+    {
+      "epoch": 45.0,
+      "grad_norm": 1.2328218221664429,
+      "learning_rate": 2e-05,
+      "loss": 0.04469647,
+      "step": 22500
+    },
+    {
+      "epoch": 45.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.998,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9860279441117764,
+        "Equal_1": 1.0,
+        "Equal_2": 0.9860279441117764,
+        "Equal_3": 0.9960079840319361,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 0.998003992015968,
+        "Parallel_1": 0.9959919839679359,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.996,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.992,
+        "Perpendicular_3": 0.8877755511022044,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 1.0,
+        "PointLiesOnCircle_3": 0.99,
+        "PointLiesOnLine_1": 1.0,
+        "PointLiesOnLine_2": 0.9939879759519038,
+        "PointLiesOnLine_3": 0.9820359281437125
+      },
+      "eval_runtime": 321.1082,
+      "eval_samples_per_second": 32.699,
+      "eval_steps_per_second": 0.654,
+      "step": 22500
+    },
+    {
+      "epoch": 45.002,
+      "grad_norm": 1.2806766033172607,
+      "learning_rate": 2e-05,
+      "loss": 0.03703031,
+      "step": 22501
+    },
+    {
+      "epoch": 45.004,
+      "grad_norm": 1.2303308248519897,
+      "learning_rate": 2e-05,
+      "loss": 0.0578556,
+      "step": 22502
+    },
+    {
+      "epoch": 45.006,
+      "grad_norm": 1.2770564556121826,
+      "learning_rate": 2e-05,
+      "loss": 0.0442674,
+      "step": 22503
+    },
+    {
+      "epoch": 45.008,
+      "grad_norm": 1.6194312572479248,
+      "learning_rate": 2e-05,
+      "loss": 0.06683954,
+      "step": 22504
+    },
+    {
+      "epoch": 45.01,
+      "grad_norm": 1.137593388557434,
+      "learning_rate": 2e-05,
+      "loss": 0.03966604,
+      "step": 22505
+    },
+    {
+      "epoch": 45.012,
+      "grad_norm": 3.7040164470672607,
+      "learning_rate": 2e-05,
+      "loss": 0.05335582,
+      "step": 22506
+    },
+    {
+      "epoch": 45.014,
+      "grad_norm": 1.019702672958374,
+      "learning_rate": 2e-05,
+      "loss": 0.039069,
+      "step": 22507
+    },
+    {
+      "epoch": 45.016,
+      "grad_norm": 1.11601984500885,
+      "learning_rate": 2e-05,
+      "loss": 0.03871304,
+      "step": 22508
+    },
+    {
+      "epoch": 45.018,
+      "grad_norm": 1.1694669723510742,
+      "learning_rate": 2e-05,
+      "loss": 0.04809815,
+      "step": 22509
+    },
+    {
+      "epoch": 45.02,
+      "grad_norm": 1.0752638578414917,
+      "learning_rate": 2e-05,
+      "loss": 0.03829714,
+      "step": 22510
+    },
+    {
+      "epoch": 45.022,
+      "grad_norm": 1.298243522644043,
+      "learning_rate": 2e-05,
+      "loss": 0.03895707,
+      "step": 22511
+    },
+    {
+      "epoch": 45.024,
+      "grad_norm": 0.9618425965309143,
+      "learning_rate": 2e-05,
+      "loss": 0.03141353,
+      "step": 22512
+    },
+    {
+      "epoch": 45.026,
+      "grad_norm": 1.0746543407440186,
+      "learning_rate": 2e-05,
+      "loss": 0.03593572,
+      "step": 22513
+    },
+    {
+      "epoch": 45.028,
+      "grad_norm": 1.0014342069625854,
+      "learning_rate": 2e-05,
+      "loss": 0.0362047,
+      "step": 22514
+    },
+    {
+      "epoch": 45.03,
+      "grad_norm": 1.230353593826294,
+      "learning_rate": 2e-05,
+      "loss": 0.04621077,
+      "step": 22515
+    },
+    {
+      "epoch": 45.032,
+      "grad_norm": 1.305424690246582,
+      "learning_rate": 2e-05,
+      "loss": 0.06496035,
+      "step": 22516
+    },
+    {
+      "epoch": 45.034,
+      "grad_norm": 0.9447730183601379,
+      "learning_rate": 2e-05,
+      "loss": 0.02817334,
+      "step": 22517
+    },
+    {
+      "epoch": 45.036,
+      "grad_norm": 1.625741958618164,
+      "learning_rate": 2e-05,
+      "loss": 0.06509583,
+      "step": 22518
+    },
+    {
+      "epoch": 45.038,
+      "grad_norm": 1.1020500659942627,
+      "learning_rate": 2e-05,
+      "loss": 0.04046689,
+      "step": 22519
+    },
+    {
+      "epoch": 45.04,
+      "grad_norm": 1.0736149549484253,
+      "learning_rate": 2e-05,
+      "loss": 0.04837882,
+      "step": 22520
+    },
+    {
+      "epoch": 45.042,
+      "grad_norm": 1.7894287109375,
+      "learning_rate": 2e-05,
+      "loss": 0.04507787,
+      "step": 22521
+    },
+    {
+      "epoch": 45.044,
+      "grad_norm": 0.9819255471229553,
+      "learning_rate": 2e-05,
+      "loss": 0.02709861,
+      "step": 22522
+    },
+    {
+      "epoch": 45.046,
+      "grad_norm": 1.2546124458312988,
+      "learning_rate": 2e-05,
+      "loss": 0.05152019,
+      "step": 22523
+    },
+    {
+      "epoch": 45.048,
+      "grad_norm": 1.2507061958312988,
+      "learning_rate": 2e-05,
+      "loss": 0.0508112,
+      "step": 22524
+    },
+    {
+      "epoch": 45.05,
+      "grad_norm": 1.0991570949554443,
+      "learning_rate": 2e-05,
+      "loss": 0.04952764,
+      "step": 22525
+    },
+    {
+      "epoch": 45.052,
+      "grad_norm": 0.895489513874054,
+      "learning_rate": 2e-05,
+      "loss": 0.02615194,
+      "step": 22526
+    },
+    {
+      "epoch": 45.054,
+      "grad_norm": 1.2041149139404297,
+      "learning_rate": 2e-05,
+      "loss": 0.05550298,
+      "step": 22527
+    },
+    {
+      "epoch": 45.056,
+      "grad_norm": 1.2021666765213013,
+      "learning_rate": 2e-05,
+      "loss": 0.0547464,
+      "step": 22528
+    },
+    {
+      "epoch": 45.058,
+      "grad_norm": 1.1511567831039429,
+      "learning_rate": 2e-05,
+      "loss": 0.04542843,
+      "step": 22529
+    },
+    {
+      "epoch": 45.06,
+      "grad_norm": 1.7193586826324463,
+      "learning_rate": 2e-05,
+      "loss": 0.06118374,
+      "step": 22530
+    },
+    {
+      "epoch": 45.062,
+      "grad_norm": 1.1019095182418823,
+      "learning_rate": 2e-05,
+      "loss": 0.04117382,
+      "step": 22531
+    },
+    {
+      "epoch": 45.064,
+      "grad_norm": 0.9580153822898865,
+      "learning_rate": 2e-05,
+      "loss": 0.04075727,
+      "step": 22532
+    },
+    {
+      "epoch": 45.066,
+      "grad_norm": 1.383394718170166,
+      "learning_rate": 2e-05,
+      "loss": 0.05107236,
+      "step": 22533
+    },
+    {
+      "epoch": 45.068,
+      "grad_norm": 1.2198097705841064,
+      "learning_rate": 2e-05,
+      "loss": 0.05111481,
+      "step": 22534
+    },
+    {
+      "epoch": 45.07,
+      "grad_norm": 1.1172356605529785,
+      "learning_rate": 2e-05,
+      "loss": 0.04869189,
+      "step": 22535
+    },
+    {
+      "epoch": 45.072,
+      "grad_norm": 1.1854000091552734,
+      "learning_rate": 2e-05,
+      "loss": 0.05265629,
+      "step": 22536
+    },
+    {
+      "epoch": 45.074,
+      "grad_norm": 1.2538152933120728,
+      "learning_rate": 2e-05,
+      "loss": 0.05185087,
+      "step": 22537
+    },
+    {
+      "epoch": 45.076,
+      "grad_norm": 1.1875842809677124,
+      "learning_rate": 2e-05,
+      "loss": 0.05062048,
+      "step": 22538
+    },
+    {
+      "epoch": 45.078,
+      "grad_norm": 1.1964948177337646,
+      "learning_rate": 2e-05,
+      "loss": 0.04546596,
+      "step": 22539
+    },
+    {
+      "epoch": 45.08,
+      "grad_norm": 0.9874855279922485,
+      "learning_rate": 2e-05,
+      "loss": 0.03447803,
+      "step": 22540
+    },
+    {
+      "epoch": 45.082,
+      "grad_norm": 1.2589225769042969,
+      "learning_rate": 2e-05,
+      "loss": 0.052671,
+      "step": 22541
+    },
+    {
+      "epoch": 45.084,
+      "grad_norm": 1.2300832271575928,
+      "learning_rate": 2e-05,
+      "loss": 0.05471198,
+      "step": 22542
+    },
+    {
+      "epoch": 45.086,
+      "grad_norm": 1.014027714729309,
+      "learning_rate": 2e-05,
+      "loss": 0.03973835,
+      "step": 22543
+    },
+    {
+      "epoch": 45.088,
+      "grad_norm": 1.0800725221633911,
+      "learning_rate": 2e-05,
+      "loss": 0.0450437,
+      "step": 22544
+    },
+    {
+      "epoch": 45.09,
+      "grad_norm": 1.2067081928253174,
+      "learning_rate": 2e-05,
+      "loss": 0.04083369,
+      "step": 22545
+    },
+    {
+      "epoch": 45.092,
+      "grad_norm": 1.1328651905059814,
+      "learning_rate": 2e-05,
+      "loss": 0.05992271,
+      "step": 22546
+    },
+    {
+      "epoch": 45.094,
+      "grad_norm": 1.180773138999939,
+      "learning_rate": 2e-05,
+      "loss": 0.04769462,
+      "step": 22547
+    },
+    {
+      "epoch": 45.096,
+      "grad_norm": 1.1579021215438843,
+      "learning_rate": 2e-05,
+      "loss": 0.05516424,
+      "step": 22548
+    },
+    {
+      "epoch": 45.098,
+      "grad_norm": 1.7966943979263306,
+      "learning_rate": 2e-05,
+      "loss": 0.0489285,
+      "step": 22549
+    },
+    {
+      "epoch": 45.1,
+      "grad_norm": 1.2300606966018677,
+      "learning_rate": 2e-05,
+      "loss": 0.05087769,
+      "step": 22550
+    },
+    {
+      "epoch": 45.102,
+      "grad_norm": 1.3096156120300293,
+      "learning_rate": 2e-05,
+      "loss": 0.04192181,
+      "step": 22551
+    },
+    {
+      "epoch": 45.104,
+      "grad_norm": 1.259109377861023,
+      "learning_rate": 2e-05,
+      "loss": 0.05914876,
+      "step": 22552
+    },
+    {
+      "epoch": 45.106,
+      "grad_norm": 0.9652984142303467,
+      "learning_rate": 2e-05,
+      "loss": 0.03940653,
+      "step": 22553
+    },
+    {
+      "epoch": 45.108,
+      "grad_norm": 1.1099121570587158,
+      "learning_rate": 2e-05,
+      "loss": 0.05075653,
+      "step": 22554
+    },
+    {
+      "epoch": 45.11,
+      "grad_norm": 1.2321460247039795,
+      "learning_rate": 2e-05,
+      "loss": 0.05328397,
+      "step": 22555
+    },
+    {
+      "epoch": 45.112,
+      "grad_norm": 1.3149770498275757,
+      "learning_rate": 2e-05,
+      "loss": 0.0476843,
+      "step": 22556
+    },
+    {
+      "epoch": 45.114,
+      "grad_norm": 0.9816151261329651,
+      "learning_rate": 2e-05,
+      "loss": 0.02892705,
+      "step": 22557
+    },
+    {
+      "epoch": 45.116,
+      "grad_norm": 1.0387154817581177,
+      "learning_rate": 2e-05,
+      "loss": 0.029668,
+      "step": 22558
+    },
+    {
+      "epoch": 45.118,
+      "grad_norm": 1.4174152612686157,
+      "learning_rate": 2e-05,
+      "loss": 0.03911356,
+      "step": 22559
+    },
+    {
+      "epoch": 45.12,
+      "grad_norm": 1.172924518585205,
+      "learning_rate": 2e-05,
+      "loss": 0.04783778,
+      "step": 22560
+    },
+    {
+      "epoch": 45.122,
+      "grad_norm": 1.978395700454712,
+      "learning_rate": 2e-05,
+      "loss": 0.06169991,
+      "step": 22561
+    },
+    {
+      "epoch": 45.124,
+      "grad_norm": 1.215388298034668,
+      "learning_rate": 2e-05,
+      "loss": 0.05454685,
+      "step": 22562
+    },
+    {
+      "epoch": 45.126,
+      "grad_norm": 0.908128559589386,
+      "learning_rate": 2e-05,
+      "loss": 0.028432,
+      "step": 22563
+    },
+    {
+      "epoch": 45.128,
+      "grad_norm": 1.1321818828582764,
+      "learning_rate": 2e-05,
+      "loss": 0.0391438,
+      "step": 22564
+    },
+    {
+      "epoch": 45.13,
+      "grad_norm": 1.0603052377700806,
+      "learning_rate": 2e-05,
+      "loss": 0.04576109,
+      "step": 22565
+    },
+    {
+      "epoch": 45.132,
+      "grad_norm": 1.1814618110656738,
+      "learning_rate": 2e-05,
+      "loss": 0.04789328,
+      "step": 22566
+    },
+    {
+      "epoch": 45.134,
+      "grad_norm": 1.4475892782211304,
+      "learning_rate": 2e-05,
+      "loss": 0.03765525,
+      "step": 22567
+    },
+    {
+      "epoch": 45.136,
+      "grad_norm": 1.2016019821166992,
+      "learning_rate": 2e-05,
+      "loss": 0.0461702,
+      "step": 22568
+    },
+    {
+      "epoch": 45.138,
+      "grad_norm": 1.2433879375457764,
+      "learning_rate": 2e-05,
+      "loss": 0.04532842,
+      "step": 22569
+    },
+    {
+      "epoch": 45.14,
+      "grad_norm": 1.0823651552200317,
+      "learning_rate": 2e-05,
+      "loss": 0.03890754,
+      "step": 22570
+    },
+    {
+      "epoch": 45.142,
+      "grad_norm": 1.3277068138122559,
+      "learning_rate": 2e-05,
+      "loss": 0.05912031,
+      "step": 22571
+    },
+    {
+      "epoch": 45.144,
+      "grad_norm": 1.2994036674499512,
+      "learning_rate": 2e-05,
+      "loss": 0.06122985,
+      "step": 22572
+    },
+    {
+      "epoch": 45.146,
+      "grad_norm": 6.131012916564941,
+      "learning_rate": 2e-05,
+      "loss": 0.06587172,
+      "step": 22573
+    },
+    {
+      "epoch": 45.148,
+      "grad_norm": 1.9716764688491821,
+      "learning_rate": 2e-05,
+      "loss": 0.06575805,
+      "step": 22574
+    },
+    {
+      "epoch": 45.15,
+      "grad_norm": 1.6159476041793823,
+      "learning_rate": 2e-05,
+      "loss": 0.04265655,
+      "step": 22575
+    },
+    {
+      "epoch": 45.152,
+      "grad_norm": 1.4026684761047363,
+      "learning_rate": 2e-05,
+      "loss": 0.04744727,
+      "step": 22576
+    },
+    {
+      "epoch": 45.154,
+      "grad_norm": 1.106533169746399,
+      "learning_rate": 2e-05,
+      "loss": 0.04791871,
+      "step": 22577
+    },
+    {
+      "epoch": 45.156,
+      "grad_norm": 1.9970992803573608,
+      "learning_rate": 2e-05,
+      "loss": 0.05857123,
+      "step": 22578
+    },
+    {
+      "epoch": 45.158,
+      "grad_norm": 2.6338624954223633,
+      "learning_rate": 2e-05,
+      "loss": 0.06465561,
+      "step": 22579
+    },
+    {
+      "epoch": 45.16,
+      "grad_norm": 3.545212507247925,
+      "learning_rate": 2e-05,
+      "loss": 0.0638639,
+      "step": 22580
+    },
+    {
+      "epoch": 45.162,
+      "grad_norm": 1.2237744331359863,
+      "learning_rate": 2e-05,
+      "loss": 0.05294487,
+      "step": 22581
+    },
+    {
+      "epoch": 45.164,
+      "grad_norm": 0.872916042804718,
+      "learning_rate": 2e-05,
+      "loss": 0.02183536,
+      "step": 22582
+    },
+    {
+      "epoch": 45.166,
+      "grad_norm": 1.3746517896652222,
+      "learning_rate": 2e-05,
+      "loss": 0.05417369,
+      "step": 22583
+    },
+    {
+      "epoch": 45.168,
+      "grad_norm": 1.1517207622528076,
+      "learning_rate": 2e-05,
+      "loss": 0.03980051,
+      "step": 22584
+    },
+    {
+      "epoch": 45.17,
+      "grad_norm": 1.2823890447616577,
+      "learning_rate": 2e-05,
+      "loss": 0.05957523,
+      "step": 22585
+    },
+    {
+      "epoch": 45.172,
+      "grad_norm": 1.4123610258102417,
+      "learning_rate": 2e-05,
+      "loss": 0.05167025,
+      "step": 22586
+    },
+    {
+      "epoch": 45.174,
+      "grad_norm": 1.3554683923721313,
+      "learning_rate": 2e-05,
+      "loss": 0.0463458,
+      "step": 22587
+    },
+    {
+      "epoch": 45.176,
+      "grad_norm": 1.6008424758911133,
+      "learning_rate": 2e-05,
+      "loss": 0.05228919,
+      "step": 22588
+    },
+    {
+      "epoch": 45.178,
+      "grad_norm": 1.1501260995864868,
+      "learning_rate": 2e-05,
+      "loss": 0.04863862,
+      "step": 22589
+    },
+    {
+      "epoch": 45.18,
+      "grad_norm": 0.9484877586364746,
+      "learning_rate": 2e-05,
+      "loss": 0.03127411,
+      "step": 22590
+    },
+    {
+      "epoch": 45.182,
+      "grad_norm": 1.6070059537887573,
+      "learning_rate": 2e-05,
+      "loss": 0.05212674,
+      "step": 22591
+    },
+    {
+      "epoch": 45.184,
+      "grad_norm": 1.3073357343673706,
+      "learning_rate": 2e-05,
+      "loss": 0.05999123,
+      "step": 22592
+    },
+    {
+      "epoch": 45.186,
+      "grad_norm": 1.1207677125930786,
+      "learning_rate": 2e-05,
+      "loss": 0.05001438,
+      "step": 22593
+    },
+    {
+      "epoch": 45.188,
+      "grad_norm": 1.6346867084503174,
+      "learning_rate": 2e-05,
+      "loss": 0.04875707,
+      "step": 22594
+    },
+    {
+      "epoch": 45.19,
+      "grad_norm": 1.2604049444198608,
+      "learning_rate": 2e-05,
+      "loss": 0.0618588,
+      "step": 22595
+    },
+    {
+      "epoch": 45.192,
+      "grad_norm": 1.1987062692642212,
+      "learning_rate": 2e-05,
+      "loss": 0.05671919,
+      "step": 22596
+    },
+    {
+      "epoch": 45.194,
+      "grad_norm": 1.0913773775100708,
+      "learning_rate": 2e-05,
+      "loss": 0.05150122,
+      "step": 22597
+    },
+    {
+      "epoch": 45.196,
+      "grad_norm": 1.0405243635177612,
+      "learning_rate": 2e-05,
+      "loss": 0.03671101,
+      "step": 22598
+    },
+    {
+      "epoch": 45.198,
+      "grad_norm": 1.0459086894989014,
+      "learning_rate": 2e-05,
+      "loss": 0.04587247,
+      "step": 22599
+    },
+    {
+      "epoch": 45.2,
+      "grad_norm": 0.8678653836250305,
+      "learning_rate": 2e-05,
+      "loss": 0.03158401,
+      "step": 22600
+    },
+    {
+      "epoch": 45.202,
+      "grad_norm": 1.2741189002990723,
+      "learning_rate": 2e-05,
+      "loss": 0.05065588,
+      "step": 22601
+    },
+    {
+      "epoch": 45.204,
+      "grad_norm": 2.527108907699585,
+      "learning_rate": 2e-05,
+      "loss": 0.05647211,
+      "step": 22602
+    },
+    {
+      "epoch": 45.206,
+      "grad_norm": 1.1058838367462158,
+      "learning_rate": 2e-05,
+      "loss": 0.04199415,
+      "step": 22603
+    },
+    {
+      "epoch": 45.208,
+      "grad_norm": 1.1132142543792725,
+      "learning_rate": 2e-05,
+      "loss": 0.06637686,
+      "step": 22604
+    },
+    {
+      "epoch": 45.21,
+      "grad_norm": 1.6548913717269897,
+      "learning_rate": 2e-05,
+      "loss": 0.05278435,
+      "step": 22605
+    },
+    {
+      "epoch": 45.212,
+      "grad_norm": 1.0226762294769287,
+      "learning_rate": 2e-05,
+      "loss": 0.03409335,
+      "step": 22606
+    },
+    {
+      "epoch": 45.214,
+      "grad_norm": 1.163483738899231,
+      "learning_rate": 2e-05,
+      "loss": 0.0428452,
+      "step": 22607
+    },
+    {
+      "epoch": 45.216,
+      "grad_norm": 0.9998294115066528,
+      "learning_rate": 2e-05,
+      "loss": 0.03900335,
+      "step": 22608
+    },
+    {
+      "epoch": 45.218,
+      "grad_norm": 1.5138134956359863,
+      "learning_rate": 2e-05,
+      "loss": 0.05385974,
+      "step": 22609
+    },
+    {
+      "epoch": 45.22,
+      "grad_norm": 1.0882453918457031,
+      "learning_rate": 2e-05,
+      "loss": 0.04675626,
+      "step": 22610
+    },
+    {
+      "epoch": 45.222,
+      "grad_norm": 1.041988730430603,
+      "learning_rate": 2e-05,
+      "loss": 0.04659366,
+      "step": 22611
+    },
+    {
+      "epoch": 45.224,
+      "grad_norm": 1.9339371919631958,
+      "learning_rate": 2e-05,
+      "loss": 0.05996863,
+      "step": 22612
+    },
+    {
+      "epoch": 45.226,
+      "grad_norm": 1.7372865676879883,
+      "learning_rate": 2e-05,
+      "loss": 0.06137735,
+      "step": 22613
+    },
+    {
+      "epoch": 45.228,
+      "grad_norm": 1.2295641899108887,
+      "learning_rate": 2e-05,
+      "loss": 0.05497001,
+      "step": 22614
+    },
+    {
+      "epoch": 45.23,
+      "grad_norm": 1.282944679260254,
+      "learning_rate": 2e-05,
+      "loss": 0.04878019,
+      "step": 22615
+    },
+    {
+      "epoch": 45.232,
+      "grad_norm": 1.0056757926940918,
+      "learning_rate": 2e-05,
+      "loss": 0.03496859,
+      "step": 22616
+    },
+    {
+      "epoch": 45.234,
+      "grad_norm": 1.1216771602630615,
+      "learning_rate": 2e-05,
+      "loss": 0.04637222,
+      "step": 22617
+    },
+    {
+      "epoch": 45.236,
+      "grad_norm": 1.1101162433624268,
+      "learning_rate": 2e-05,
+      "loss": 0.05418177,
+      "step": 22618
+    },
+    {
+      "epoch": 45.238,
+      "grad_norm": 2.1140050888061523,
+      "learning_rate": 2e-05,
+      "loss": 0.05144568,
+      "step": 22619
+    },
+    {
+      "epoch": 45.24,
+      "grad_norm": 1.0529059171676636,
+      "learning_rate": 2e-05,
+      "loss": 0.04527819,
+      "step": 22620
+    },
+    {
+      "epoch": 45.242,
+      "grad_norm": 1.2318655252456665,
+      "learning_rate": 2e-05,
+      "loss": 0.07417886,
+      "step": 22621
+    },
+    {
+      "epoch": 45.244,
+      "grad_norm": 1.1625219583511353,
+      "learning_rate": 2e-05,
+      "loss": 0.04924101,
+      "step": 22622
+    },
+    {
+      "epoch": 45.246,
+      "grad_norm": 1.1424137353897095,
+      "learning_rate": 2e-05,
+      "loss": 0.0479248,
+      "step": 22623
+    },
+    {
+      "epoch": 45.248,
+      "grad_norm": 0.9770327806472778,
+      "learning_rate": 2e-05,
+      "loss": 0.04924466,
+      "step": 22624
+    },
+    {
+      "epoch": 45.25,
+      "grad_norm": 1.0025733709335327,
+      "learning_rate": 2e-05,
+      "loss": 0.04422404,
+      "step": 22625
+    },
+    {
+      "epoch": 45.252,
+      "grad_norm": 1.084976315498352,
+      "learning_rate": 2e-05,
+      "loss": 0.04491837,
+      "step": 22626
+    },
+    {
+      "epoch": 45.254,
+      "grad_norm": 1.000255823135376,
+      "learning_rate": 2e-05,
+      "loss": 0.04638529,
+      "step": 22627
+    },
+    {
+      "epoch": 45.256,
+      "grad_norm": 1.7140803337097168,
+      "learning_rate": 2e-05,
+      "loss": 0.0516253,
+      "step": 22628
+    },
+    {
+      "epoch": 45.258,
+      "grad_norm": 1.0502322912216187,
+      "learning_rate": 2e-05,
+      "loss": 0.03990895,
+      "step": 22629
+    },
+    {
+      "epoch": 45.26,
+      "grad_norm": 1.12234628200531,
+      "learning_rate": 2e-05,
+      "loss": 0.04589079,
+      "step": 22630
+    },
+    {
+      "epoch": 45.262,
+      "grad_norm": 1.0489299297332764,
+      "learning_rate": 2e-05,
+      "loss": 0.03381399,
+      "step": 22631
+    },
+    {
+      "epoch": 45.264,
+      "grad_norm": 1.171677589416504,
+      "learning_rate": 2e-05,
+      "loss": 0.05054912,
+      "step": 22632
+    },
+    {
+      "epoch": 45.266,
+      "grad_norm": 1.0880128145217896,
+      "learning_rate": 2e-05,
+      "loss": 0.03867623,
+      "step": 22633
+    },
+    {
+      "epoch": 45.268,
+      "grad_norm": 2.91977596282959,
+      "learning_rate": 2e-05,
+      "loss": 0.05142038,
+      "step": 22634
+    },
+    {
+      "epoch": 45.27,
+      "grad_norm": 1.1017811298370361,
+      "learning_rate": 2e-05,
+      "loss": 0.04421684,
+      "step": 22635
+    },
+    {
+      "epoch": 45.272,
+      "grad_norm": 1.2263803482055664,
+      "learning_rate": 2e-05,
+      "loss": 0.0688946,
+      "step": 22636
+    },
+    {
+      "epoch": 45.274,
+      "grad_norm": 1.1422224044799805,
+      "learning_rate": 2e-05,
+      "loss": 0.04059188,
+      "step": 22637
+    },
+    {
+      "epoch": 45.276,
+      "grad_norm": 1.0902363061904907,
+      "learning_rate": 2e-05,
+      "loss": 0.04885675,
+      "step": 22638
+    },
+    {
+      "epoch": 45.278,
+      "grad_norm": 1.2119808197021484,
+      "learning_rate": 2e-05,
+      "loss": 0.04337999,
+      "step": 22639
+    },
+    {
+      "epoch": 45.28,
+      "grad_norm": 1.135478138923645,
+      "learning_rate": 2e-05,
+      "loss": 0.04769107,
+      "step": 22640
+    },
+    {
+      "epoch": 45.282,
+      "grad_norm": 0.9367907643318176,
+      "learning_rate": 2e-05,
+      "loss": 0.02835327,
+      "step": 22641
+    },
+    {
+      "epoch": 45.284,
+      "grad_norm": 1.128351092338562,
+      "learning_rate": 2e-05,
+      "loss": 0.05129791,
+      "step": 22642
+    },
+    {
+      "epoch": 45.286,
+      "grad_norm": 1.2856783866882324,
+      "learning_rate": 2e-05,
+      "loss": 0.05416092,
+      "step": 22643
+    },
+    {
+      "epoch": 45.288,
+      "grad_norm": 1.115014672279358,
+      "learning_rate": 2e-05,
+      "loss": 0.03043066,
+      "step": 22644
+    },
+    {
+      "epoch": 45.29,
+      "grad_norm": 1.2709392309188843,
+      "learning_rate": 2e-05,
+      "loss": 0.04415888,
+      "step": 22645
+    },
+    {
+      "epoch": 45.292,
+      "grad_norm": 1.6946849822998047,
+      "learning_rate": 2e-05,
+      "loss": 0.0498039,
+      "step": 22646
+    },
+    {
+      "epoch": 45.294,
+      "grad_norm": 3.176492929458618,
+      "learning_rate": 2e-05,
+      "loss": 0.08013065,
+      "step": 22647
+    },
+    {
+      "epoch": 45.296,
+      "grad_norm": 1.2074462175369263,
+      "learning_rate": 2e-05,
+      "loss": 0.05039094,
+      "step": 22648
+    },
+    {
+      "epoch": 45.298,
+      "grad_norm": 1.2501564025878906,
+      "learning_rate": 2e-05,
+      "loss": 0.05088755,
+      "step": 22649
+    },
+    {
+      "epoch": 45.3,
+      "grad_norm": 0.9809415340423584,
+      "learning_rate": 2e-05,
+      "loss": 0.038453,
+      "step": 22650
+    },
+    {
+      "epoch": 45.302,
+      "grad_norm": 1.3970571756362915,
+      "learning_rate": 2e-05,
+      "loss": 0.04629774,
+      "step": 22651
+    },
+    {
+      "epoch": 45.304,
+      "grad_norm": 1.4258559942245483,
+      "learning_rate": 2e-05,
+      "loss": 0.04505184,
+      "step": 22652
+    },
+    {
+      "epoch": 45.306,
+      "grad_norm": 1.1975897550582886,
+      "learning_rate": 2e-05,
+      "loss": 0.04722192,
+      "step": 22653
+    },
+    {
+      "epoch": 45.308,
+      "grad_norm": 1.2232824563980103,
+      "learning_rate": 2e-05,
+      "loss": 0.06708793,
+      "step": 22654
+    },
+    {
+      "epoch": 45.31,
+      "grad_norm": 0.874752938747406,
+      "learning_rate": 2e-05,
+      "loss": 0.03515249,
+      "step": 22655
+    },
+    {
+      "epoch": 45.312,
+      "grad_norm": 1.2584034204483032,
+      "learning_rate": 2e-05,
+      "loss": 0.04658163,
+      "step": 22656
+    },
+    {
+      "epoch": 45.314,
+      "grad_norm": 1.2609401941299438,
+      "learning_rate": 2e-05,
+      "loss": 0.05220157,
+      "step": 22657
+    },
+    {
+      "epoch": 45.316,
+      "grad_norm": 2.2510557174682617,
+      "learning_rate": 2e-05,
+      "loss": 0.04448275,
+      "step": 22658
+    },
+    {
+      "epoch": 45.318,
+      "grad_norm": 1.55043625831604,
+      "learning_rate": 2e-05,
+      "loss": 0.05576836,
+      "step": 22659
+    },
+    {
+      "epoch": 45.32,
+      "grad_norm": 1.1508514881134033,
+      "learning_rate": 2e-05,
+      "loss": 0.05758848,
+      "step": 22660
+    },
+    {
+      "epoch": 45.322,
+      "grad_norm": 1.7243223190307617,
+      "learning_rate": 2e-05,
+      "loss": 0.0692929,
+      "step": 22661
+    },
+    {
+      "epoch": 45.324,
+      "grad_norm": 1.220308780670166,
+      "learning_rate": 2e-05,
+      "loss": 0.04048155,
+      "step": 22662
+    },
+    {
+      "epoch": 45.326,
+      "grad_norm": 1.149865984916687,
+      "learning_rate": 2e-05,
+      "loss": 0.04047617,
+      "step": 22663
+    },
+    {
+      "epoch": 45.328,
+      "grad_norm": 1.3912262916564941,
+      "learning_rate": 2e-05,
+      "loss": 0.04581297,
+      "step": 22664
+    },
+    {
+      "epoch": 45.33,
+      "grad_norm": 1.1557172536849976,
+      "learning_rate": 2e-05,
+      "loss": 0.04854126,
+      "step": 22665
+    },
+    {
+      "epoch": 45.332,
+      "grad_norm": 1.5989816188812256,
+      "learning_rate": 2e-05,
+      "loss": 0.05772961,
+      "step": 22666
+    },
+    {
+      "epoch": 45.334,
+      "grad_norm": 0.9497352838516235,
+      "learning_rate": 2e-05,
+      "loss": 0.03526705,
+      "step": 22667
+    },
+    {
+      "epoch": 45.336,
+      "grad_norm": 1.0269160270690918,
+      "learning_rate": 2e-05,
+      "loss": 0.0354956,
+      "step": 22668
+    },
+    {
+      "epoch": 45.338,
+      "grad_norm": 0.9239309430122375,
+      "learning_rate": 2e-05,
+      "loss": 0.03152385,
+      "step": 22669
+    },
+    {
+      "epoch": 45.34,
+      "grad_norm": 1.0795482397079468,
+      "learning_rate": 2e-05,
+      "loss": 0.04576925,
+      "step": 22670
+    },
+    {
+      "epoch": 45.342,
+      "grad_norm": 1.2527785301208496,
+      "learning_rate": 2e-05,
+      "loss": 0.0519852,
+      "step": 22671
+    },
+    {
+      "epoch": 45.344,
+      "grad_norm": 1.2454344034194946,
+      "learning_rate": 2e-05,
+      "loss": 0.05516186,
+      "step": 22672
+    },
+    {
+      "epoch": 45.346,
+      "grad_norm": 1.122444748878479,
+      "learning_rate": 2e-05,
+      "loss": 0.05303994,
+      "step": 22673
+    },
+    {
+      "epoch": 45.348,
+      "grad_norm": 1.169447660446167,
+      "learning_rate": 2e-05,
+      "loss": 0.05172439,
+      "step": 22674
+    },
+    {
+      "epoch": 45.35,
+      "grad_norm": 1.2407636642456055,
+      "learning_rate": 2e-05,
+      "loss": 0.05318564,
+      "step": 22675
+    },
+    {
+      "epoch": 45.352,
+      "grad_norm": 1.0374010801315308,
+      "learning_rate": 2e-05,
+      "loss": 0.0399787,
+      "step": 22676
+    },
+    {
+      "epoch": 45.354,
+      "grad_norm": 1.6039947271347046,
+      "learning_rate": 2e-05,
+      "loss": 0.04658148,
+      "step": 22677
+    },
+    {
+      "epoch": 45.356,
+      "grad_norm": 2.6198055744171143,
+      "learning_rate": 2e-05,
+      "loss": 0.04925442,
+      "step": 22678
+    },
+    {
+      "epoch": 45.358,
+      "grad_norm": 1.2547622919082642,
+      "learning_rate": 2e-05,
+      "loss": 0.05453364,
+      "step": 22679
+    },
+    {
+      "epoch": 45.36,
+      "grad_norm": 1.2854013442993164,
+      "learning_rate": 2e-05,
+      "loss": 0.05297067,
+      "step": 22680
+    },
+    {
+      "epoch": 45.362,
+      "grad_norm": 1.3244527578353882,
+      "learning_rate": 2e-05,
+      "loss": 0.0462161,
+      "step": 22681
+    },
+    {
+      "epoch": 45.364,
+      "grad_norm": 1.1322252750396729,
+      "learning_rate": 2e-05,
+      "loss": 0.04703336,
+      "step": 22682
+    },
+    {
+      "epoch": 45.366,
+      "grad_norm": 1.50275456905365,
+      "learning_rate": 2e-05,
+      "loss": 0.0487467,
+      "step": 22683
+    },
+    {
+      "epoch": 45.368,
+      "grad_norm": 1.0968610048294067,
+      "learning_rate": 2e-05,
+      "loss": 0.03857952,
+      "step": 22684
+    },
+    {
+      "epoch": 45.37,
+      "grad_norm": 1.4444911479949951,
+      "learning_rate": 2e-05,
+      "loss": 0.0504374,
+      "step": 22685
+    },
+    {
+      "epoch": 45.372,
+      "grad_norm": 1.5214262008666992,
+      "learning_rate": 2e-05,
+      "loss": 0.05310171,
+      "step": 22686
+    },
+    {
+      "epoch": 45.374,
+      "grad_norm": 1.0097070932388306,
+      "learning_rate": 2e-05,
+      "loss": 0.04260479,
+      "step": 22687
+    },
+    {
+      "epoch": 45.376,
+      "grad_norm": 0.984809160232544,
+      "learning_rate": 2e-05,
+      "loss": 0.03620268,
+      "step": 22688
+    },
+    {
+      "epoch": 45.378,
+      "grad_norm": 3.517108201980591,
+      "learning_rate": 2e-05,
+      "loss": 0.05183204,
+      "step": 22689
+    },
+    {
+      "epoch": 45.38,
+      "grad_norm": 1.4225305318832397,
+      "learning_rate": 2e-05,
+      "loss": 0.05968772,
+      "step": 22690
+    },
+    {
+      "epoch": 45.382,
+      "grad_norm": 1.4418306350708008,
+      "learning_rate": 2e-05,
+      "loss": 0.04406624,
+      "step": 22691
+    },
+    {
+      "epoch": 45.384,
+      "grad_norm": 1.0155731439590454,
+      "learning_rate": 2e-05,
+      "loss": 0.04117178,
+      "step": 22692
+    },
+    {
+      "epoch": 45.386,
+      "grad_norm": 1.1497960090637207,
+      "learning_rate": 2e-05,
+      "loss": 0.04193247,
+      "step": 22693
+    },
+    {
+      "epoch": 45.388,
+      "grad_norm": 1.2186371088027954,
+      "learning_rate": 2e-05,
+      "loss": 0.05743311,
+      "step": 22694
+    },
+    {
+      "epoch": 45.39,
+      "grad_norm": 1.2327183485031128,
+      "learning_rate": 2e-05,
+      "loss": 0.04604057,
+      "step": 22695
+    },
+    {
+      "epoch": 45.392,
+      "grad_norm": 1.4449176788330078,
+      "learning_rate": 2e-05,
+      "loss": 0.06641807,
+      "step": 22696
+    },
+    {
+      "epoch": 45.394,
+      "grad_norm": 1.2744368314743042,
+      "learning_rate": 2e-05,
+      "loss": 0.0488281,
+      "step": 22697
+    },
+    {
+      "epoch": 45.396,
+      "grad_norm": 1.1756504774093628,
+      "learning_rate": 2e-05,
+      "loss": 0.04841977,
+      "step": 22698
+    },
+    {
+      "epoch": 45.398,
+      "grad_norm": 1.5969294309616089,
+      "learning_rate": 2e-05,
+      "loss": 0.0381519,
+      "step": 22699
+    },
+    {
+      "epoch": 45.4,
+      "grad_norm": 1.177388310432434,
+      "learning_rate": 2e-05,
+      "loss": 0.05469193,
+      "step": 22700
+    },
+    {
+      "epoch": 45.402,
+      "grad_norm": 1.2466199398040771,
+      "learning_rate": 2e-05,
+      "loss": 0.04355872,
+      "step": 22701
+    },
+    {
+      "epoch": 45.404,
+      "grad_norm": 1.1346672773361206,
+      "learning_rate": 2e-05,
+      "loss": 0.0505007,
+      "step": 22702
+    },
+    {
+      "epoch": 45.406,
+      "grad_norm": 1.0802158117294312,
+      "learning_rate": 2e-05,
+      "loss": 0.03773396,
+      "step": 22703
+    },
+    {
+      "epoch": 45.408,
+      "grad_norm": 1.0188031196594238,
+      "learning_rate": 2e-05,
+      "loss": 0.03674341,
+      "step": 22704
+    },
+    {
+      "epoch": 45.41,
+      "grad_norm": 0.9986094236373901,
+      "learning_rate": 2e-05,
+      "loss": 0.04618426,
+      "step": 22705
+    },
+    {
+      "epoch": 45.412,
+      "grad_norm": 1.0472339391708374,
+      "learning_rate": 2e-05,
+      "loss": 0.04356561,
+      "step": 22706
+    },
+    {
+      "epoch": 45.414,
+      "grad_norm": 1.2004998922348022,
+      "learning_rate": 2e-05,
+      "loss": 0.05084341,
+      "step": 22707
+    },
+    {
+      "epoch": 45.416,
+      "grad_norm": 1.2121773958206177,
+      "learning_rate": 2e-05,
+      "loss": 0.04511089,
+      "step": 22708
+    },
+    {
+      "epoch": 45.418,
+      "grad_norm": 1.7242828607559204,
+      "learning_rate": 2e-05,
+      "loss": 0.0549062,
+      "step": 22709
+    },
+    {
+      "epoch": 45.42,
+      "grad_norm": 1.154462218284607,
+      "learning_rate": 2e-05,
+      "loss": 0.04893161,
+      "step": 22710
+    },
+    {
+      "epoch": 45.422,
+      "grad_norm": 1.1248419284820557,
+      "learning_rate": 2e-05,
+      "loss": 0.04025434,
+      "step": 22711
+    },
+    {
+      "epoch": 45.424,
+      "grad_norm": 1.2144428491592407,
+      "learning_rate": 2e-05,
+      "loss": 0.04915642,
+      "step": 22712
+    },
+    {
+      "epoch": 45.426,
+      "grad_norm": 1.1011919975280762,
+      "learning_rate": 2e-05,
+      "loss": 0.04912124,
+      "step": 22713
+    },
+    {
+      "epoch": 45.428,
+      "grad_norm": 1.134437084197998,
+      "learning_rate": 2e-05,
+      "loss": 0.0618168,
+      "step": 22714
+    },
+    {
+      "epoch": 45.43,
+      "grad_norm": 1.2620517015457153,
+      "learning_rate": 2e-05,
+      "loss": 0.06262559,
+      "step": 22715
+    },
+    {
+      "epoch": 45.432,
+      "grad_norm": 1.0983620882034302,
+      "learning_rate": 2e-05,
+      "loss": 0.04276703,
+      "step": 22716
+    },
+    {
+      "epoch": 45.434,
+      "grad_norm": 1.1140186786651611,
+      "learning_rate": 2e-05,
+      "loss": 0.05332418,
+      "step": 22717
+    },
+    {
+      "epoch": 45.436,
+      "grad_norm": 1.3023191690444946,
+      "learning_rate": 2e-05,
+      "loss": 0.06473353,
+      "step": 22718
+    },
+    {
+      "epoch": 45.438,
+      "grad_norm": 0.8834937810897827,
+      "learning_rate": 2e-05,
+      "loss": 0.04048298,
+      "step": 22719
+    },
+    {
+      "epoch": 45.44,
+      "grad_norm": 1.276069164276123,
+      "learning_rate": 2e-05,
+      "loss": 0.0446801,
+      "step": 22720
+    },
+    {
+      "epoch": 45.442,
+      "grad_norm": 1.2004058361053467,
+      "learning_rate": 2e-05,
+      "loss": 0.05174657,
+      "step": 22721
+    },
+    {
+      "epoch": 45.444,
+      "grad_norm": 1.3467042446136475,
+      "learning_rate": 2e-05,
+      "loss": 0.06729078,
+      "step": 22722
+    },
+    {
+      "epoch": 45.446,
+      "grad_norm": 1.200899600982666,
+      "learning_rate": 2e-05,
+      "loss": 0.04949638,
+      "step": 22723
+    },
+    {
+      "epoch": 45.448,
+      "grad_norm": 1.2946702241897583,
+      "learning_rate": 2e-05,
+      "loss": 0.05751118,
+      "step": 22724
+    },
+    {
+      "epoch": 45.45,
+      "grad_norm": 1.094891905784607,
+      "learning_rate": 2e-05,
+      "loss": 0.04523872,
+      "step": 22725
+    },
+    {
+      "epoch": 45.452,
+      "grad_norm": 1.049796462059021,
+      "learning_rate": 2e-05,
+      "loss": 0.05196452,
+      "step": 22726
+    },
+    {
+      "epoch": 45.454,
+      "grad_norm": 1.2776124477386475,
+      "learning_rate": 2e-05,
+      "loss": 0.04980614,
+      "step": 22727
+    },
+    {
+      "epoch": 45.456,
+      "grad_norm": 1.2622987031936646,
+      "learning_rate": 2e-05,
+      "loss": 0.05976017,
+      "step": 22728
+    },
+    {
+      "epoch": 45.458,
+      "grad_norm": 1.2927825450897217,
+      "learning_rate": 2e-05,
+      "loss": 0.05953868,
+      "step": 22729
+    },
+    {
+      "epoch": 45.46,
+      "grad_norm": 1.0884168148040771,
+      "learning_rate": 2e-05,
+      "loss": 0.03435563,
+      "step": 22730
+    },
+    {
+      "epoch": 45.462,
+      "grad_norm": 1.2159061431884766,
+      "learning_rate": 2e-05,
+      "loss": 0.05816945,
+      "step": 22731
+    },
+    {
+      "epoch": 45.464,
+      "grad_norm": 1.9903509616851807,
+      "learning_rate": 2e-05,
+      "loss": 0.05068534,
+      "step": 22732
+    },
+    {
+      "epoch": 45.466,
+      "grad_norm": 1.7449239492416382,
+      "learning_rate": 2e-05,
+      "loss": 0.05164981,
+      "step": 22733
+    },
+    {
+      "epoch": 45.468,
+      "grad_norm": 0.8933517336845398,
+      "learning_rate": 2e-05,
+      "loss": 0.03138543,
+      "step": 22734
+    },
+    {
+      "epoch": 45.47,
+      "grad_norm": 1.1336041688919067,
+      "learning_rate": 2e-05,
+      "loss": 0.04539839,
+      "step": 22735
+    },
+    {
+      "epoch": 45.472,
+      "grad_norm": 0.8588364720344543,
+      "learning_rate": 2e-05,
+      "loss": 0.03250713,
+      "step": 22736
+    },
+    {
+      "epoch": 45.474,
+      "grad_norm": 0.9573074579238892,
+      "learning_rate": 2e-05,
+      "loss": 0.03433614,
+      "step": 22737
+    },
+    {
+      "epoch": 45.476,
+      "grad_norm": 3.1477766036987305,
+      "learning_rate": 2e-05,
+      "loss": 0.07124802,
+      "step": 22738
+    },
+    {
+      "epoch": 45.478,
+      "grad_norm": 1.1136095523834229,
+      "learning_rate": 2e-05,
+      "loss": 0.04518589,
+      "step": 22739
+    },
+    {
+      "epoch": 45.48,
+      "grad_norm": 1.7395950555801392,
+      "learning_rate": 2e-05,
+      "loss": 0.03977414,
+      "step": 22740
+    },
+    {
+      "epoch": 45.482,
+      "grad_norm": 1.2946288585662842,
+      "learning_rate": 2e-05,
+      "loss": 0.04495868,
+      "step": 22741
+    },
+    {
+      "epoch": 45.484,
+      "grad_norm": 1.1849554777145386,
+      "learning_rate": 2e-05,
+      "loss": 0.04799411,
+      "step": 22742
+    },
+    {
+      "epoch": 45.486,
+      "grad_norm": 1.13692307472229,
+      "learning_rate": 2e-05,
+      "loss": 0.04666179,
+      "step": 22743
+    },
+    {
+      "epoch": 45.488,
+      "grad_norm": 1.1994067430496216,
+      "learning_rate": 2e-05,
+      "loss": 0.04230376,
+      "step": 22744
+    },
+    {
+      "epoch": 45.49,
+      "grad_norm": 1.5030912160873413,
+      "learning_rate": 2e-05,
+      "loss": 0.05279317,
+      "step": 22745
+    },
+    {
+      "epoch": 45.492,
+      "grad_norm": 0.8699735999107361,
+      "learning_rate": 2e-05,
+      "loss": 0.02855832,
+      "step": 22746
+    },
+    {
+      "epoch": 45.494,
+      "grad_norm": 1.2766979932785034,
+      "learning_rate": 2e-05,
+      "loss": 0.04953785,
+      "step": 22747
+    },
+    {
+      "epoch": 45.496,
+      "grad_norm": 1.0743104219436646,
+      "learning_rate": 2e-05,
+      "loss": 0.0503507,
+      "step": 22748
+    },
+    {
+      "epoch": 45.498,
+      "grad_norm": 1.15504789352417,
+      "learning_rate": 2e-05,
+      "loss": 0.04353083,
+      "step": 22749
+    },
+    {
+      "epoch": 45.5,
+      "grad_norm": 1.5795592069625854,
+      "learning_rate": 2e-05,
+      "loss": 0.05220584,
+      "step": 22750
+    },
+    {
+      "epoch": 45.502,
+      "grad_norm": 1.1085896492004395,
+      "learning_rate": 2e-05,
+      "loss": 0.05066767,
+      "step": 22751
+    },
+    {
+      "epoch": 45.504,
+      "grad_norm": 1.2716381549835205,
+      "learning_rate": 2e-05,
+      "loss": 0.06609177,
+      "step": 22752
+    },
+    {
+      "epoch": 45.506,
+      "grad_norm": 1.9817065000534058,
+      "learning_rate": 2e-05,
+      "loss": 0.06183188,
+      "step": 22753
+    },
+    {
+      "epoch": 45.508,
+      "grad_norm": 1.3558555841445923,
+      "learning_rate": 2e-05,
+      "loss": 0.05611308,
+      "step": 22754
+    },
+    {
+      "epoch": 45.51,
+      "grad_norm": 1.2409502267837524,
+      "learning_rate": 2e-05,
+      "loss": 0.03493334,
+      "step": 22755
+    },
+    {
+      "epoch": 45.512,
+      "grad_norm": 1.2743219137191772,
+      "learning_rate": 2e-05,
+      "loss": 0.04769837,
+      "step": 22756
+    },
+    {
+      "epoch": 45.514,
+      "grad_norm": 0.9491370320320129,
+      "learning_rate": 2e-05,
+      "loss": 0.03029821,
+      "step": 22757
+    },
+    {
+      "epoch": 45.516,
+      "grad_norm": 4.352717876434326,
+      "learning_rate": 2e-05,
+      "loss": 0.03928834,
+      "step": 22758
+    },
+    {
+      "epoch": 45.518,
+      "grad_norm": 1.3825864791870117,
+      "learning_rate": 2e-05,
+      "loss": 0.05772555,
+      "step": 22759
+    },
+    {
+      "epoch": 45.52,
+      "grad_norm": 1.626591682434082,
+      "learning_rate": 2e-05,
+      "loss": 0.05705488,
+      "step": 22760
+    },
+    {
+      "epoch": 45.522,
+      "grad_norm": 1.2819281816482544,
+      "learning_rate": 2e-05,
+      "loss": 0.06068337,
+      "step": 22761
+    },
+    {
+      "epoch": 45.524,
+      "grad_norm": 2.8456923961639404,
+      "learning_rate": 2e-05,
+      "loss": 0.05577563,
+      "step": 22762
+    },
+    {
+      "epoch": 45.526,
+      "grad_norm": 1.33793044090271,
+      "learning_rate": 2e-05,
+      "loss": 0.05010552,
+      "step": 22763
+    },
+    {
+      "epoch": 45.528,
+      "grad_norm": 1.059351921081543,
+      "learning_rate": 2e-05,
+      "loss": 0.04426561,
+      "step": 22764
+    },
+    {
+      "epoch": 45.53,
+      "grad_norm": 1.1769965887069702,
+      "learning_rate": 2e-05,
+      "loss": 0.03966656,
+      "step": 22765
+    },
+    {
+      "epoch": 45.532,
+      "grad_norm": 1.2695751190185547,
+      "learning_rate": 2e-05,
+      "loss": 0.04549221,
+      "step": 22766
+    },
+    {
+      "epoch": 45.534,
+      "grad_norm": 1.2788498401641846,
+      "learning_rate": 2e-05,
+      "loss": 0.05216346,
+      "step": 22767
+    },
+    {
+      "epoch": 45.536,
+      "grad_norm": 0.9913758039474487,
+      "learning_rate": 2e-05,
+      "loss": 0.03447246,
+      "step": 22768
+    },
+    {
+      "epoch": 45.538,
+      "grad_norm": 1.0714240074157715,
+      "learning_rate": 2e-05,
+      "loss": 0.04273647,
+      "step": 22769
+    },
+    {
+      "epoch": 45.54,
+      "grad_norm": 1.2439073324203491,
+      "learning_rate": 2e-05,
+      "loss": 0.04950628,
+      "step": 22770
+    },
+    {
+      "epoch": 45.542,
+      "grad_norm": 1.4593321084976196,
+      "learning_rate": 2e-05,
+      "loss": 0.03826152,
+      "step": 22771
+    },
+    {
+      "epoch": 45.544,
+      "grad_norm": 1.562803864479065,
+      "learning_rate": 2e-05,
+      "loss": 0.05784161,
+      "step": 22772
+    },
+    {
+      "epoch": 45.546,
+      "grad_norm": 0.9718835949897766,
+      "learning_rate": 2e-05,
+      "loss": 0.04087242,
+      "step": 22773
+    },
+    {
+      "epoch": 45.548,
+      "grad_norm": 1.1946215629577637,
+      "learning_rate": 2e-05,
+      "loss": 0.03955045,
+      "step": 22774
+    },
+    {
+      "epoch": 45.55,
+      "grad_norm": 5.945659160614014,
+      "learning_rate": 2e-05,
+      "loss": 0.05163043,
+      "step": 22775
+    },
+    {
+      "epoch": 45.552,
+      "grad_norm": 0.9503419995307922,
+      "learning_rate": 2e-05,
+      "loss": 0.03446332,
+      "step": 22776
+    },
+    {
+      "epoch": 45.554,
+      "grad_norm": 2.860668897628784,
+      "learning_rate": 2e-05,
+      "loss": 0.06032272,
+      "step": 22777
+    },
+    {
+      "epoch": 45.556,
+      "grad_norm": 1.0413198471069336,
+      "learning_rate": 2e-05,
+      "loss": 0.03605777,
+      "step": 22778
+    },
+    {
+      "epoch": 45.558,
+      "grad_norm": 1.3972159624099731,
+      "learning_rate": 2e-05,
+      "loss": 0.0614309,
+      "step": 22779
+    },
+    {
+      "epoch": 45.56,
+      "grad_norm": 1.1884814500808716,
+      "learning_rate": 2e-05,
+      "loss": 0.04031287,
+      "step": 22780
+    },
+    {
+      "epoch": 45.562,
+      "grad_norm": 1.1417438983917236,
+      "learning_rate": 2e-05,
+      "loss": 0.04732093,
+      "step": 22781
+    },
+    {
+      "epoch": 45.564,
+      "grad_norm": 1.4334638118743896,
+      "learning_rate": 2e-05,
+      "loss": 0.05984236,
+      "step": 22782
+    },
+    {
+      "epoch": 45.566,
+      "grad_norm": 2.67592453956604,
+      "learning_rate": 2e-05,
+      "loss": 0.07209148,
+      "step": 22783
+    },
+    {
+      "epoch": 45.568,
+      "grad_norm": 1.1358227729797363,
+      "learning_rate": 2e-05,
+      "loss": 0.04441652,
+      "step": 22784
+    },
+    {
+      "epoch": 45.57,
+      "grad_norm": 1.161139726638794,
+      "learning_rate": 2e-05,
+      "loss": 0.05294163,
+      "step": 22785
+    },
+    {
+      "epoch": 45.572,
+      "grad_norm": 4.65118932723999,
+      "learning_rate": 2e-05,
+      "loss": 0.03936018,
+      "step": 22786
+    },
+    {
+      "epoch": 45.574,
+      "grad_norm": 1.0236941576004028,
+      "learning_rate": 2e-05,
+      "loss": 0.05270987,
+      "step": 22787
+    },
+    {
+      "epoch": 45.576,
+      "grad_norm": 1.3920187950134277,
+      "learning_rate": 2e-05,
+      "loss": 0.05220519,
+      "step": 22788
+    },
+    {
+      "epoch": 45.578,
+      "grad_norm": 1.4506982564926147,
+      "learning_rate": 2e-05,
+      "loss": 0.05476096,
+      "step": 22789
+    },
+    {
+      "epoch": 45.58,
+      "grad_norm": 1.6124314069747925,
+      "learning_rate": 2e-05,
+      "loss": 0.04182716,
+      "step": 22790
+    },
+    {
+      "epoch": 45.582,
+      "grad_norm": 1.353986144065857,
+      "learning_rate": 2e-05,
+      "loss": 0.04613897,
+      "step": 22791
+    },
+    {
+      "epoch": 45.584,
+      "grad_norm": 1.1616084575653076,
+      "learning_rate": 2e-05,
+      "loss": 0.06366609,
+      "step": 22792
+    },
+    {
+      "epoch": 45.586,
+      "grad_norm": 1.0128250122070312,
+      "learning_rate": 2e-05,
+      "loss": 0.03711364,
+      "step": 22793
+    },
+    {
+      "epoch": 45.588,
+      "grad_norm": 1.2062525749206543,
+      "learning_rate": 2e-05,
+      "loss": 0.04811207,
+      "step": 22794
+    },
+    {
+      "epoch": 45.59,
+      "grad_norm": 1.031899094581604,
+      "learning_rate": 2e-05,
+      "loss": 0.03934745,
+      "step": 22795
+    },
+    {
+      "epoch": 45.592,
+      "grad_norm": 1.0343753099441528,
+      "learning_rate": 2e-05,
+      "loss": 0.04823088,
+      "step": 22796
+    },
+    {
+      "epoch": 45.594,
+      "grad_norm": 3.7844700813293457,
+      "learning_rate": 2e-05,
+      "loss": 0.05894583,
+      "step": 22797
+    },
+    {
+      "epoch": 45.596,
+      "grad_norm": 1.1267212629318237,
+      "learning_rate": 2e-05,
+      "loss": 0.03568856,
+      "step": 22798
+    },
+    {
+      "epoch": 45.598,
+      "grad_norm": 1.3324168920516968,
+      "learning_rate": 2e-05,
+      "loss": 0.05654506,
+      "step": 22799
+    },
+    {
+      "epoch": 45.6,
+      "grad_norm": 1.0916633605957031,
+      "learning_rate": 2e-05,
+      "loss": 0.04748604,
+      "step": 22800
+    },
+    {
+      "epoch": 45.602,
+      "grad_norm": 1.7681325674057007,
+      "learning_rate": 2e-05,
+      "loss": 0.04749766,
+      "step": 22801
+    },
+    {
+      "epoch": 45.604,
+      "grad_norm": 1.385416030883789,
+      "learning_rate": 2e-05,
+      "loss": 0.05868257,
+      "step": 22802
+    },
+    {
+      "epoch": 45.606,
+      "grad_norm": 2.559743642807007,
+      "learning_rate": 2e-05,
+      "loss": 0.06657992,
+      "step": 22803
+    },
+    {
+      "epoch": 45.608,
+      "grad_norm": 1.5238560438156128,
+      "learning_rate": 2e-05,
+      "loss": 0.05722816,
+      "step": 22804
+    },
+    {
+      "epoch": 45.61,
+      "grad_norm": 1.215009331703186,
+      "learning_rate": 2e-05,
+      "loss": 0.05467505,
+      "step": 22805
+    },
+    {
+      "epoch": 45.612,
+      "grad_norm": 1.3188890218734741,
+      "learning_rate": 2e-05,
+      "loss": 0.07242645,
+      "step": 22806
+    },
+    {
+      "epoch": 45.614,
+      "grad_norm": 1.3299347162246704,
+      "learning_rate": 2e-05,
+      "loss": 0.04678376,
+      "step": 22807
+    },
+    {
+      "epoch": 45.616,
+      "grad_norm": 1.213125467300415,
+      "learning_rate": 2e-05,
+      "loss": 0.05465173,
+      "step": 22808
+    },
+    {
+      "epoch": 45.618,
+      "grad_norm": 0.9832747578620911,
+      "learning_rate": 2e-05,
+      "loss": 0.03593368,
+      "step": 22809
+    },
+    {
+      "epoch": 45.62,
+      "grad_norm": 1.4957003593444824,
+      "learning_rate": 2e-05,
+      "loss": 0.04334493,
+      "step": 22810
+    },
+    {
+      "epoch": 45.622,
+      "grad_norm": 1.1476776599884033,
+      "learning_rate": 2e-05,
+      "loss": 0.05104943,
+      "step": 22811
+    },
+    {
+      "epoch": 45.624,
+      "grad_norm": 1.1928304433822632,
+      "learning_rate": 2e-05,
+      "loss": 0.05153049,
+      "step": 22812
+    },
+    {
+      "epoch": 45.626,
+      "grad_norm": 3.3560004234313965,
+      "learning_rate": 2e-05,
+      "loss": 0.04769015,
+      "step": 22813
+    },
+    {
+      "epoch": 45.628,
+      "grad_norm": 1.2147530317306519,
+      "learning_rate": 2e-05,
+      "loss": 0.04993775,
+      "step": 22814
+    },
+    {
+      "epoch": 45.63,
+      "grad_norm": 1.3720028400421143,
+      "learning_rate": 2e-05,
+      "loss": 0.04633579,
+      "step": 22815
+    },
+    {
+      "epoch": 45.632,
+      "grad_norm": 1.1289563179016113,
+      "learning_rate": 2e-05,
+      "loss": 0.0430528,
+      "step": 22816
+    },
+    {
+      "epoch": 45.634,
+      "grad_norm": 1.3510793447494507,
+      "learning_rate": 2e-05,
+      "loss": 0.04878941,
+      "step": 22817
+    },
+    {
+      "epoch": 45.636,
+      "grad_norm": 1.4363319873809814,
+      "learning_rate": 2e-05,
+      "loss": 0.05465399,
+      "step": 22818
+    },
+    {
+      "epoch": 45.638,
+      "grad_norm": 2.460465669631958,
+      "learning_rate": 2e-05,
+      "loss": 0.05995677,
+      "step": 22819
+    },
+    {
+      "epoch": 45.64,
+      "grad_norm": 1.1262542009353638,
+      "learning_rate": 2e-05,
+      "loss": 0.03892493,
+      "step": 22820
+    },
+    {
+      "epoch": 45.642,
+      "grad_norm": 1.2960659265518188,
+      "learning_rate": 2e-05,
+      "loss": 0.0484963,
+      "step": 22821
+    },
+    {
+      "epoch": 45.644,
+      "grad_norm": 1.1812695264816284,
+      "learning_rate": 2e-05,
+      "loss": 0.05082157,
+      "step": 22822
+    },
+    {
+      "epoch": 45.646,
+      "grad_norm": 1.172400712966919,
+      "learning_rate": 2e-05,
+      "loss": 0.04413823,
+      "step": 22823
+    },
+    {
+      "epoch": 45.648,
+      "grad_norm": 1.1976510286331177,
+      "learning_rate": 2e-05,
+      "loss": 0.04225794,
+      "step": 22824
+    },
+    {
+      "epoch": 45.65,
+      "grad_norm": 1.4453816413879395,
+      "learning_rate": 2e-05,
+      "loss": 0.05087319,
+      "step": 22825
+    },
+    {
+      "epoch": 45.652,
+      "grad_norm": 0.993677020072937,
+      "learning_rate": 2e-05,
+      "loss": 0.02660798,
+      "step": 22826
+    },
+    {
+      "epoch": 45.654,
+      "grad_norm": 1.0433624982833862,
+      "learning_rate": 2e-05,
+      "loss": 0.04253418,
+      "step": 22827
+    },
+    {
+      "epoch": 45.656,
+      "grad_norm": 1.026785969734192,
+      "learning_rate": 2e-05,
+      "loss": 0.04078073,
+      "step": 22828
+    },
+    {
+      "epoch": 45.658,
+      "grad_norm": 1.125288963317871,
+      "learning_rate": 2e-05,
+      "loss": 0.04690692,
+      "step": 22829
+    },
+    {
+      "epoch": 45.66,
+      "grad_norm": 1.1184576749801636,
+      "learning_rate": 2e-05,
+      "loss": 0.02984325,
+      "step": 22830
+    },
+    {
+      "epoch": 45.662,
+      "grad_norm": 1.30687415599823,
+      "learning_rate": 2e-05,
+      "loss": 0.05822021,
+      "step": 22831
+    },
+    {
+      "epoch": 45.664,
+      "grad_norm": 1.2212743759155273,
+      "learning_rate": 2e-05,
+      "loss": 0.04939518,
+      "step": 22832
+    },
+    {
+      "epoch": 45.666,
+      "grad_norm": 1.186759352684021,
+      "learning_rate": 2e-05,
+      "loss": 0.05301164,
+      "step": 22833
+    },
+    {
+      "epoch": 45.668,
+      "grad_norm": 1.2905839681625366,
+      "learning_rate": 2e-05,
+      "loss": 0.04207082,
+      "step": 22834
+    },
+    {
+      "epoch": 45.67,
+      "grad_norm": 1.1900533437728882,
+      "learning_rate": 2e-05,
+      "loss": 0.05991127,
+      "step": 22835
+    },
+    {
+      "epoch": 45.672,
+      "grad_norm": 2.1371090412139893,
+      "learning_rate": 2e-05,
+      "loss": 0.05556452,
+      "step": 22836
+    },
+    {
+      "epoch": 45.674,
+      "grad_norm": 1.6555566787719727,
+      "learning_rate": 2e-05,
+      "loss": 0.0666324,
+      "step": 22837
+    },
+    {
+      "epoch": 45.676,
+      "grad_norm": 1.2142423391342163,
+      "learning_rate": 2e-05,
+      "loss": 0.04500743,
+      "step": 22838
+    },
+    {
+      "epoch": 45.678,
+      "grad_norm": 1.204254150390625,
+      "learning_rate": 2e-05,
+      "loss": 0.05000075,
+      "step": 22839
+    },
+    {
+      "epoch": 45.68,
+      "grad_norm": 1.3275655508041382,
+      "learning_rate": 2e-05,
+      "loss": 0.05476985,
+      "step": 22840
+    },
+    {
+      "epoch": 45.682,
+      "grad_norm": 1.0724529027938843,
+      "learning_rate": 2e-05,
+      "loss": 0.03965185,
+      "step": 22841
+    },
+    {
+      "epoch": 45.684,
+      "grad_norm": 1.363209843635559,
+      "learning_rate": 2e-05,
+      "loss": 0.0393138,
+      "step": 22842
+    },
+    {
+      "epoch": 45.686,
+      "grad_norm": 1.3756287097930908,
+      "learning_rate": 2e-05,
+      "loss": 0.0670438,
+      "step": 22843
+    },
+    {
+      "epoch": 45.688,
+      "grad_norm": 1.252241849899292,
+      "learning_rate": 2e-05,
+      "loss": 0.05263539,
+      "step": 22844
+    },
+    {
+      "epoch": 45.69,
+      "grad_norm": 1.1118369102478027,
+      "learning_rate": 2e-05,
+      "loss": 0.05109958,
+      "step": 22845
+    },
+    {
+      "epoch": 45.692,
+      "grad_norm": 1.2781487703323364,
+      "learning_rate": 2e-05,
+      "loss": 0.05006749,
+      "step": 22846
+    },
+    {
+      "epoch": 45.694,
+      "grad_norm": 1.309794306755066,
+      "learning_rate": 2e-05,
+      "loss": 0.03993253,
+      "step": 22847
+    },
+    {
+      "epoch": 45.696,
+      "grad_norm": 1.3862566947937012,
+      "learning_rate": 2e-05,
+      "loss": 0.03329,
+      "step": 22848
+    },
+    {
+      "epoch": 45.698,
+      "grad_norm": 1.0639959573745728,
+      "learning_rate": 2e-05,
+      "loss": 0.03904609,
+      "step": 22849
+    },
+    {
+      "epoch": 45.7,
+      "grad_norm": 1.0626970529556274,
+      "learning_rate": 2e-05,
+      "loss": 0.03999314,
+      "step": 22850
+    },
+    {
+      "epoch": 45.702,
+      "grad_norm": 1.2711509466171265,
+      "learning_rate": 2e-05,
+      "loss": 0.0375726,
+      "step": 22851
+    },
+    {
+      "epoch": 45.704,
+      "grad_norm": 0.9790001511573792,
+      "learning_rate": 2e-05,
+      "loss": 0.03540372,
+      "step": 22852
+    },
+    {
+      "epoch": 45.706,
+      "grad_norm": 0.975598931312561,
+      "learning_rate": 2e-05,
+      "loss": 0.03365985,
+      "step": 22853
+    },
+    {
+      "epoch": 45.708,
+      "grad_norm": 1.0220794677734375,
+      "learning_rate": 2e-05,
+      "loss": 0.04530801,
+      "step": 22854
+    },
+    {
+      "epoch": 45.71,
+      "grad_norm": 1.2147080898284912,
+      "learning_rate": 2e-05,
+      "loss": 0.06349006,
+      "step": 22855
+    },
+    {
+      "epoch": 45.712,
+      "grad_norm": 1.1468337774276733,
+      "learning_rate": 2e-05,
+      "loss": 0.04973483,
+      "step": 22856
+    },
+    {
+      "epoch": 45.714,
+      "grad_norm": 1.4470347166061401,
+      "learning_rate": 2e-05,
+      "loss": 0.04426311,
+      "step": 22857
+    },
+    {
+      "epoch": 45.716,
+      "grad_norm": 1.1811226606369019,
+      "learning_rate": 2e-05,
+      "loss": 0.04253963,
+      "step": 22858
+    },
+    {
+      "epoch": 45.718,
+      "grad_norm": 1.0529688596725464,
+      "learning_rate": 2e-05,
+      "loss": 0.04702899,
+      "step": 22859
+    },
+    {
+      "epoch": 45.72,
+      "grad_norm": 1.1447519063949585,
+      "learning_rate": 2e-05,
+      "loss": 0.06699675,
+      "step": 22860
+    },
+    {
+      "epoch": 45.722,
+      "grad_norm": 1.0626235008239746,
+      "learning_rate": 2e-05,
+      "loss": 0.04488755,
+      "step": 22861
+    },
+    {
+      "epoch": 45.724,
+      "grad_norm": 1.3267689943313599,
+      "learning_rate": 2e-05,
+      "loss": 0.05437113,
+      "step": 22862
+    },
+    {
+      "epoch": 45.726,
+      "grad_norm": 1.0401638746261597,
+      "learning_rate": 2e-05,
+      "loss": 0.03539321,
+      "step": 22863
+    },
+    {
+      "epoch": 45.728,
+      "grad_norm": 1.124693512916565,
+      "learning_rate": 2e-05,
+      "loss": 0.03614509,
+      "step": 22864
+    },
+    {
+      "epoch": 45.73,
+      "grad_norm": 1.4313117265701294,
+      "learning_rate": 2e-05,
+      "loss": 0.04222115,
+      "step": 22865
+    },
+    {
+      "epoch": 45.732,
+      "grad_norm": 2.3040075302124023,
+      "learning_rate": 2e-05,
+      "loss": 0.03782343,
+      "step": 22866
+    },
+    {
+      "epoch": 45.734,
+      "grad_norm": 1.12531316280365,
+      "learning_rate": 2e-05,
+      "loss": 0.05212967,
+      "step": 22867
+    },
+    {
+      "epoch": 45.736,
+      "grad_norm": 1.1863642930984497,
+      "learning_rate": 2e-05,
+      "loss": 0.04659764,
+      "step": 22868
+    },
+    {
+      "epoch": 45.738,
+      "grad_norm": 3.5154294967651367,
+      "learning_rate": 2e-05,
+      "loss": 0.0452398,
+      "step": 22869
+    },
+    {
+      "epoch": 45.74,
+      "grad_norm": 1.2669914960861206,
+      "learning_rate": 2e-05,
+      "loss": 0.04121526,
+      "step": 22870
+    },
+    {
+      "epoch": 45.742,
+      "grad_norm": 1.5299526453018188,
+      "learning_rate": 2e-05,
+      "loss": 0.06856737,
+      "step": 22871
+    },
+    {
+      "epoch": 45.744,
+      "grad_norm": 1.943005919456482,
+      "learning_rate": 2e-05,
+      "loss": 0.03736391,
+      "step": 22872
+    },
+    {
+      "epoch": 45.746,
+      "grad_norm": 1.5813325643539429,
+      "learning_rate": 2e-05,
+      "loss": 0.05658036,
+      "step": 22873
+    },
+    {
+      "epoch": 45.748,
+      "grad_norm": 1.1869730949401855,
+      "learning_rate": 2e-05,
+      "loss": 0.04118659,
+      "step": 22874
+    },
+    {
+      "epoch": 45.75,
+      "grad_norm": 1.1262238025665283,
+      "learning_rate": 2e-05,
+      "loss": 0.05163476,
+      "step": 22875
+    },
+    {
+      "epoch": 45.752,
+      "grad_norm": 1.8035818338394165,
+      "learning_rate": 2e-05,
+      "loss": 0.03897575,
+      "step": 22876
+    },
+    {
+      "epoch": 45.754,
+      "grad_norm": 1.2717819213867188,
+      "learning_rate": 2e-05,
+      "loss": 0.06012269,
+      "step": 22877
+    },
+    {
+      "epoch": 45.756,
+      "grad_norm": 1.081311583518982,
+      "learning_rate": 2e-05,
+      "loss": 0.05380928,
+      "step": 22878
+    },
+    {
+      "epoch": 45.758,
+      "grad_norm": 1.1609777212142944,
+      "learning_rate": 2e-05,
+      "loss": 0.04177097,
+      "step": 22879
+    },
+    {
+      "epoch": 45.76,
+      "grad_norm": 2.053823471069336,
+      "learning_rate": 2e-05,
+      "loss": 0.04388625,
+      "step": 22880
+    },
+    {
+      "epoch": 45.762,
+      "grad_norm": 1.1730915307998657,
+      "learning_rate": 2e-05,
+      "loss": 0.03498461,
+      "step": 22881
+    },
+    {
+      "epoch": 45.764,
+      "grad_norm": 1.1314507722854614,
+      "learning_rate": 2e-05,
+      "loss": 0.04693135,
+      "step": 22882
+    },
+    {
+      "epoch": 45.766,
+      "grad_norm": 1.214177131652832,
+      "learning_rate": 2e-05,
+      "loss": 0.05761575,
+      "step": 22883
+    },
+    {
+      "epoch": 45.768,
+      "grad_norm": 1.2264306545257568,
+      "learning_rate": 2e-05,
+      "loss": 0.05583285,
+      "step": 22884
+    },
+    {
+      "epoch": 45.77,
+      "grad_norm": 1.1142363548278809,
+      "learning_rate": 2e-05,
+      "loss": 0.03938077,
+      "step": 22885
+    },
+    {
+      "epoch": 45.772,
+      "grad_norm": 1.2358603477478027,
+      "learning_rate": 2e-05,
+      "loss": 0.05267806,
+      "step": 22886
+    },
+    {
+      "epoch": 45.774,
+      "grad_norm": 2.0392813682556152,
+      "learning_rate": 2e-05,
+      "loss": 0.05020905,
+      "step": 22887
+    },
+    {
+      "epoch": 45.776,
+      "grad_norm": 1.9601068496704102,
+      "learning_rate": 2e-05,
+      "loss": 0.06155182,
+      "step": 22888
+    },
+    {
+      "epoch": 45.778,
+      "grad_norm": 1.2073707580566406,
+      "learning_rate": 2e-05,
+      "loss": 0.04262544,
+      "step": 22889
+    },
+    {
+      "epoch": 45.78,
+      "grad_norm": 1.61870276927948,
+      "learning_rate": 2e-05,
+      "loss": 0.04001756,
+      "step": 22890
+    },
+    {
+      "epoch": 45.782,
+      "grad_norm": 1.62678062915802,
+      "learning_rate": 2e-05,
+      "loss": 0.04077552,
+      "step": 22891
+    },
+    {
+      "epoch": 45.784,
+      "grad_norm": 1.0999619960784912,
+      "learning_rate": 2e-05,
+      "loss": 0.0418282,
+      "step": 22892
+    },
+    {
+      "epoch": 45.786,
+      "grad_norm": 3.3379061222076416,
+      "learning_rate": 2e-05,
+      "loss": 0.06497736,
+      "step": 22893
+    },
+    {
+      "epoch": 45.788,
+      "grad_norm": 1.6847137212753296,
+      "learning_rate": 2e-05,
+      "loss": 0.05242064,
+      "step": 22894
+    },
+    {
+      "epoch": 45.79,
+      "grad_norm": 1.1532042026519775,
+      "learning_rate": 2e-05,
+      "loss": 0.0519854,
+      "step": 22895
+    },
+    {
+      "epoch": 45.792,
+      "grad_norm": 1.1070436239242554,
+      "learning_rate": 2e-05,
+      "loss": 0.04739845,
+      "step": 22896
+    },
+    {
+      "epoch": 45.794,
+      "grad_norm": 1.0771516561508179,
+      "learning_rate": 2e-05,
+      "loss": 0.04649124,
+      "step": 22897
+    },
+    {
+      "epoch": 45.796,
+      "grad_norm": 1.0672521591186523,
+      "learning_rate": 2e-05,
+      "loss": 0.04606741,
+      "step": 22898
+    },
+    {
+      "epoch": 45.798,
+      "grad_norm": 1.7498153448104858,
+      "learning_rate": 2e-05,
+      "loss": 0.0525641,
+      "step": 22899
+    },
+    {
+      "epoch": 45.8,
+      "grad_norm": 3.885802984237671,
+      "learning_rate": 2e-05,
+      "loss": 0.05591378,
+      "step": 22900
+    },
+    {
+      "epoch": 45.802,
+      "grad_norm": 0.9939262866973877,
+      "learning_rate": 2e-05,
+      "loss": 0.0356969,
+      "step": 22901
+    },
+    {
+      "epoch": 45.804,
+      "grad_norm": 1.0101487636566162,
+      "learning_rate": 2e-05,
+      "loss": 0.03555669,
+      "step": 22902
+    },
+    {
+      "epoch": 45.806,
+      "grad_norm": 1.3214828968048096,
+      "learning_rate": 2e-05,
+      "loss": 0.04506984,
+      "step": 22903
+    },
+    {
+      "epoch": 45.808,
+      "grad_norm": 1.2031276226043701,
+      "learning_rate": 2e-05,
+      "loss": 0.04604453,
+      "step": 22904
+    },
+    {
+      "epoch": 45.81,
+      "grad_norm": 0.993742048740387,
+      "learning_rate": 2e-05,
+      "loss": 0.04115344,
+      "step": 22905
+    },
+    {
+      "epoch": 45.812,
+      "grad_norm": 1.3306998014450073,
+      "learning_rate": 2e-05,
+      "loss": 0.05285704,
+      "step": 22906
+    },
+    {
+      "epoch": 45.814,
+      "grad_norm": 1.5680806636810303,
+      "learning_rate": 2e-05,
+      "loss": 0.06785664,
+      "step": 22907
+    },
+    {
+      "epoch": 45.816,
+      "grad_norm": 1.360222339630127,
+      "learning_rate": 2e-05,
+      "loss": 0.05301679,
+      "step": 22908
+    },
+    {
+      "epoch": 45.818,
+      "grad_norm": 1.7606545686721802,
+      "learning_rate": 2e-05,
+      "loss": 0.05103129,
+      "step": 22909
+    },
+    {
+      "epoch": 45.82,
+      "grad_norm": 0.9953403472900391,
+      "learning_rate": 2e-05,
+      "loss": 0.03279923,
+      "step": 22910
+    },
+    {
+      "epoch": 45.822,
+      "grad_norm": 2.362177610397339,
+      "learning_rate": 2e-05,
+      "loss": 0.04696232,
+      "step": 22911
+    },
+    {
+      "epoch": 45.824,
+      "grad_norm": 1.3651307821273804,
+      "learning_rate": 2e-05,
+      "loss": 0.05074629,
+      "step": 22912
+    },
+    {
+      "epoch": 45.826,
+      "grad_norm": 1.4231796264648438,
+      "learning_rate": 2e-05,
+      "loss": 0.04402232,
+      "step": 22913
+    },
+    {
+      "epoch": 45.828,
+      "grad_norm": 1.3840693235397339,
+      "learning_rate": 2e-05,
+      "loss": 0.05228634,
+      "step": 22914
+    },
+    {
+      "epoch": 45.83,
+      "grad_norm": 1.2382279634475708,
+      "learning_rate": 2e-05,
+      "loss": 0.03767666,
+      "step": 22915
+    },
+    {
+      "epoch": 45.832,
+      "grad_norm": 1.1215736865997314,
+      "learning_rate": 2e-05,
+      "loss": 0.03750758,
+      "step": 22916
+    },
+    {
+      "epoch": 45.834,
+      "grad_norm": 1.5836503505706787,
+      "learning_rate": 2e-05,
+      "loss": 0.05021936,
+      "step": 22917
+    },
+    {
+      "epoch": 45.836,
+      "grad_norm": 1.175191879272461,
+      "learning_rate": 2e-05,
+      "loss": 0.05768834,
+      "step": 22918
+    },
+    {
+      "epoch": 45.838,
+      "grad_norm": 2.3862361907958984,
+      "learning_rate": 2e-05,
+      "loss": 0.06385323,
+      "step": 22919
+    },
+    {
+      "epoch": 45.84,
+      "grad_norm": 1.1213129758834839,
+      "learning_rate": 2e-05,
+      "loss": 0.0474508,
+      "step": 22920
+    },
+    {
+      "epoch": 45.842,
+      "grad_norm": 1.1670286655426025,
+      "learning_rate": 2e-05,
+      "loss": 0.05094395,
+      "step": 22921
+    },
+    {
+      "epoch": 45.844,
+      "grad_norm": 1.0269542932510376,
+      "learning_rate": 2e-05,
+      "loss": 0.03310216,
+      "step": 22922
+    },
+    {
+      "epoch": 45.846,
+      "grad_norm": 1.220439076423645,
+      "learning_rate": 2e-05,
+      "loss": 0.05242105,
+      "step": 22923
+    },
+    {
+      "epoch": 45.848,
+      "grad_norm": 1.1493821144104004,
+      "learning_rate": 2e-05,
+      "loss": 0.05762265,
+      "step": 22924
+    },
+    {
+      "epoch": 45.85,
+      "grad_norm": 1.2592790126800537,
+      "learning_rate": 2e-05,
+      "loss": 0.05583587,
+      "step": 22925
+    },
+    {
+      "epoch": 45.852,
+      "grad_norm": 1.1037945747375488,
+      "learning_rate": 2e-05,
+      "loss": 0.05163608,
+      "step": 22926
+    },
+    {
+      "epoch": 45.854,
+      "grad_norm": 1.039101243019104,
+      "learning_rate": 2e-05,
+      "loss": 0.04163966,
+      "step": 22927
+    },
+    {
+      "epoch": 45.856,
+      "grad_norm": 1.087335228919983,
+      "learning_rate": 2e-05,
+      "loss": 0.04357848,
+      "step": 22928
+    },
+    {
+      "epoch": 45.858,
+      "grad_norm": 1.1296930313110352,
+      "learning_rate": 2e-05,
+      "loss": 0.04551987,
+      "step": 22929
+    },
+    {
+      "epoch": 45.86,
+      "grad_norm": 1.1426345109939575,
+      "learning_rate": 2e-05,
+      "loss": 0.05805442,
+      "step": 22930
+    },
+    {
+      "epoch": 45.862,
+      "grad_norm": 2.9095516204833984,
+      "learning_rate": 2e-05,
+      "loss": 0.06481321,
+      "step": 22931
+    },
+    {
+      "epoch": 45.864,
+      "grad_norm": 1.4706426858901978,
+      "learning_rate": 2e-05,
+      "loss": 0.06707012,
+      "step": 22932
+    },
+    {
+      "epoch": 45.866,
+      "grad_norm": 1.0578680038452148,
+      "learning_rate": 2e-05,
+      "loss": 0.03972709,
+      "step": 22933
+    },
+    {
+      "epoch": 45.868,
+      "grad_norm": 1.6905008554458618,
+      "learning_rate": 2e-05,
+      "loss": 0.03994934,
+      "step": 22934
+    },
+    {
+      "epoch": 45.87,
+      "grad_norm": 1.2955090999603271,
+      "learning_rate": 2e-05,
+      "loss": 0.06763369,
+      "step": 22935
+    },
+    {
+      "epoch": 45.872,
+      "grad_norm": 1.0367980003356934,
+      "learning_rate": 2e-05,
+      "loss": 0.03746669,
+      "step": 22936
+    },
+    {
+      "epoch": 45.874,
+      "grad_norm": 1.5050883293151855,
+      "learning_rate": 2e-05,
+      "loss": 0.06505072,
+      "step": 22937
+    },
+    {
+      "epoch": 45.876,
+      "grad_norm": 1.0128837823867798,
+      "learning_rate": 2e-05,
+      "loss": 0.03502842,
+      "step": 22938
+    },
+    {
+      "epoch": 45.878,
+      "grad_norm": 3.16584849357605,
+      "learning_rate": 2e-05,
+      "loss": 0.05251692,
+      "step": 22939
+    },
+    {
+      "epoch": 45.88,
+      "grad_norm": 1.3227653503417969,
+      "learning_rate": 2e-05,
+      "loss": 0.02910842,
+      "step": 22940
+    },
+    {
+      "epoch": 45.882,
+      "grad_norm": 1.2259905338287354,
+      "learning_rate": 2e-05,
+      "loss": 0.06085608,
+      "step": 22941
+    },
+    {
+      "epoch": 45.884,
+      "grad_norm": 1.1765010356903076,
+      "learning_rate": 2e-05,
+      "loss": 0.04740379,
+      "step": 22942
+    },
+    {
+      "epoch": 45.886,
+      "grad_norm": 1.1642879247665405,
+      "learning_rate": 2e-05,
+      "loss": 0.05026664,
+      "step": 22943
+    },
+    {
+      "epoch": 45.888,
+      "grad_norm": 1.3020238876342773,
+      "learning_rate": 2e-05,
+      "loss": 0.06652694,
+      "step": 22944
+    },
+    {
+      "epoch": 45.89,
+      "grad_norm": 1.2435083389282227,
+      "learning_rate": 2e-05,
+      "loss": 0.05145294,
+      "step": 22945
+    },
+    {
+      "epoch": 45.892,
+      "grad_norm": 1.1023855209350586,
+      "learning_rate": 2e-05,
+      "loss": 0.03766239,
+      "step": 22946
+    },
+    {
+      "epoch": 45.894,
+      "grad_norm": 0.9923415780067444,
+      "learning_rate": 2e-05,
+      "loss": 0.03449097,
+      "step": 22947
+    },
+    {
+      "epoch": 45.896,
+      "grad_norm": 1.2809637784957886,
+      "learning_rate": 2e-05,
+      "loss": 0.0631137,
+      "step": 22948
+    },
+    {
+      "epoch": 45.898,
+      "grad_norm": 1.0583648681640625,
+      "learning_rate": 2e-05,
+      "loss": 0.04882384,
+      "step": 22949
+    },
+    {
+      "epoch": 45.9,
+      "grad_norm": 1.2955833673477173,
+      "learning_rate": 2e-05,
+      "loss": 0.0419897,
+      "step": 22950
+    },
+    {
+      "epoch": 45.902,
+      "grad_norm": 1.0099056959152222,
+      "learning_rate": 2e-05,
+      "loss": 0.03409104,
+      "step": 22951
+    },
+    {
+      "epoch": 45.904,
+      "grad_norm": 1.4930360317230225,
+      "learning_rate": 2e-05,
+      "loss": 0.06223853,
+      "step": 22952
+    },
+    {
+      "epoch": 45.906,
+      "grad_norm": 1.2988282442092896,
+      "learning_rate": 2e-05,
+      "loss": 0.04434647,
+      "step": 22953
+    },
+    {
+      "epoch": 45.908,
+      "grad_norm": 1.2352509498596191,
+      "learning_rate": 2e-05,
+      "loss": 0.04742352,
+      "step": 22954
+    },
+    {
+      "epoch": 45.91,
+      "grad_norm": 1.0645052194595337,
+      "learning_rate": 2e-05,
+      "loss": 0.03650692,
+      "step": 22955
+    },
+    {
+      "epoch": 45.912,
+      "grad_norm": 1.607302188873291,
+      "learning_rate": 2e-05,
+      "loss": 0.03560692,
+      "step": 22956
+    },
+    {
+      "epoch": 45.914,
+      "grad_norm": 0.9931079149246216,
+      "learning_rate": 2e-05,
+      "loss": 0.03896011,
+      "step": 22957
+    },
+    {
+      "epoch": 45.916,
+      "grad_norm": 1.081117033958435,
+      "learning_rate": 2e-05,
+      "loss": 0.03914691,
+      "step": 22958
+    },
+    {
+      "epoch": 45.918,
+      "grad_norm": 2.213714361190796,
+      "learning_rate": 2e-05,
+      "loss": 0.07680227,
+      "step": 22959
+    },
+    {
+      "epoch": 45.92,
+      "grad_norm": 1.328762173652649,
+      "learning_rate": 2e-05,
+      "loss": 0.04840521,
+      "step": 22960
+    },
+    {
+      "epoch": 45.922,
+      "grad_norm": 1.999614953994751,
+      "learning_rate": 2e-05,
+      "loss": 0.04826399,
+      "step": 22961
+    },
+    {
+      "epoch": 45.924,
+      "grad_norm": 1.5177665948867798,
+      "learning_rate": 2e-05,
+      "loss": 0.04357665,
+      "step": 22962
+    },
+    {
+      "epoch": 45.926,
+      "grad_norm": 1.822375774383545,
+      "learning_rate": 2e-05,
+      "loss": 0.05791084,
+      "step": 22963
+    },
+    {
+      "epoch": 45.928,
+      "grad_norm": 1.217348575592041,
+      "learning_rate": 2e-05,
+      "loss": 0.03876396,
+      "step": 22964
+    },
+    {
+      "epoch": 45.93,
+      "grad_norm": 1.5601190328598022,
+      "learning_rate": 2e-05,
+      "loss": 0.06204118,
+      "step": 22965
+    },
+    {
+      "epoch": 45.932,
+      "grad_norm": 2.090487480163574,
+      "learning_rate": 2e-05,
+      "loss": 0.0422308,
+      "step": 22966
+    },
+    {
+      "epoch": 45.934,
+      "grad_norm": 1.2206273078918457,
+      "learning_rate": 2e-05,
+      "loss": 0.04056178,
+      "step": 22967
+    },
+    {
+      "epoch": 45.936,
+      "grad_norm": 1.1585288047790527,
+      "learning_rate": 2e-05,
+      "loss": 0.03958793,
+      "step": 22968
+    },
+    {
+      "epoch": 45.938,
+      "grad_norm": 1.2946537733078003,
+      "learning_rate": 2e-05,
+      "loss": 0.04267664,
+      "step": 22969
+    },
+    {
+      "epoch": 45.94,
+      "grad_norm": 1.4737160205841064,
+      "learning_rate": 2e-05,
+      "loss": 0.07163599,
+      "step": 22970
+    },
+    {
+      "epoch": 45.942,
+      "grad_norm": 1.1721677780151367,
+      "learning_rate": 2e-05,
+      "loss": 0.04065248,
+      "step": 22971
+    },
+    {
+      "epoch": 45.944,
+      "grad_norm": 1.7377208471298218,
+      "learning_rate": 2e-05,
+      "loss": 0.05867632,
+      "step": 22972
+    },
+    {
+      "epoch": 45.946,
+      "grad_norm": 1.7599306106567383,
+      "learning_rate": 2e-05,
+      "loss": 0.04502322,
+      "step": 22973
+    },
+    {
+      "epoch": 45.948,
+      "grad_norm": 1.2876641750335693,
+      "learning_rate": 2e-05,
+      "loss": 0.0509927,
+      "step": 22974
+    },
+    {
+      "epoch": 45.95,
+      "grad_norm": 1.1776494979858398,
+      "learning_rate": 2e-05,
+      "loss": 0.05537301,
+      "step": 22975
+    },
+    {
+      "epoch": 45.952,
+      "grad_norm": 1.3018521070480347,
+      "learning_rate": 2e-05,
+      "loss": 0.0640357,
+      "step": 22976
+    },
+    {
+      "epoch": 45.954,
+      "grad_norm": 1.2216527462005615,
+      "learning_rate": 2e-05,
+      "loss": 0.0506551,
+      "step": 22977
+    },
+    {
+      "epoch": 45.956,
+      "grad_norm": 1.313719630241394,
+      "learning_rate": 2e-05,
+      "loss": 0.05280958,
+      "step": 22978
+    },
+    {
+      "epoch": 45.958,
+      "grad_norm": 1.1301958560943604,
+      "learning_rate": 2e-05,
+      "loss": 0.04503809,
+      "step": 22979
+    },
+    {
+      "epoch": 45.96,
+      "grad_norm": 1.7027695178985596,
+      "learning_rate": 2e-05,
+      "loss": 0.05137224,
+      "step": 22980
+    },
+    {
+      "epoch": 45.962,
+      "grad_norm": 2.217540740966797,
+      "learning_rate": 2e-05,
+      "loss": 0.05492135,
+      "step": 22981
+    },
+    {
+      "epoch": 45.964,
+      "grad_norm": 1.163046956062317,
+      "learning_rate": 2e-05,
+      "loss": 0.03381425,
+      "step": 22982
+    },
+    {
+      "epoch": 45.966,
+      "grad_norm": 1.2386566400527954,
+      "learning_rate": 2e-05,
+      "loss": 0.04658375,
+      "step": 22983
+    },
+    {
+      "epoch": 45.968,
+      "grad_norm": 1.490594506263733,
+      "learning_rate": 2e-05,
+      "loss": 0.06036313,
+      "step": 22984
+    },
+    {
+      "epoch": 45.97,
+      "grad_norm": 1.3342573642730713,
+      "learning_rate": 2e-05,
+      "loss": 0.04109773,
+      "step": 22985
+    },
+    {
+      "epoch": 45.972,
+      "grad_norm": 2.1110265254974365,
+      "learning_rate": 2e-05,
+      "loss": 0.04837161,
+      "step": 22986
+    },
+    {
+      "epoch": 45.974,
+      "grad_norm": 1.1389340162277222,
+      "learning_rate": 2e-05,
+      "loss": 0.04660567,
+      "step": 22987
+    },
+    {
+      "epoch": 45.976,
+      "grad_norm": 1.324904441833496,
+      "learning_rate": 2e-05,
+      "loss": 0.05071661,
+      "step": 22988
+    },
+    {
+      "epoch": 45.978,
+      "grad_norm": 1.7974005937576294,
+      "learning_rate": 2e-05,
+      "loss": 0.05388502,
+      "step": 22989
+    },
+    {
+      "epoch": 45.98,
+      "grad_norm": 1.4797214269638062,
+      "learning_rate": 2e-05,
+      "loss": 0.04529134,
+      "step": 22990
+    },
+    {
+      "epoch": 45.982,
+      "grad_norm": 2.093899965286255,
+      "learning_rate": 2e-05,
+      "loss": 0.04731727,
+      "step": 22991
+    },
+    {
+      "epoch": 45.984,
+      "grad_norm": 1.1308255195617676,
+      "learning_rate": 2e-05,
+      "loss": 0.04801143,
+      "step": 22992
+    },
+    {
+      "epoch": 45.986,
+      "grad_norm": 1.4428658485412598,
+      "learning_rate": 2e-05,
+      "loss": 0.04077057,
+      "step": 22993
+    },
+    {
+      "epoch": 45.988,
+      "grad_norm": 1.423751950263977,
+      "learning_rate": 2e-05,
+      "loss": 0.05186245,
+      "step": 22994
+    },
+    {
+      "epoch": 45.99,
+      "grad_norm": 1.1405518054962158,
+      "learning_rate": 2e-05,
+      "loss": 0.04192315,
+      "step": 22995
+    },
+    {
+      "epoch": 45.992,
+      "grad_norm": 1.020820140838623,
+      "learning_rate": 2e-05,
+      "loss": 0.04345158,
+      "step": 22996
+    },
+    {
+      "epoch": 45.994,
+      "grad_norm": 1.4293239116668701,
+      "learning_rate": 2e-05,
+      "loss": 0.04355782,
+      "step": 22997
+    },
+    {
+      "epoch": 45.996,
+      "grad_norm": 2.0200562477111816,
+      "learning_rate": 2e-05,
+      "loss": 0.06576998,
+      "step": 22998
+    },
+    {
+      "epoch": 45.998,
+      "grad_norm": 1.1396820545196533,
+      "learning_rate": 2e-05,
+      "loss": 0.0453399,
+      "step": 22999
+    },
+    {
+      "epoch": 46.0,
+      "grad_norm": 1.615691900253296,
+      "learning_rate": 2e-05,
+      "loss": 0.04784571,
+      "step": 23000
+    },
+    {
+      "epoch": 46.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9840319361277445,
+        "Equal_1": 0.998,
+        "Equal_2": 0.9820359281437125,
+        "Equal_3": 0.9920159680638723,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 1.0,
+        "Parallel_1": 0.9959919839679359,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.998,
+        "Perpendicular_3": 0.905811623246493,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 1.0,
+        "PointLiesOnCircle_3": 0.992,
+        "PointLiesOnLine_1": 0.9939879759519038,
+        "PointLiesOnLine_2": 0.9939879759519038,
+        "PointLiesOnLine_3": 0.9860279441117764
+      },
+      "eval_runtime": 324.2615,
+      "eval_samples_per_second": 32.381,
+      "eval_steps_per_second": 0.648,
+      "step": 23000
+    },
+    {
+      "epoch": 46.002,
+      "grad_norm": 1.4302409887313843,
+      "learning_rate": 2e-05,
+      "loss": 0.05861004,
+      "step": 23001
+    },
+    {
+      "epoch": 46.004,
+      "grad_norm": 1.39664626121521,
+      "learning_rate": 2e-05,
+      "loss": 0.05934343,
+      "step": 23002
+    },
+    {
+      "epoch": 46.006,
+      "grad_norm": 1.1906030178070068,
+      "learning_rate": 2e-05,
+      "loss": 0.03054802,
+      "step": 23003
+    },
+    {
+      "epoch": 46.008,
+      "grad_norm": 1.7340916395187378,
+      "learning_rate": 2e-05,
+      "loss": 0.05384779,
+      "step": 23004
+    },
+    {
+      "epoch": 46.01,
+      "grad_norm": 1.309027075767517,
+      "learning_rate": 2e-05,
+      "loss": 0.04369318,
+      "step": 23005
+    },
+    {
+      "epoch": 46.012,
+      "grad_norm": 1.002219796180725,
+      "learning_rate": 2e-05,
+      "loss": 0.03218366,
+      "step": 23006
+    },
+    {
+      "epoch": 46.014,
+      "grad_norm": 1.6894913911819458,
+      "learning_rate": 2e-05,
+      "loss": 0.05528177,
+      "step": 23007
+    },
+    {
+      "epoch": 46.016,
+      "grad_norm": 1.18147873878479,
+      "learning_rate": 2e-05,
+      "loss": 0.03384054,
+      "step": 23008
+    },
+    {
+      "epoch": 46.018,
+      "grad_norm": 1.331606149673462,
+      "learning_rate": 2e-05,
+      "loss": 0.05830061,
+      "step": 23009
+    },
+    {
+      "epoch": 46.02,
+      "grad_norm": 1.5706523656845093,
+      "learning_rate": 2e-05,
+      "loss": 0.05562139,
+      "step": 23010
+    },
+    {
+      "epoch": 46.022,
+      "grad_norm": 1.2435215711593628,
+      "learning_rate": 2e-05,
+      "loss": 0.05742179,
+      "step": 23011
+    },
+    {
+      "epoch": 46.024,
+      "grad_norm": 1.2738981246948242,
+      "learning_rate": 2e-05,
+      "loss": 0.04086342,
+      "step": 23012
+    },
+    {
+      "epoch": 46.026,
+      "grad_norm": 1.4569175243377686,
+      "learning_rate": 2e-05,
+      "loss": 0.04479681,
+      "step": 23013
+    },
+    {
+      "epoch": 46.028,
+      "grad_norm": 1.147181749343872,
+      "learning_rate": 2e-05,
+      "loss": 0.04108036,
+      "step": 23014
+    },
+    {
+      "epoch": 46.03,
+      "grad_norm": 1.1991140842437744,
+      "learning_rate": 2e-05,
+      "loss": 0.04342316,
+      "step": 23015
+    },
+    {
+      "epoch": 46.032,
+      "grad_norm": 1.3934859037399292,
+      "learning_rate": 2e-05,
+      "loss": 0.05477795,
+      "step": 23016
+    },
+    {
+      "epoch": 46.034,
+      "grad_norm": 1.2585614919662476,
+      "learning_rate": 2e-05,
+      "loss": 0.04793346,
+      "step": 23017
+    },
+    {
+      "epoch": 46.036,
+      "grad_norm": 1.1003919839859009,
+      "learning_rate": 2e-05,
+      "loss": 0.05579524,
+      "step": 23018
+    },
+    {
+      "epoch": 46.038,
+      "grad_norm": 1.2833575010299683,
+      "learning_rate": 2e-05,
+      "loss": 0.05675605,
+      "step": 23019
+    },
+    {
+      "epoch": 46.04,
+      "grad_norm": 3.453655242919922,
+      "learning_rate": 2e-05,
+      "loss": 0.05313053,
+      "step": 23020
+    },
+    {
+      "epoch": 46.042,
+      "grad_norm": 1.191473126411438,
+      "learning_rate": 2e-05,
+      "loss": 0.04780373,
+      "step": 23021
+    },
+    {
+      "epoch": 46.044,
+      "grad_norm": 1.29380202293396,
+      "learning_rate": 2e-05,
+      "loss": 0.04844226,
+      "step": 23022
+    },
+    {
+      "epoch": 46.046,
+      "grad_norm": 1.3847426176071167,
+      "learning_rate": 2e-05,
+      "loss": 0.05267187,
+      "step": 23023
+    },
+    {
+      "epoch": 46.048,
+      "grad_norm": 2.212348222732544,
+      "learning_rate": 2e-05,
+      "loss": 0.0534815,
+      "step": 23024
+    },
+    {
+      "epoch": 46.05,
+      "grad_norm": 1.3633710145950317,
+      "learning_rate": 2e-05,
+      "loss": 0.06192131,
+      "step": 23025
+    },
+    {
+      "epoch": 46.052,
+      "grad_norm": 1.212673544883728,
+      "learning_rate": 2e-05,
+      "loss": 0.03544668,
+      "step": 23026
+    },
+    {
+      "epoch": 46.054,
+      "grad_norm": 1.0772227048873901,
+      "learning_rate": 2e-05,
+      "loss": 0.03711303,
+      "step": 23027
+    },
+    {
+      "epoch": 46.056,
+      "grad_norm": 2.53438401222229,
+      "learning_rate": 2e-05,
+      "loss": 0.05098771,
+      "step": 23028
+    },
+    {
+      "epoch": 46.058,
+      "grad_norm": 1.814626932144165,
+      "learning_rate": 2e-05,
+      "loss": 0.04822916,
+      "step": 23029
+    },
+    {
+      "epoch": 46.06,
+      "grad_norm": 1.6201591491699219,
+      "learning_rate": 2e-05,
+      "loss": 0.0526498,
+      "step": 23030
+    },
+    {
+      "epoch": 46.062,
+      "grad_norm": 1.464896559715271,
+      "learning_rate": 2e-05,
+      "loss": 0.05728604,
+      "step": 23031
+    },
+    {
+      "epoch": 46.064,
+      "grad_norm": 1.0204657316207886,
+      "learning_rate": 2e-05,
+      "loss": 0.04259542,
+      "step": 23032
+    },
+    {
+      "epoch": 46.066,
+      "grad_norm": 1.1545565128326416,
+      "learning_rate": 2e-05,
+      "loss": 0.05285469,
+      "step": 23033
+    },
+    {
+      "epoch": 46.068,
+      "grad_norm": 1.0158534049987793,
+      "learning_rate": 2e-05,
+      "loss": 0.03832763,
+      "step": 23034
+    },
+    {
+      "epoch": 46.07,
+      "grad_norm": 1.1786905527114868,
+      "learning_rate": 2e-05,
+      "loss": 0.03449775,
+      "step": 23035
+    },
+    {
+      "epoch": 46.072,
+      "grad_norm": 1.2744739055633545,
+      "learning_rate": 2e-05,
+      "loss": 0.04379936,
+      "step": 23036
+    },
+    {
+      "epoch": 46.074,
+      "grad_norm": 1.868301510810852,
+      "learning_rate": 2e-05,
+      "loss": 0.0644619,
+      "step": 23037
+    },
+    {
+      "epoch": 46.076,
+      "grad_norm": 1.1725568771362305,
+      "learning_rate": 2e-05,
+      "loss": 0.04825684,
+      "step": 23038
+    },
+    {
+      "epoch": 46.078,
+      "grad_norm": 1.6947202682495117,
+      "learning_rate": 2e-05,
+      "loss": 0.06334668,
+      "step": 23039
+    },
+    {
+      "epoch": 46.08,
+      "grad_norm": 1.4566727876663208,
+      "learning_rate": 2e-05,
+      "loss": 0.05566946,
+      "step": 23040
+    },
+    {
+      "epoch": 46.082,
+      "grad_norm": 1.2158526182174683,
+      "learning_rate": 2e-05,
+      "loss": 0.03828437,
+      "step": 23041
+    },
+    {
+      "epoch": 46.084,
+      "grad_norm": 1.1454522609710693,
+      "learning_rate": 2e-05,
+      "loss": 0.04942612,
+      "step": 23042
+    },
+    {
+      "epoch": 46.086,
+      "grad_norm": 1.1992695331573486,
+      "learning_rate": 2e-05,
+      "loss": 0.04454256,
+      "step": 23043
+    },
+    {
+      "epoch": 46.088,
+      "grad_norm": 1.242703914642334,
+      "learning_rate": 2e-05,
+      "loss": 0.04919573,
+      "step": 23044
+    },
+    {
+      "epoch": 46.09,
+      "grad_norm": 1.1189968585968018,
+      "learning_rate": 2e-05,
+      "loss": 0.04536124,
+      "step": 23045
+    },
+    {
+      "epoch": 46.092,
+      "grad_norm": 1.1488397121429443,
+      "learning_rate": 2e-05,
+      "loss": 0.05430396,
+      "step": 23046
+    },
+    {
+      "epoch": 46.094,
+      "grad_norm": 1.281959891319275,
+      "learning_rate": 2e-05,
+      "loss": 0.06264284,
+      "step": 23047
+    },
+    {
+      "epoch": 46.096,
+      "grad_norm": 1.4672093391418457,
+      "learning_rate": 2e-05,
+      "loss": 0.06311803,
+      "step": 23048
+    },
+    {
+      "epoch": 46.098,
+      "grad_norm": 1.124530553817749,
+      "learning_rate": 2e-05,
+      "loss": 0.03884341,
+      "step": 23049
+    },
+    {
+      "epoch": 46.1,
+      "grad_norm": 1.1475732326507568,
+      "learning_rate": 2e-05,
+      "loss": 0.05629781,
+      "step": 23050
+    },
+    {
+      "epoch": 46.102,
+      "grad_norm": 1.4845279455184937,
+      "learning_rate": 2e-05,
+      "loss": 0.05716734,
+      "step": 23051
+    },
+    {
+      "epoch": 46.104,
+      "grad_norm": 0.9782443642616272,
+      "learning_rate": 2e-05,
+      "loss": 0.0321027,
+      "step": 23052
+    },
+    {
+      "epoch": 46.106,
+      "grad_norm": 1.0824551582336426,
+      "learning_rate": 2e-05,
+      "loss": 0.04053795,
+      "step": 23053
+    },
+    {
+      "epoch": 46.108,
+      "grad_norm": 1.1406644582748413,
+      "learning_rate": 2e-05,
+      "loss": 0.03841585,
+      "step": 23054
+    },
+    {
+      "epoch": 46.11,
+      "grad_norm": 1.2297335863113403,
+      "learning_rate": 2e-05,
+      "loss": 0.05340029,
+      "step": 23055
+    },
+    {
+      "epoch": 46.112,
+      "grad_norm": 1.1259406805038452,
+      "learning_rate": 2e-05,
+      "loss": 0.05276688,
+      "step": 23056
+    },
+    {
+      "epoch": 46.114,
+      "grad_norm": 2.047508478164673,
+      "learning_rate": 2e-05,
+      "loss": 0.05007156,
+      "step": 23057
+    },
+    {
+      "epoch": 46.116,
+      "grad_norm": 1.338058352470398,
+      "learning_rate": 2e-05,
+      "loss": 0.04387624,
+      "step": 23058
+    },
+    {
+      "epoch": 46.118,
+      "grad_norm": 1.2544599771499634,
+      "learning_rate": 2e-05,
+      "loss": 0.04295979,
+      "step": 23059
+    },
+    {
+      "epoch": 46.12,
+      "grad_norm": 1.185205101966858,
+      "learning_rate": 2e-05,
+      "loss": 0.05217368,
+      "step": 23060
+    },
+    {
+      "epoch": 46.122,
+      "grad_norm": 1.1495543718338013,
+      "learning_rate": 2e-05,
+      "loss": 0.042289,
+      "step": 23061
+    },
+    {
+      "epoch": 46.124,
+      "grad_norm": 1.7538362741470337,
+      "learning_rate": 2e-05,
+      "loss": 0.0680445,
+      "step": 23062
+    },
+    {
+      "epoch": 46.126,
+      "grad_norm": 1.0621302127838135,
+      "learning_rate": 2e-05,
+      "loss": 0.04465366,
+      "step": 23063
+    },
+    {
+      "epoch": 46.128,
+      "grad_norm": 1.2388436794281006,
+      "learning_rate": 2e-05,
+      "loss": 0.05579409,
+      "step": 23064
+    },
+    {
+      "epoch": 46.13,
+      "grad_norm": 1.4218683242797852,
+      "learning_rate": 2e-05,
+      "loss": 0.0443549,
+      "step": 23065
+    },
+    {
+      "epoch": 46.132,
+      "grad_norm": 1.1729727983474731,
+      "learning_rate": 2e-05,
+      "loss": 0.05611319,
+      "step": 23066
+    },
+    {
+      "epoch": 46.134,
+      "grad_norm": 1.1419835090637207,
+      "learning_rate": 2e-05,
+      "loss": 0.04782356,
+      "step": 23067
+    },
+    {
+      "epoch": 46.136,
+      "grad_norm": 1.0463011264801025,
+      "learning_rate": 2e-05,
+      "loss": 0.04206122,
+      "step": 23068
+    },
+    {
+      "epoch": 46.138,
+      "grad_norm": 1.1902068853378296,
+      "learning_rate": 2e-05,
+      "loss": 0.0524741,
+      "step": 23069
+    },
+    {
+      "epoch": 46.14,
+      "grad_norm": 1.0246272087097168,
+      "learning_rate": 2e-05,
+      "loss": 0.03112867,
+      "step": 23070
+    },
+    {
+      "epoch": 46.142,
+      "grad_norm": 1.05487859249115,
+      "learning_rate": 2e-05,
+      "loss": 0.04408978,
+      "step": 23071
+    },
+    {
+      "epoch": 46.144,
+      "grad_norm": 1.0214228630065918,
+      "learning_rate": 2e-05,
+      "loss": 0.03995924,
+      "step": 23072
+    },
+    {
+      "epoch": 46.146,
+      "grad_norm": 1.0368109941482544,
+      "learning_rate": 2e-05,
+      "loss": 0.04691656,
+      "step": 23073
+    },
+    {
+      "epoch": 46.148,
+      "grad_norm": 1.1551309823989868,
+      "learning_rate": 2e-05,
+      "loss": 0.04764878,
+      "step": 23074
+    },
+    {
+      "epoch": 46.15,
+      "grad_norm": 1.2423347234725952,
+      "learning_rate": 2e-05,
+      "loss": 0.05095094,
+      "step": 23075
+    },
+    {
+      "epoch": 46.152,
+      "grad_norm": 1.4114395380020142,
+      "learning_rate": 2e-05,
+      "loss": 0.05528051,
+      "step": 23076
+    },
+    {
+      "epoch": 46.154,
+      "grad_norm": 0.9848634600639343,
+      "learning_rate": 2e-05,
+      "loss": 0.04308025,
+      "step": 23077
+    },
+    {
+      "epoch": 46.156,
+      "grad_norm": 1.071808099746704,
+      "learning_rate": 2e-05,
+      "loss": 0.0475583,
+      "step": 23078
+    },
+    {
+      "epoch": 46.158,
+      "grad_norm": 1.093819499015808,
+      "learning_rate": 2e-05,
+      "loss": 0.03855433,
+      "step": 23079
+    },
+    {
+      "epoch": 46.16,
+      "grad_norm": 1.9816279411315918,
+      "learning_rate": 2e-05,
+      "loss": 0.04996374,
+      "step": 23080
+    },
+    {
+      "epoch": 46.162,
+      "grad_norm": 1.2489104270935059,
+      "learning_rate": 2e-05,
+      "loss": 0.05128529,
+      "step": 23081
+    },
+    {
+      "epoch": 46.164,
+      "grad_norm": 1.3830829858779907,
+      "learning_rate": 2e-05,
+      "loss": 0.05532495,
+      "step": 23082
+    },
+    {
+      "epoch": 46.166,
+      "grad_norm": 1.2296466827392578,
+      "learning_rate": 2e-05,
+      "loss": 0.04965352,
+      "step": 23083
+    },
+    {
+      "epoch": 46.168,
+      "grad_norm": 1.0486325025558472,
+      "learning_rate": 2e-05,
+      "loss": 0.03422453,
+      "step": 23084
+    },
+    {
+      "epoch": 46.17,
+      "grad_norm": 1.4689314365386963,
+      "learning_rate": 2e-05,
+      "loss": 0.05122601,
+      "step": 23085
+    },
+    {
+      "epoch": 46.172,
+      "grad_norm": 1.9588061571121216,
+      "learning_rate": 2e-05,
+      "loss": 0.04775591,
+      "step": 23086
+    },
+    {
+      "epoch": 46.174,
+      "grad_norm": 1.2059226036071777,
+      "learning_rate": 2e-05,
+      "loss": 0.04841192,
+      "step": 23087
+    },
+    {
+      "epoch": 46.176,
+      "grad_norm": 1.1085312366485596,
+      "learning_rate": 2e-05,
+      "loss": 0.04657616,
+      "step": 23088
+    },
+    {
+      "epoch": 46.178,
+      "grad_norm": 2.187915563583374,
+      "learning_rate": 2e-05,
+      "loss": 0.05317451,
+      "step": 23089
+    },
+    {
+      "epoch": 46.18,
+      "grad_norm": 1.0216542482376099,
+      "learning_rate": 2e-05,
+      "loss": 0.0392207,
+      "step": 23090
+    },
+    {
+      "epoch": 46.182,
+      "grad_norm": 1.3469269275665283,
+      "learning_rate": 2e-05,
+      "loss": 0.05570158,
+      "step": 23091
+    },
+    {
+      "epoch": 46.184,
+      "grad_norm": 1.4849958419799805,
+      "learning_rate": 2e-05,
+      "loss": 0.04750857,
+      "step": 23092
+    },
+    {
+      "epoch": 46.186,
+      "grad_norm": 1.8725115060806274,
+      "learning_rate": 2e-05,
+      "loss": 0.06257782,
+      "step": 23093
+    },
+    {
+      "epoch": 46.188,
+      "grad_norm": 1.5166374444961548,
+      "learning_rate": 2e-05,
+      "loss": 0.04668888,
+      "step": 23094
+    },
+    {
+      "epoch": 46.19,
+      "grad_norm": 1.2905293703079224,
+      "learning_rate": 2e-05,
+      "loss": 0.04647534,
+      "step": 23095
+    },
+    {
+      "epoch": 46.192,
+      "grad_norm": 1.208664894104004,
+      "learning_rate": 2e-05,
+      "loss": 0.05198023,
+      "step": 23096
+    },
+    {
+      "epoch": 46.194,
+      "grad_norm": 1.0935642719268799,
+      "learning_rate": 2e-05,
+      "loss": 0.05911352,
+      "step": 23097
+    },
+    {
+      "epoch": 46.196,
+      "grad_norm": 1.104860782623291,
+      "learning_rate": 2e-05,
+      "loss": 0.03793868,
+      "step": 23098
+    },
+    {
+      "epoch": 46.198,
+      "grad_norm": 1.06515371799469,
+      "learning_rate": 2e-05,
+      "loss": 0.04790756,
+      "step": 23099
+    },
+    {
+      "epoch": 46.2,
+      "grad_norm": 1.6914348602294922,
+      "learning_rate": 2e-05,
+      "loss": 0.04272515,
+      "step": 23100
+    },
+    {
+      "epoch": 46.202,
+      "grad_norm": 1.292098879814148,
+      "learning_rate": 2e-05,
+      "loss": 0.04760931,
+      "step": 23101
+    },
+    {
+      "epoch": 46.204,
+      "grad_norm": 1.197635531425476,
+      "learning_rate": 2e-05,
+      "loss": 0.05021203,
+      "step": 23102
+    },
+    {
+      "epoch": 46.206,
+      "grad_norm": 1.267522931098938,
+      "learning_rate": 2e-05,
+      "loss": 0.06113753,
+      "step": 23103
+    },
+    {
+      "epoch": 46.208,
+      "grad_norm": 0.9828411936759949,
+      "learning_rate": 2e-05,
+      "loss": 0.04029978,
+      "step": 23104
+    },
+    {
+      "epoch": 46.21,
+      "grad_norm": 1.4888331890106201,
+      "learning_rate": 2e-05,
+      "loss": 0.05187474,
+      "step": 23105
+    },
+    {
+      "epoch": 46.212,
+      "grad_norm": 1.055316686630249,
+      "learning_rate": 2e-05,
+      "loss": 0.04340184,
+      "step": 23106
+    },
+    {
+      "epoch": 46.214,
+      "grad_norm": 1.2184720039367676,
+      "learning_rate": 2e-05,
+      "loss": 0.0478823,
+      "step": 23107
+    },
+    {
+      "epoch": 46.216,
+      "grad_norm": 1.2196934223175049,
+      "learning_rate": 2e-05,
+      "loss": 0.05798684,
+      "step": 23108
+    },
+    {
+      "epoch": 46.218,
+      "grad_norm": 1.194522738456726,
+      "learning_rate": 2e-05,
+      "loss": 0.05176704,
+      "step": 23109
+    },
+    {
+      "epoch": 46.22,
+      "grad_norm": 1.0654971599578857,
+      "learning_rate": 2e-05,
+      "loss": 0.03646301,
+      "step": 23110
+    },
+    {
+      "epoch": 46.222,
+      "grad_norm": 1.3132420778274536,
+      "learning_rate": 2e-05,
+      "loss": 0.03887681,
+      "step": 23111
+    },
+    {
+      "epoch": 46.224,
+      "grad_norm": 1.870571255683899,
+      "learning_rate": 2e-05,
+      "loss": 0.0637666,
+      "step": 23112
+    },
+    {
+      "epoch": 46.226,
+      "grad_norm": 1.1391980648040771,
+      "learning_rate": 2e-05,
+      "loss": 0.04590259,
+      "step": 23113
+    },
+    {
+      "epoch": 46.228,
+      "grad_norm": 1.1913096904754639,
+      "learning_rate": 2e-05,
+      "loss": 0.03562557,
+      "step": 23114
+    },
+    {
+      "epoch": 46.23,
+      "grad_norm": 3.3259854316711426,
+      "learning_rate": 2e-05,
+      "loss": 0.05996282,
+      "step": 23115
+    },
+    {
+      "epoch": 46.232,
+      "grad_norm": 1.2023781538009644,
+      "learning_rate": 2e-05,
+      "loss": 0.05674819,
+      "step": 23116
+    },
+    {
+      "epoch": 46.234,
+      "grad_norm": 1.2463189363479614,
+      "learning_rate": 2e-05,
+      "loss": 0.05823443,
+      "step": 23117
+    },
+    {
+      "epoch": 46.236,
+      "grad_norm": 1.352786660194397,
+      "learning_rate": 2e-05,
+      "loss": 0.0599811,
+      "step": 23118
+    },
+    {
+      "epoch": 46.238,
+      "grad_norm": 1.1087812185287476,
+      "learning_rate": 2e-05,
+      "loss": 0.05076696,
+      "step": 23119
+    },
+    {
+      "epoch": 46.24,
+      "grad_norm": 1.2660101652145386,
+      "learning_rate": 2e-05,
+      "loss": 0.0315996,
+      "step": 23120
+    },
+    {
+      "epoch": 46.242,
+      "grad_norm": 1.2009081840515137,
+      "learning_rate": 2e-05,
+      "loss": 0.04985572,
+      "step": 23121
+    },
+    {
+      "epoch": 46.244,
+      "grad_norm": 0.9989933371543884,
+      "learning_rate": 2e-05,
+      "loss": 0.04121695,
+      "step": 23122
+    },
+    {
+      "epoch": 46.246,
+      "grad_norm": 2.8735687732696533,
+      "learning_rate": 2e-05,
+      "loss": 0.07063636,
+      "step": 23123
+    },
+    {
+      "epoch": 46.248,
+      "grad_norm": 1.0923492908477783,
+      "learning_rate": 2e-05,
+      "loss": 0.04320874,
+      "step": 23124
+    },
+    {
+      "epoch": 46.25,
+      "grad_norm": 1.0142641067504883,
+      "learning_rate": 2e-05,
+      "loss": 0.03465679,
+      "step": 23125
+    },
+    {
+      "epoch": 46.252,
+      "grad_norm": 1.1704773902893066,
+      "learning_rate": 2e-05,
+      "loss": 0.04421233,
+      "step": 23126
+    },
+    {
+      "epoch": 46.254,
+      "grad_norm": 1.2731939554214478,
+      "learning_rate": 2e-05,
+      "loss": 0.07911802,
+      "step": 23127
+    },
+    {
+      "epoch": 46.256,
+      "grad_norm": 2.1010546684265137,
+      "learning_rate": 2e-05,
+      "loss": 0.04964682,
+      "step": 23128
+    },
+    {
+      "epoch": 46.258,
+      "grad_norm": 3.078134536743164,
+      "learning_rate": 2e-05,
+      "loss": 0.0332694,
+      "step": 23129
+    },
+    {
+      "epoch": 46.26,
+      "grad_norm": 1.6000622510910034,
+      "learning_rate": 2e-05,
+      "loss": 0.05290427,
+      "step": 23130
+    },
+    {
+      "epoch": 46.262,
+      "grad_norm": 2.217181921005249,
+      "learning_rate": 2e-05,
+      "loss": 0.04330307,
+      "step": 23131
+    },
+    {
+      "epoch": 46.264,
+      "grad_norm": 1.268926739692688,
+      "learning_rate": 2e-05,
+      "loss": 0.05655915,
+      "step": 23132
+    },
+    {
+      "epoch": 46.266,
+      "grad_norm": 5.429689407348633,
+      "learning_rate": 2e-05,
+      "loss": 0.05056974,
+      "step": 23133
+    },
+    {
+      "epoch": 46.268,
+      "grad_norm": 4.090033054351807,
+      "learning_rate": 2e-05,
+      "loss": 0.05439257,
+      "step": 23134
+    },
+    {
+      "epoch": 46.27,
+      "grad_norm": 4.116130352020264,
+      "learning_rate": 2e-05,
+      "loss": 0.06643784,
+      "step": 23135
+    },
+    {
+      "epoch": 46.272,
+      "grad_norm": 1.2675787210464478,
+      "learning_rate": 2e-05,
+      "loss": 0.05370174,
+      "step": 23136
+    },
+    {
+      "epoch": 46.274,
+      "grad_norm": 1.1625384092330933,
+      "learning_rate": 2e-05,
+      "loss": 0.04453802,
+      "step": 23137
+    },
+    {
+      "epoch": 46.276,
+      "grad_norm": 1.1202532052993774,
+      "learning_rate": 2e-05,
+      "loss": 0.04463059,
+      "step": 23138
+    },
+    {
+      "epoch": 46.278,
+      "grad_norm": 1.0193384885787964,
+      "learning_rate": 2e-05,
+      "loss": 0.03918165,
+      "step": 23139
+    },
+    {
+      "epoch": 46.28,
+      "grad_norm": 0.9413808584213257,
+      "learning_rate": 2e-05,
+      "loss": 0.02721583,
+      "step": 23140
+    },
+    {
+      "epoch": 46.282,
+      "grad_norm": 1.1726583242416382,
+      "learning_rate": 2e-05,
+      "loss": 0.03906541,
+      "step": 23141
+    },
+    {
+      "epoch": 46.284,
+      "grad_norm": 1.1463528871536255,
+      "learning_rate": 2e-05,
+      "loss": 0.04882437,
+      "step": 23142
+    },
+    {
+      "epoch": 46.286,
+      "grad_norm": 1.02104651927948,
+      "learning_rate": 2e-05,
+      "loss": 0.04882137,
+      "step": 23143
+    },
+    {
+      "epoch": 46.288,
+      "grad_norm": 2.9313247203826904,
+      "learning_rate": 2e-05,
+      "loss": 0.04990475,
+      "step": 23144
+    },
+    {
+      "epoch": 46.29,
+      "grad_norm": 1.2045042514801025,
+      "learning_rate": 2e-05,
+      "loss": 0.04847421,
+      "step": 23145
+    },
+    {
+      "epoch": 46.292,
+      "grad_norm": 1.4009835720062256,
+      "learning_rate": 2e-05,
+      "loss": 0.05024692,
+      "step": 23146
+    },
+    {
+      "epoch": 46.294,
+      "grad_norm": 1.1895580291748047,
+      "learning_rate": 2e-05,
+      "loss": 0.04267966,
+      "step": 23147
+    },
+    {
+      "epoch": 46.296,
+      "grad_norm": 1.1182502508163452,
+      "learning_rate": 2e-05,
+      "loss": 0.05323965,
+      "step": 23148
+    },
+    {
+      "epoch": 46.298,
+      "grad_norm": 1.719334363937378,
+      "learning_rate": 2e-05,
+      "loss": 0.04970006,
+      "step": 23149
+    },
+    {
+      "epoch": 46.3,
+      "grad_norm": 2.0725491046905518,
+      "learning_rate": 2e-05,
+      "loss": 0.06950231,
+      "step": 23150
+    },
+    {
+      "epoch": 46.302,
+      "grad_norm": 3.3280606269836426,
+      "learning_rate": 2e-05,
+      "loss": 0.066084,
+      "step": 23151
+    },
+    {
+      "epoch": 46.304,
+      "grad_norm": 1.1809648275375366,
+      "learning_rate": 2e-05,
+      "loss": 0.04454755,
+      "step": 23152
+    },
+    {
+      "epoch": 46.306,
+      "grad_norm": 1.2297167778015137,
+      "learning_rate": 2e-05,
+      "loss": 0.06474146,
+      "step": 23153
+    },
+    {
+      "epoch": 46.308,
+      "grad_norm": 1.6546833515167236,
+      "learning_rate": 2e-05,
+      "loss": 0.04665916,
+      "step": 23154
+    },
+    {
+      "epoch": 46.31,
+      "grad_norm": 0.9898395538330078,
+      "learning_rate": 2e-05,
+      "loss": 0.03971004,
+      "step": 23155
+    },
+    {
+      "epoch": 46.312,
+      "grad_norm": 1.4820255041122437,
+      "learning_rate": 2e-05,
+      "loss": 0.04219128,
+      "step": 23156
+    },
+    {
+      "epoch": 46.314,
+      "grad_norm": 1.7158340215682983,
+      "learning_rate": 2e-05,
+      "loss": 0.0606416,
+      "step": 23157
+    },
+    {
+      "epoch": 46.316,
+      "grad_norm": 1.0219082832336426,
+      "learning_rate": 2e-05,
+      "loss": 0.04473948,
+      "step": 23158
+    },
+    {
+      "epoch": 46.318,
+      "grad_norm": 1.4255625009536743,
+      "learning_rate": 2e-05,
+      "loss": 0.04729429,
+      "step": 23159
+    },
+    {
+      "epoch": 46.32,
+      "grad_norm": 1.1542085409164429,
+      "learning_rate": 2e-05,
+      "loss": 0.04332086,
+      "step": 23160
+    },
+    {
+      "epoch": 46.322,
+      "grad_norm": 1.1608744859695435,
+      "learning_rate": 2e-05,
+      "loss": 0.05311103,
+      "step": 23161
+    },
+    {
+      "epoch": 46.324,
+      "grad_norm": 1.333052396774292,
+      "learning_rate": 2e-05,
+      "loss": 0.05798706,
+      "step": 23162
+    },
+    {
+      "epoch": 46.326,
+      "grad_norm": 1.1479754447937012,
+      "learning_rate": 2e-05,
+      "loss": 0.0463843,
+      "step": 23163
+    },
+    {
+      "epoch": 46.328,
+      "grad_norm": 1.8615238666534424,
+      "learning_rate": 2e-05,
+      "loss": 0.04511591,
+      "step": 23164
+    },
+    {
+      "epoch": 46.33,
+      "grad_norm": 1.2631683349609375,
+      "learning_rate": 2e-05,
+      "loss": 0.05475502,
+      "step": 23165
+    },
+    {
+      "epoch": 46.332,
+      "grad_norm": 1.4224528074264526,
+      "learning_rate": 2e-05,
+      "loss": 0.06586679,
+      "step": 23166
+    },
+    {
+      "epoch": 46.334,
+      "grad_norm": 1.1078559160232544,
+      "learning_rate": 2e-05,
+      "loss": 0.03966166,
+      "step": 23167
+    },
+    {
+      "epoch": 46.336,
+      "grad_norm": 0.9279854893684387,
+      "learning_rate": 2e-05,
+      "loss": 0.03631292,
+      "step": 23168
+    },
+    {
+      "epoch": 46.338,
+      "grad_norm": 2.5856709480285645,
+      "learning_rate": 2e-05,
+      "loss": 0.05210607,
+      "step": 23169
+    },
+    {
+      "epoch": 46.34,
+      "grad_norm": 1.2316886186599731,
+      "learning_rate": 2e-05,
+      "loss": 0.04556869,
+      "step": 23170
+    },
+    {
+      "epoch": 46.342,
+      "grad_norm": 1.1155706644058228,
+      "learning_rate": 2e-05,
+      "loss": 0.05066471,
+      "step": 23171
+    },
+    {
+      "epoch": 46.344,
+      "grad_norm": 2.491442918777466,
+      "learning_rate": 2e-05,
+      "loss": 0.05617189,
+      "step": 23172
+    },
+    {
+      "epoch": 46.346,
+      "grad_norm": 1.3126240968704224,
+      "learning_rate": 2e-05,
+      "loss": 0.04718146,
+      "step": 23173
+    },
+    {
+      "epoch": 46.348,
+      "grad_norm": 1.0606178045272827,
+      "learning_rate": 2e-05,
+      "loss": 0.04762572,
+      "step": 23174
+    },
+    {
+      "epoch": 46.35,
+      "grad_norm": 1.0557273626327515,
+      "learning_rate": 2e-05,
+      "loss": 0.03899936,
+      "step": 23175
+    },
+    {
+      "epoch": 46.352,
+      "grad_norm": 1.1816285848617554,
+      "learning_rate": 2e-05,
+      "loss": 0.04925738,
+      "step": 23176
+    },
+    {
+      "epoch": 46.354,
+      "grad_norm": 1.4543360471725464,
+      "learning_rate": 2e-05,
+      "loss": 0.04988147,
+      "step": 23177
+    },
+    {
+      "epoch": 46.356,
+      "grad_norm": 1.1029762029647827,
+      "learning_rate": 2e-05,
+      "loss": 0.04883325,
+      "step": 23178
+    },
+    {
+      "epoch": 46.358,
+      "grad_norm": 1.2085076570510864,
+      "learning_rate": 2e-05,
+      "loss": 0.0439341,
+      "step": 23179
+    },
+    {
+      "epoch": 46.36,
+      "grad_norm": 0.9505863785743713,
+      "learning_rate": 2e-05,
+      "loss": 0.03014771,
+      "step": 23180
+    },
+    {
+      "epoch": 46.362,
+      "grad_norm": 0.9543641805648804,
+      "learning_rate": 2e-05,
+      "loss": 0.03300781,
+      "step": 23181
+    },
+    {
+      "epoch": 46.364,
+      "grad_norm": 0.9900519251823425,
+      "learning_rate": 2e-05,
+      "loss": 0.04158117,
+      "step": 23182
+    },
+    {
+      "epoch": 46.366,
+      "grad_norm": 2.153928279876709,
+      "learning_rate": 2e-05,
+      "loss": 0.04417425,
+      "step": 23183
+    },
+    {
+      "epoch": 46.368,
+      "grad_norm": 2.409182071685791,
+      "learning_rate": 2e-05,
+      "loss": 0.06157666,
+      "step": 23184
+    },
+    {
+      "epoch": 46.37,
+      "grad_norm": 1.057466745376587,
+      "learning_rate": 2e-05,
+      "loss": 0.04592596,
+      "step": 23185
+    },
+    {
+      "epoch": 46.372,
+      "grad_norm": 1.1165255308151245,
+      "learning_rate": 2e-05,
+      "loss": 0.04613017,
+      "step": 23186
+    },
+    {
+      "epoch": 46.374,
+      "grad_norm": 0.9036556482315063,
+      "learning_rate": 2e-05,
+      "loss": 0.03239109,
+      "step": 23187
+    },
+    {
+      "epoch": 46.376,
+      "grad_norm": 1.1022392511367798,
+      "learning_rate": 2e-05,
+      "loss": 0.05009172,
+      "step": 23188
+    },
+    {
+      "epoch": 46.378,
+      "grad_norm": 1.0640677213668823,
+      "learning_rate": 2e-05,
+      "loss": 0.04262567,
+      "step": 23189
+    },
+    {
+      "epoch": 46.38,
+      "grad_norm": 1.11314857006073,
+      "learning_rate": 2e-05,
+      "loss": 0.04713321,
+      "step": 23190
+    },
+    {
+      "epoch": 46.382,
+      "grad_norm": 1.3541191816329956,
+      "learning_rate": 2e-05,
+      "loss": 0.04420114,
+      "step": 23191
+    },
+    {
+      "epoch": 46.384,
+      "grad_norm": 1.3517026901245117,
+      "learning_rate": 2e-05,
+      "loss": 0.03507008,
+      "step": 23192
+    },
+    {
+      "epoch": 46.386,
+      "grad_norm": 1.197635293006897,
+      "learning_rate": 2e-05,
+      "loss": 0.04363513,
+      "step": 23193
+    },
+    {
+      "epoch": 46.388,
+      "grad_norm": 1.3382245302200317,
+      "learning_rate": 2e-05,
+      "loss": 0.05595472,
+      "step": 23194
+    },
+    {
+      "epoch": 46.39,
+      "grad_norm": 1.074084758758545,
+      "learning_rate": 2e-05,
+      "loss": 0.05243614,
+      "step": 23195
+    },
+    {
+      "epoch": 46.392,
+      "grad_norm": 1.0921239852905273,
+      "learning_rate": 2e-05,
+      "loss": 0.03115225,
+      "step": 23196
+    },
+    {
+      "epoch": 46.394,
+      "grad_norm": 2.5413079261779785,
+      "learning_rate": 2e-05,
+      "loss": 0.04841086,
+      "step": 23197
+    },
+    {
+      "epoch": 46.396,
+      "grad_norm": 1.2901527881622314,
+      "learning_rate": 2e-05,
+      "loss": 0.05885056,
+      "step": 23198
+    },
+    {
+      "epoch": 46.398,
+      "grad_norm": 1.157203197479248,
+      "learning_rate": 2e-05,
+      "loss": 0.05642793,
+      "step": 23199
+    },
+    {
+      "epoch": 46.4,
+      "grad_norm": 1.1579469442367554,
+      "learning_rate": 2e-05,
+      "loss": 0.04676284,
+      "step": 23200
+    },
+    {
+      "epoch": 46.402,
+      "grad_norm": 1.3478295803070068,
+      "learning_rate": 2e-05,
+      "loss": 0.05287962,
+      "step": 23201
+    },
+    {
+      "epoch": 46.404,
+      "grad_norm": 0.9474977254867554,
+      "learning_rate": 2e-05,
+      "loss": 0.0353672,
+      "step": 23202
+    },
+    {
+      "epoch": 46.406,
+      "grad_norm": 1.401186466217041,
+      "learning_rate": 2e-05,
+      "loss": 0.05350681,
+      "step": 23203
+    },
+    {
+      "epoch": 46.408,
+      "grad_norm": 1.6736927032470703,
+      "learning_rate": 2e-05,
+      "loss": 0.05108627,
+      "step": 23204
+    },
+    {
+      "epoch": 46.41,
+      "grad_norm": 1.5280506610870361,
+      "learning_rate": 2e-05,
+      "loss": 0.06310539,
+      "step": 23205
+    },
+    {
+      "epoch": 46.412,
+      "grad_norm": 1.2769758701324463,
+      "learning_rate": 2e-05,
+      "loss": 0.04531645,
+      "step": 23206
+    },
+    {
+      "epoch": 46.414,
+      "grad_norm": 1.0869319438934326,
+      "learning_rate": 2e-05,
+      "loss": 0.04933855,
+      "step": 23207
+    },
+    {
+      "epoch": 46.416,
+      "grad_norm": 1.8246020078659058,
+      "learning_rate": 2e-05,
+      "loss": 0.04469419,
+      "step": 23208
+    },
+    {
+      "epoch": 46.418,
+      "grad_norm": 1.1753547191619873,
+      "learning_rate": 2e-05,
+      "loss": 0.03421785,
+      "step": 23209
+    },
+    {
+      "epoch": 46.42,
+      "grad_norm": 1.1583861112594604,
+      "learning_rate": 2e-05,
+      "loss": 0.05171474,
+      "step": 23210
+    },
+    {
+      "epoch": 46.422,
+      "grad_norm": 1.1681585311889648,
+      "learning_rate": 2e-05,
+      "loss": 0.04288517,
+      "step": 23211
+    },
+    {
+      "epoch": 46.424,
+      "grad_norm": 1.8628696203231812,
+      "learning_rate": 2e-05,
+      "loss": 0.06078832,
+      "step": 23212
+    },
+    {
+      "epoch": 46.426,
+      "grad_norm": 2.069993495941162,
+      "learning_rate": 2e-05,
+      "loss": 0.06477488,
+      "step": 23213
+    },
+    {
+      "epoch": 46.428,
+      "grad_norm": 3.0566177368164062,
+      "learning_rate": 2e-05,
+      "loss": 0.05633872,
+      "step": 23214
+    },
+    {
+      "epoch": 46.43,
+      "grad_norm": 1.5551815032958984,
+      "learning_rate": 2e-05,
+      "loss": 0.06199846,
+      "step": 23215
+    },
+    {
+      "epoch": 46.432,
+      "grad_norm": 1.2828153371810913,
+      "learning_rate": 2e-05,
+      "loss": 0.04329269,
+      "step": 23216
+    },
+    {
+      "epoch": 46.434,
+      "grad_norm": 1.1735610961914062,
+      "learning_rate": 2e-05,
+      "loss": 0.05622312,
+      "step": 23217
+    },
+    {
+      "epoch": 46.436,
+      "grad_norm": 2.0017809867858887,
+      "learning_rate": 2e-05,
+      "loss": 0.03825869,
+      "step": 23218
+    },
+    {
+      "epoch": 46.438,
+      "grad_norm": 1.1812978982925415,
+      "learning_rate": 2e-05,
+      "loss": 0.05794305,
+      "step": 23219
+    },
+    {
+      "epoch": 46.44,
+      "grad_norm": 1.3604928255081177,
+      "learning_rate": 2e-05,
+      "loss": 0.05252583,
+      "step": 23220
+    },
+    {
+      "epoch": 46.442,
+      "grad_norm": 1.1032557487487793,
+      "learning_rate": 2e-05,
+      "loss": 0.03681751,
+      "step": 23221
+    },
+    {
+      "epoch": 46.444,
+      "grad_norm": 1.597011685371399,
+      "learning_rate": 2e-05,
+      "loss": 0.03853673,
+      "step": 23222
+    },
+    {
+      "epoch": 46.446,
+      "grad_norm": 1.193235158920288,
+      "learning_rate": 2e-05,
+      "loss": 0.05152307,
+      "step": 23223
+    },
+    {
+      "epoch": 46.448,
+      "grad_norm": 1.071548581123352,
+      "learning_rate": 2e-05,
+      "loss": 0.04468784,
+      "step": 23224
+    },
+    {
+      "epoch": 46.45,
+      "grad_norm": 1.2598882913589478,
+      "learning_rate": 2e-05,
+      "loss": 0.05761843,
+      "step": 23225
+    },
+    {
+      "epoch": 46.452,
+      "grad_norm": 1.1971979141235352,
+      "learning_rate": 2e-05,
+      "loss": 0.05141561,
+      "step": 23226
+    },
+    {
+      "epoch": 46.454,
+      "grad_norm": 2.1761207580566406,
+      "learning_rate": 2e-05,
+      "loss": 0.04906508,
+      "step": 23227
+    },
+    {
+      "epoch": 46.456,
+      "grad_norm": 1.4585120677947998,
+      "learning_rate": 2e-05,
+      "loss": 0.0423772,
+      "step": 23228
+    },
+    {
+      "epoch": 46.458,
+      "grad_norm": 1.1392666101455688,
+      "learning_rate": 2e-05,
+      "loss": 0.04400862,
+      "step": 23229
+    },
+    {
+      "epoch": 46.46,
+      "grad_norm": 1.4353058338165283,
+      "learning_rate": 2e-05,
+      "loss": 0.0567156,
+      "step": 23230
+    },
+    {
+      "epoch": 46.462,
+      "grad_norm": 1.12968909740448,
+      "learning_rate": 2e-05,
+      "loss": 0.03764338,
+      "step": 23231
+    },
+    {
+      "epoch": 46.464,
+      "grad_norm": 1.2651554346084595,
+      "learning_rate": 2e-05,
+      "loss": 0.06105966,
+      "step": 23232
+    },
+    {
+      "epoch": 46.466,
+      "grad_norm": 1.1130878925323486,
+      "learning_rate": 2e-05,
+      "loss": 0.03365841,
+      "step": 23233
+    },
+    {
+      "epoch": 46.468,
+      "grad_norm": 1.5134097337722778,
+      "learning_rate": 2e-05,
+      "loss": 0.04425268,
+      "step": 23234
+    },
+    {
+      "epoch": 46.47,
+      "grad_norm": 1.2247387170791626,
+      "learning_rate": 2e-05,
+      "loss": 0.04427869,
+      "step": 23235
+    },
+    {
+      "epoch": 46.472,
+      "grad_norm": 1.1638450622558594,
+      "learning_rate": 2e-05,
+      "loss": 0.04198921,
+      "step": 23236
+    },
+    {
+      "epoch": 46.474,
+      "grad_norm": 1.113433599472046,
+      "learning_rate": 2e-05,
+      "loss": 0.03829671,
+      "step": 23237
+    },
+    {
+      "epoch": 46.476,
+      "grad_norm": 1.3648087978363037,
+      "learning_rate": 2e-05,
+      "loss": 0.03742781,
+      "step": 23238
+    },
+    {
+      "epoch": 46.478,
+      "grad_norm": 1.5586036443710327,
+      "learning_rate": 2e-05,
+      "loss": 0.04454248,
+      "step": 23239
+    },
+    {
+      "epoch": 46.48,
+      "grad_norm": 1.5988510847091675,
+      "learning_rate": 2e-05,
+      "loss": 0.04805886,
+      "step": 23240
+    },
+    {
+      "epoch": 46.482,
+      "grad_norm": 1.0325082540512085,
+      "learning_rate": 2e-05,
+      "loss": 0.03924419,
+      "step": 23241
+    },
+    {
+      "epoch": 46.484,
+      "grad_norm": 1.351863980293274,
+      "learning_rate": 2e-05,
+      "loss": 0.05655452,
+      "step": 23242
+    },
+    {
+      "epoch": 46.486,
+      "grad_norm": 1.4276981353759766,
+      "learning_rate": 2e-05,
+      "loss": 0.05009124,
+      "step": 23243
+    },
+    {
+      "epoch": 46.488,
+      "grad_norm": 1.474223017692566,
+      "learning_rate": 2e-05,
+      "loss": 0.04846952,
+      "step": 23244
+    },
+    {
+      "epoch": 46.49,
+      "grad_norm": 0.9883701205253601,
+      "learning_rate": 2e-05,
+      "loss": 0.03598514,
+      "step": 23245
+    },
+    {
+      "epoch": 46.492,
+      "grad_norm": 1.3215599060058594,
+      "learning_rate": 2e-05,
+      "loss": 0.05669083,
+      "step": 23246
+    },
+    {
+      "epoch": 46.494,
+      "grad_norm": 1.0561680793762207,
+      "learning_rate": 2e-05,
+      "loss": 0.04358659,
+      "step": 23247
+    },
+    {
+      "epoch": 46.496,
+      "grad_norm": 1.151599407196045,
+      "learning_rate": 2e-05,
+      "loss": 0.05015893,
+      "step": 23248
+    },
+    {
+      "epoch": 46.498,
+      "grad_norm": 1.4524661302566528,
+      "learning_rate": 2e-05,
+      "loss": 0.04662398,
+      "step": 23249
+    },
+    {
+      "epoch": 46.5,
+      "grad_norm": 1.5353271961212158,
+      "learning_rate": 2e-05,
+      "loss": 0.03597783,
+      "step": 23250
+    },
+    {
+      "epoch": 46.502,
+      "grad_norm": 2.727353811264038,
+      "learning_rate": 2e-05,
+      "loss": 0.03727568,
+      "step": 23251
+    },
+    {
+      "epoch": 46.504,
+      "grad_norm": 1.1945043802261353,
+      "learning_rate": 2e-05,
+      "loss": 0.03968314,
+      "step": 23252
+    },
+    {
+      "epoch": 46.506,
+      "grad_norm": 1.0994758605957031,
+      "learning_rate": 2e-05,
+      "loss": 0.05075628,
+      "step": 23253
+    },
+    {
+      "epoch": 46.508,
+      "grad_norm": 1.2463784217834473,
+      "learning_rate": 2e-05,
+      "loss": 0.05148967,
+      "step": 23254
+    },
+    {
+      "epoch": 46.51,
+      "grad_norm": 1.2693138122558594,
+      "learning_rate": 2e-05,
+      "loss": 0.04996495,
+      "step": 23255
+    },
+    {
+      "epoch": 46.512,
+      "grad_norm": 1.0111137628555298,
+      "learning_rate": 2e-05,
+      "loss": 0.04151868,
+      "step": 23256
+    },
+    {
+      "epoch": 46.514,
+      "grad_norm": 1.1659035682678223,
+      "learning_rate": 2e-05,
+      "loss": 0.03361614,
+      "step": 23257
+    },
+    {
+      "epoch": 46.516,
+      "grad_norm": 1.3636959791183472,
+      "learning_rate": 2e-05,
+      "loss": 0.054345,
+      "step": 23258
+    },
+    {
+      "epoch": 46.518,
+      "grad_norm": 1.228248953819275,
+      "learning_rate": 2e-05,
+      "loss": 0.06052267,
+      "step": 23259
+    },
+    {
+      "epoch": 46.52,
+      "grad_norm": 1.2913553714752197,
+      "learning_rate": 2e-05,
+      "loss": 0.05008456,
+      "step": 23260
+    },
+    {
+      "epoch": 46.522,
+      "grad_norm": 1.2020117044448853,
+      "learning_rate": 2e-05,
+      "loss": 0.05029326,
+      "step": 23261
+    },
+    {
+      "epoch": 46.524,
+      "grad_norm": 1.0171692371368408,
+      "learning_rate": 2e-05,
+      "loss": 0.03987342,
+      "step": 23262
+    },
+    {
+      "epoch": 46.526,
+      "grad_norm": 1.2745869159698486,
+      "learning_rate": 2e-05,
+      "loss": 0.06460331,
+      "step": 23263
+    },
+    {
+      "epoch": 46.528,
+      "grad_norm": 1.0925174951553345,
+      "learning_rate": 2e-05,
+      "loss": 0.05129455,
+      "step": 23264
+    },
+    {
+      "epoch": 46.53,
+      "grad_norm": 1.1556943655014038,
+      "learning_rate": 2e-05,
+      "loss": 0.0476981,
+      "step": 23265
+    },
+    {
+      "epoch": 46.532,
+      "grad_norm": 1.4467393159866333,
+      "learning_rate": 2e-05,
+      "loss": 0.06611196,
+      "step": 23266
+    },
+    {
+      "epoch": 46.534,
+      "grad_norm": 1.8696622848510742,
+      "learning_rate": 2e-05,
+      "loss": 0.04994733,
+      "step": 23267
+    },
+    {
+      "epoch": 46.536,
+      "grad_norm": 1.0204404592514038,
+      "learning_rate": 2e-05,
+      "loss": 0.04660401,
+      "step": 23268
+    },
+    {
+      "epoch": 46.538,
+      "grad_norm": 1.2084547281265259,
+      "learning_rate": 2e-05,
+      "loss": 0.05732334,
+      "step": 23269
+    },
+    {
+      "epoch": 46.54,
+      "grad_norm": 1.333622932434082,
+      "learning_rate": 2e-05,
+      "loss": 0.05636217,
+      "step": 23270
+    },
+    {
+      "epoch": 46.542,
+      "grad_norm": 1.1122244596481323,
+      "learning_rate": 2e-05,
+      "loss": 0.04155293,
+      "step": 23271
+    },
+    {
+      "epoch": 46.544,
+      "grad_norm": 1.6370881795883179,
+      "learning_rate": 2e-05,
+      "loss": 0.05667079,
+      "step": 23272
+    },
+    {
+      "epoch": 46.546,
+      "grad_norm": 1.505200743675232,
+      "learning_rate": 2e-05,
+      "loss": 0.05249736,
+      "step": 23273
+    },
+    {
+      "epoch": 46.548,
+      "grad_norm": 1.155645728111267,
+      "learning_rate": 2e-05,
+      "loss": 0.05694194,
+      "step": 23274
+    },
+    {
+      "epoch": 46.55,
+      "grad_norm": 1.1647768020629883,
+      "learning_rate": 2e-05,
+      "loss": 0.05435292,
+      "step": 23275
+    },
+    {
+      "epoch": 46.552,
+      "grad_norm": 1.0532338619232178,
+      "learning_rate": 2e-05,
+      "loss": 0.04438793,
+      "step": 23276
+    },
+    {
+      "epoch": 46.554,
+      "grad_norm": 1.6303856372833252,
+      "learning_rate": 2e-05,
+      "loss": 0.05212019,
+      "step": 23277
+    },
+    {
+      "epoch": 46.556,
+      "grad_norm": 1.217422604560852,
+      "learning_rate": 2e-05,
+      "loss": 0.03759392,
+      "step": 23278
+    },
+    {
+      "epoch": 46.558,
+      "grad_norm": 1.3669236898422241,
+      "learning_rate": 2e-05,
+      "loss": 0.05771047,
+      "step": 23279
+    },
+    {
+      "epoch": 46.56,
+      "grad_norm": 2.1932220458984375,
+      "learning_rate": 2e-05,
+      "loss": 0.04165822,
+      "step": 23280
+    },
+    {
+      "epoch": 46.562,
+      "grad_norm": 1.2292122840881348,
+      "learning_rate": 2e-05,
+      "loss": 0.04785195,
+      "step": 23281
+    },
+    {
+      "epoch": 46.564,
+      "grad_norm": 1.196970820426941,
+      "learning_rate": 2e-05,
+      "loss": 0.03689167,
+      "step": 23282
+    },
+    {
+      "epoch": 46.566,
+      "grad_norm": 1.020137071609497,
+      "learning_rate": 2e-05,
+      "loss": 0.03734909,
+      "step": 23283
+    },
+    {
+      "epoch": 46.568,
+      "grad_norm": 2.541544198989868,
+      "learning_rate": 2e-05,
+      "loss": 0.06038415,
+      "step": 23284
+    },
+    {
+      "epoch": 46.57,
+      "grad_norm": 1.083556890487671,
+      "learning_rate": 2e-05,
+      "loss": 0.03873578,
+      "step": 23285
+    },
+    {
+      "epoch": 46.572,
+      "grad_norm": 1.4096356630325317,
+      "learning_rate": 2e-05,
+      "loss": 0.05036553,
+      "step": 23286
+    },
+    {
+      "epoch": 46.574,
+      "grad_norm": 0.9817960858345032,
+      "learning_rate": 2e-05,
+      "loss": 0.03779167,
+      "step": 23287
+    },
+    {
+      "epoch": 46.576,
+      "grad_norm": 1.2446906566619873,
+      "learning_rate": 2e-05,
+      "loss": 0.05619381,
+      "step": 23288
+    },
+    {
+      "epoch": 46.578,
+      "grad_norm": 1.0761586427688599,
+      "learning_rate": 2e-05,
+      "loss": 0.04057213,
+      "step": 23289
+    },
+    {
+      "epoch": 46.58,
+      "grad_norm": 1.518162727355957,
+      "learning_rate": 2e-05,
+      "loss": 0.03429618,
+      "step": 23290
+    },
+    {
+      "epoch": 46.582,
+      "grad_norm": 1.1717662811279297,
+      "learning_rate": 2e-05,
+      "loss": 0.0466333,
+      "step": 23291
+    },
+    {
+      "epoch": 46.584,
+      "grad_norm": 3.0718834400177,
+      "learning_rate": 2e-05,
+      "loss": 0.04770675,
+      "step": 23292
+    },
+    {
+      "epoch": 46.586,
+      "grad_norm": 1.1569914817810059,
+      "learning_rate": 2e-05,
+      "loss": 0.05019266,
+      "step": 23293
+    },
+    {
+      "epoch": 46.588,
+      "grad_norm": 1.0586425065994263,
+      "learning_rate": 2e-05,
+      "loss": 0.03773466,
+      "step": 23294
+    },
+    {
+      "epoch": 46.59,
+      "grad_norm": 1.4067730903625488,
+      "learning_rate": 2e-05,
+      "loss": 0.06203558,
+      "step": 23295
+    },
+    {
+      "epoch": 46.592,
+      "grad_norm": 0.9604422450065613,
+      "learning_rate": 2e-05,
+      "loss": 0.03419396,
+      "step": 23296
+    },
+    {
+      "epoch": 46.594,
+      "grad_norm": 1.4648443460464478,
+      "learning_rate": 2e-05,
+      "loss": 0.06012858,
+      "step": 23297
+    },
+    {
+      "epoch": 46.596,
+      "grad_norm": 1.02535080909729,
+      "learning_rate": 2e-05,
+      "loss": 0.03938122,
+      "step": 23298
+    },
+    {
+      "epoch": 46.598,
+      "grad_norm": 1.0963574647903442,
+      "learning_rate": 2e-05,
+      "loss": 0.0514155,
+      "step": 23299
+    },
+    {
+      "epoch": 46.6,
+      "grad_norm": 1.111864447593689,
+      "learning_rate": 2e-05,
+      "loss": 0.03777619,
+      "step": 23300
+    },
+    {
+      "epoch": 46.602,
+      "grad_norm": 1.2542434930801392,
+      "learning_rate": 2e-05,
+      "loss": 0.03768213,
+      "step": 23301
+    },
+    {
+      "epoch": 46.604,
+      "grad_norm": 0.9642258286476135,
+      "learning_rate": 2e-05,
+      "loss": 0.040147,
+      "step": 23302
+    },
+    {
+      "epoch": 46.606,
+      "grad_norm": 1.2112300395965576,
+      "learning_rate": 2e-05,
+      "loss": 0.06443147,
+      "step": 23303
+    },
+    {
+      "epoch": 46.608,
+      "grad_norm": 0.9941690564155579,
+      "learning_rate": 2e-05,
+      "loss": 0.04848402,
+      "step": 23304
+    },
+    {
+      "epoch": 46.61,
+      "grad_norm": 1.9957858324050903,
+      "learning_rate": 2e-05,
+      "loss": 0.04321967,
+      "step": 23305
+    },
+    {
+      "epoch": 46.612,
+      "grad_norm": 0.9449677467346191,
+      "learning_rate": 2e-05,
+      "loss": 0.02798765,
+      "step": 23306
+    },
+    {
+      "epoch": 46.614,
+      "grad_norm": 1.068254828453064,
+      "learning_rate": 2e-05,
+      "loss": 0.04512742,
+      "step": 23307
+    },
+    {
+      "epoch": 46.616,
+      "grad_norm": 1.142982006072998,
+      "learning_rate": 2e-05,
+      "loss": 0.04882696,
+      "step": 23308
+    },
+    {
+      "epoch": 46.618,
+      "grad_norm": 1.2351460456848145,
+      "learning_rate": 2e-05,
+      "loss": 0.05512179,
+      "step": 23309
+    },
+    {
+      "epoch": 46.62,
+      "grad_norm": 2.279547691345215,
+      "learning_rate": 2e-05,
+      "loss": 0.03314951,
+      "step": 23310
+    },
+    {
+      "epoch": 46.622,
+      "grad_norm": 1.0746487379074097,
+      "learning_rate": 2e-05,
+      "loss": 0.04710182,
+      "step": 23311
+    },
+    {
+      "epoch": 46.624,
+      "grad_norm": 1.222518801689148,
+      "learning_rate": 2e-05,
+      "loss": 0.06075151,
+      "step": 23312
+    },
+    {
+      "epoch": 46.626,
+      "grad_norm": 0.9651801586151123,
+      "learning_rate": 2e-05,
+      "loss": 0.04442292,
+      "step": 23313
+    },
+    {
+      "epoch": 46.628,
+      "grad_norm": 0.9503448605537415,
+      "learning_rate": 2e-05,
+      "loss": 0.03715204,
+      "step": 23314
+    },
+    {
+      "epoch": 46.63,
+      "grad_norm": 1.1955111026763916,
+      "learning_rate": 2e-05,
+      "loss": 0.0486602,
+      "step": 23315
+    },
+    {
+      "epoch": 46.632,
+      "grad_norm": 1.020430088043213,
+      "learning_rate": 2e-05,
+      "loss": 0.0444079,
+      "step": 23316
+    },
+    {
+      "epoch": 46.634,
+      "grad_norm": 1.1504944562911987,
+      "learning_rate": 2e-05,
+      "loss": 0.0579015,
+      "step": 23317
+    },
+    {
+      "epoch": 46.636,
+      "grad_norm": 1.0248736143112183,
+      "learning_rate": 2e-05,
+      "loss": 0.03767063,
+      "step": 23318
+    },
+    {
+      "epoch": 46.638,
+      "grad_norm": 1.0812240839004517,
+      "learning_rate": 2e-05,
+      "loss": 0.03341661,
+      "step": 23319
+    },
+    {
+      "epoch": 46.64,
+      "grad_norm": 1.3688080310821533,
+      "learning_rate": 2e-05,
+      "loss": 0.04959516,
+      "step": 23320
+    },
+    {
+      "epoch": 46.642,
+      "grad_norm": 1.0399272441864014,
+      "learning_rate": 2e-05,
+      "loss": 0.04377398,
+      "step": 23321
+    },
+    {
+      "epoch": 46.644,
+      "grad_norm": 1.2335994243621826,
+      "learning_rate": 2e-05,
+      "loss": 0.06242413,
+      "step": 23322
+    },
+    {
+      "epoch": 46.646,
+      "grad_norm": 1.2382959127426147,
+      "learning_rate": 2e-05,
+      "loss": 0.04701665,
+      "step": 23323
+    },
+    {
+      "epoch": 46.648,
+      "grad_norm": 1.318410873413086,
+      "learning_rate": 2e-05,
+      "loss": 0.03696349,
+      "step": 23324
+    },
+    {
+      "epoch": 46.65,
+      "grad_norm": 2.3140316009521484,
+      "learning_rate": 2e-05,
+      "loss": 0.05928268,
+      "step": 23325
+    },
+    {
+      "epoch": 46.652,
+      "grad_norm": 1.1803500652313232,
+      "learning_rate": 2e-05,
+      "loss": 0.03985074,
+      "step": 23326
+    },
+    {
+      "epoch": 46.654,
+      "grad_norm": 1.5823149681091309,
+      "learning_rate": 2e-05,
+      "loss": 0.06254095,
+      "step": 23327
+    },
+    {
+      "epoch": 46.656,
+      "grad_norm": 1.4812148809432983,
+      "learning_rate": 2e-05,
+      "loss": 0.05604445,
+      "step": 23328
+    },
+    {
+      "epoch": 46.658,
+      "grad_norm": 2.1350300312042236,
+      "learning_rate": 2e-05,
+      "loss": 0.05875448,
+      "step": 23329
+    },
+    {
+      "epoch": 46.66,
+      "grad_norm": 1.400207757949829,
+      "learning_rate": 2e-05,
+      "loss": 0.07081041,
+      "step": 23330
+    },
+    {
+      "epoch": 46.662,
+      "grad_norm": 1.0994622707366943,
+      "learning_rate": 2e-05,
+      "loss": 0.04271101,
+      "step": 23331
+    },
+    {
+      "epoch": 46.664,
+      "grad_norm": 1.0609034299850464,
+      "learning_rate": 2e-05,
+      "loss": 0.04732626,
+      "step": 23332
+    },
+    {
+      "epoch": 46.666,
+      "grad_norm": 1.0104788541793823,
+      "learning_rate": 2e-05,
+      "loss": 0.03551992,
+      "step": 23333
+    },
+    {
+      "epoch": 46.668,
+      "grad_norm": 1.1824249029159546,
+      "learning_rate": 2e-05,
+      "loss": 0.05614135,
+      "step": 23334
+    },
+    {
+      "epoch": 46.67,
+      "grad_norm": 2.1531567573547363,
+      "learning_rate": 2e-05,
+      "loss": 0.04873116,
+      "step": 23335
+    },
+    {
+      "epoch": 46.672,
+      "grad_norm": 1.3467340469360352,
+      "learning_rate": 2e-05,
+      "loss": 0.03813969,
+      "step": 23336
+    },
+    {
+      "epoch": 46.674,
+      "grad_norm": 1.037055253982544,
+      "learning_rate": 2e-05,
+      "loss": 0.04405836,
+      "step": 23337
+    },
+    {
+      "epoch": 46.676,
+      "grad_norm": 1.868750810623169,
+      "learning_rate": 2e-05,
+      "loss": 0.06937762,
+      "step": 23338
+    },
+    {
+      "epoch": 46.678,
+      "grad_norm": 1.8962420225143433,
+      "learning_rate": 2e-05,
+      "loss": 0.04944228,
+      "step": 23339
+    },
+    {
+      "epoch": 46.68,
+      "grad_norm": 1.170283317565918,
+      "learning_rate": 2e-05,
+      "loss": 0.04979547,
+      "step": 23340
+    },
+    {
+      "epoch": 46.682,
+      "grad_norm": 1.2021749019622803,
+      "learning_rate": 2e-05,
+      "loss": 0.05005801,
+      "step": 23341
+    },
+    {
+      "epoch": 46.684,
+      "grad_norm": 1.635096549987793,
+      "learning_rate": 2e-05,
+      "loss": 0.05956845,
+      "step": 23342
+    },
+    {
+      "epoch": 46.686,
+      "grad_norm": 2.5779504776000977,
+      "learning_rate": 2e-05,
+      "loss": 0.05944379,
+      "step": 23343
+    },
+    {
+      "epoch": 46.688,
+      "grad_norm": 1.169403076171875,
+      "learning_rate": 2e-05,
+      "loss": 0.04585174,
+      "step": 23344
+    },
+    {
+      "epoch": 46.69,
+      "grad_norm": 1.3411033153533936,
+      "learning_rate": 2e-05,
+      "loss": 0.06493143,
+      "step": 23345
+    },
+    {
+      "epoch": 46.692,
+      "grad_norm": 1.2918158769607544,
+      "learning_rate": 2e-05,
+      "loss": 0.04256565,
+      "step": 23346
+    },
+    {
+      "epoch": 46.694,
+      "grad_norm": 0.985222578048706,
+      "learning_rate": 2e-05,
+      "loss": 0.0372174,
+      "step": 23347
+    },
+    {
+      "epoch": 46.696,
+      "grad_norm": 1.0924230813980103,
+      "learning_rate": 2e-05,
+      "loss": 0.03921843,
+      "step": 23348
+    },
+    {
+      "epoch": 46.698,
+      "grad_norm": 1.3069814443588257,
+      "learning_rate": 2e-05,
+      "loss": 0.04222802,
+      "step": 23349
+    },
+    {
+      "epoch": 46.7,
+      "grad_norm": 1.1846057176589966,
+      "learning_rate": 2e-05,
+      "loss": 0.04912648,
+      "step": 23350
+    },
+    {
+      "epoch": 46.702,
+      "grad_norm": 1.3495811223983765,
+      "learning_rate": 2e-05,
+      "loss": 0.04603204,
+      "step": 23351
+    },
+    {
+      "epoch": 46.704,
+      "grad_norm": 1.0049281120300293,
+      "learning_rate": 2e-05,
+      "loss": 0.04690091,
+      "step": 23352
+    },
+    {
+      "epoch": 46.706,
+      "grad_norm": 1.2509807348251343,
+      "learning_rate": 2e-05,
+      "loss": 0.05150412,
+      "step": 23353
+    },
+    {
+      "epoch": 46.708,
+      "grad_norm": 1.5264276266098022,
+      "learning_rate": 2e-05,
+      "loss": 0.05389606,
+      "step": 23354
+    },
+    {
+      "epoch": 46.71,
+      "grad_norm": 1.7153115272521973,
+      "learning_rate": 2e-05,
+      "loss": 0.06340338,
+      "step": 23355
+    },
+    {
+      "epoch": 46.712,
+      "grad_norm": 1.010880708694458,
+      "learning_rate": 2e-05,
+      "loss": 0.0329822,
+      "step": 23356
+    },
+    {
+      "epoch": 46.714,
+      "grad_norm": 1.157622218132019,
+      "learning_rate": 2e-05,
+      "loss": 0.05921149,
+      "step": 23357
+    },
+    {
+      "epoch": 46.716,
+      "grad_norm": 0.9842050075531006,
+      "learning_rate": 2e-05,
+      "loss": 0.03667116,
+      "step": 23358
+    },
+    {
+      "epoch": 46.718,
+      "grad_norm": 1.1801451444625854,
+      "learning_rate": 2e-05,
+      "loss": 0.04451838,
+      "step": 23359
+    },
+    {
+      "epoch": 46.72,
+      "grad_norm": 1.2267253398895264,
+      "learning_rate": 2e-05,
+      "loss": 0.05214031,
+      "step": 23360
+    },
+    {
+      "epoch": 46.722,
+      "grad_norm": 1.019111156463623,
+      "learning_rate": 2e-05,
+      "loss": 0.04116342,
+      "step": 23361
+    },
+    {
+      "epoch": 46.724,
+      "grad_norm": 1.3958607912063599,
+      "learning_rate": 2e-05,
+      "loss": 0.05170007,
+      "step": 23362
+    },
+    {
+      "epoch": 46.726,
+      "grad_norm": 1.5343711376190186,
+      "learning_rate": 2e-05,
+      "loss": 0.05161713,
+      "step": 23363
+    },
+    {
+      "epoch": 46.728,
+      "grad_norm": 0.9997346997261047,
+      "learning_rate": 2e-05,
+      "loss": 0.0342833,
+      "step": 23364
+    },
+    {
+      "epoch": 46.73,
+      "grad_norm": 1.421368956565857,
+      "learning_rate": 2e-05,
+      "loss": 0.04781113,
+      "step": 23365
+    },
+    {
+      "epoch": 46.732,
+      "grad_norm": 1.6076453924179077,
+      "learning_rate": 2e-05,
+      "loss": 0.04223795,
+      "step": 23366
+    },
+    {
+      "epoch": 46.734,
+      "grad_norm": 1.1187045574188232,
+      "learning_rate": 2e-05,
+      "loss": 0.03693566,
+      "step": 23367
+    },
+    {
+      "epoch": 46.736,
+      "grad_norm": 1.6310001611709595,
+      "learning_rate": 2e-05,
+      "loss": 0.04587207,
+      "step": 23368
+    },
+    {
+      "epoch": 46.738,
+      "grad_norm": 1.0003893375396729,
+      "learning_rate": 2e-05,
+      "loss": 0.02953731,
+      "step": 23369
+    },
+    {
+      "epoch": 46.74,
+      "grad_norm": 2.397538900375366,
+      "learning_rate": 2e-05,
+      "loss": 0.04967777,
+      "step": 23370
+    },
+    {
+      "epoch": 46.742,
+      "grad_norm": 2.9524292945861816,
+      "learning_rate": 2e-05,
+      "loss": 0.06176184,
+      "step": 23371
+    },
+    {
+      "epoch": 46.744,
+      "grad_norm": 1.3411389589309692,
+      "learning_rate": 2e-05,
+      "loss": 0.05731054,
+      "step": 23372
+    },
+    {
+      "epoch": 46.746,
+      "grad_norm": 0.9929517507553101,
+      "learning_rate": 2e-05,
+      "loss": 0.04180113,
+      "step": 23373
+    },
+    {
+      "epoch": 46.748,
+      "grad_norm": 1.270747184753418,
+      "learning_rate": 2e-05,
+      "loss": 0.05174856,
+      "step": 23374
+    },
+    {
+      "epoch": 46.75,
+      "grad_norm": 1.058241605758667,
+      "learning_rate": 2e-05,
+      "loss": 0.04240339,
+      "step": 23375
+    },
+    {
+      "epoch": 46.752,
+      "grad_norm": 1.1424481868743896,
+      "learning_rate": 2e-05,
+      "loss": 0.04190686,
+      "step": 23376
+    },
+    {
+      "epoch": 46.754,
+      "grad_norm": 1.0503220558166504,
+      "learning_rate": 2e-05,
+      "loss": 0.05099996,
+      "step": 23377
+    },
+    {
+      "epoch": 46.756,
+      "grad_norm": 1.255510687828064,
+      "learning_rate": 2e-05,
+      "loss": 0.03783688,
+      "step": 23378
+    },
+    {
+      "epoch": 46.758,
+      "grad_norm": 1.1512470245361328,
+      "learning_rate": 2e-05,
+      "loss": 0.04825658,
+      "step": 23379
+    },
+    {
+      "epoch": 46.76,
+      "grad_norm": 1.2274504899978638,
+      "learning_rate": 2e-05,
+      "loss": 0.05738919,
+      "step": 23380
+    },
+    {
+      "epoch": 46.762,
+      "grad_norm": 2.000230073928833,
+      "learning_rate": 2e-05,
+      "loss": 0.05466451,
+      "step": 23381
+    },
+    {
+      "epoch": 46.764,
+      "grad_norm": 1.1941397190093994,
+      "learning_rate": 2e-05,
+      "loss": 0.0490576,
+      "step": 23382
+    },
+    {
+      "epoch": 46.766,
+      "grad_norm": 1.1872297525405884,
+      "learning_rate": 2e-05,
+      "loss": 0.04906765,
+      "step": 23383
+    },
+    {
+      "epoch": 46.768,
+      "grad_norm": 1.068359136581421,
+      "learning_rate": 2e-05,
+      "loss": 0.04766599,
+      "step": 23384
+    },
+    {
+      "epoch": 46.77,
+      "grad_norm": 1.1551222801208496,
+      "learning_rate": 2e-05,
+      "loss": 0.05015551,
+      "step": 23385
+    },
+    {
+      "epoch": 46.772,
+      "grad_norm": 1.1826947927474976,
+      "learning_rate": 2e-05,
+      "loss": 0.04717891,
+      "step": 23386
+    },
+    {
+      "epoch": 46.774,
+      "grad_norm": 1.3930027484893799,
+      "learning_rate": 2e-05,
+      "loss": 0.04331183,
+      "step": 23387
+    },
+    {
+      "epoch": 46.776,
+      "grad_norm": 1.001609206199646,
+      "learning_rate": 2e-05,
+      "loss": 0.03368718,
+      "step": 23388
+    },
+    {
+      "epoch": 46.778,
+      "grad_norm": 1.371207356452942,
+      "learning_rate": 2e-05,
+      "loss": 0.05175086,
+      "step": 23389
+    },
+    {
+      "epoch": 46.78,
+      "grad_norm": 1.378040075302124,
+      "learning_rate": 2e-05,
+      "loss": 0.04431616,
+      "step": 23390
+    },
+    {
+      "epoch": 46.782,
+      "grad_norm": 1.5609763860702515,
+      "learning_rate": 2e-05,
+      "loss": 0.04517556,
+      "step": 23391
+    },
+    {
+      "epoch": 46.784,
+      "grad_norm": 1.9839285612106323,
+      "learning_rate": 2e-05,
+      "loss": 0.05456992,
+      "step": 23392
+    },
+    {
+      "epoch": 46.786,
+      "grad_norm": 0.9844745993614197,
+      "learning_rate": 2e-05,
+      "loss": 0.0431798,
+      "step": 23393
+    },
+    {
+      "epoch": 46.788,
+      "grad_norm": 1.6684527397155762,
+      "learning_rate": 2e-05,
+      "loss": 0.03772101,
+      "step": 23394
+    },
+    {
+      "epoch": 46.79,
+      "grad_norm": 1.1648509502410889,
+      "learning_rate": 2e-05,
+      "loss": 0.05023489,
+      "step": 23395
+    },
+    {
+      "epoch": 46.792,
+      "grad_norm": 1.1498349905014038,
+      "learning_rate": 2e-05,
+      "loss": 0.05116352,
+      "step": 23396
+    },
+    {
+      "epoch": 46.794,
+      "grad_norm": 2.005505084991455,
+      "learning_rate": 2e-05,
+      "loss": 0.04006296,
+      "step": 23397
+    },
+    {
+      "epoch": 46.796,
+      "grad_norm": 4.387477874755859,
+      "learning_rate": 2e-05,
+      "loss": 0.05183373,
+      "step": 23398
+    },
+    {
+      "epoch": 46.798,
+      "grad_norm": 2.7198569774627686,
+      "learning_rate": 2e-05,
+      "loss": 0.05188341,
+      "step": 23399
+    },
+    {
+      "epoch": 46.8,
+      "grad_norm": 1.1940276622772217,
+      "learning_rate": 2e-05,
+      "loss": 0.05574021,
+      "step": 23400
+    },
+    {
+      "epoch": 46.802,
+      "grad_norm": 1.0419164896011353,
+      "learning_rate": 2e-05,
+      "loss": 0.04050265,
+      "step": 23401
+    },
+    {
+      "epoch": 46.804,
+      "grad_norm": 3.387681484222412,
+      "learning_rate": 2e-05,
+      "loss": 0.04175478,
+      "step": 23402
+    },
+    {
+      "epoch": 46.806,
+      "grad_norm": 1.285062551498413,
+      "learning_rate": 2e-05,
+      "loss": 0.04971816,
+      "step": 23403
+    },
+    {
+      "epoch": 46.808,
+      "grad_norm": 1.3174958229064941,
+      "learning_rate": 2e-05,
+      "loss": 0.04881707,
+      "step": 23404
+    },
+    {
+      "epoch": 46.81,
+      "grad_norm": 3.6777937412261963,
+      "learning_rate": 2e-05,
+      "loss": 0.0468701,
+      "step": 23405
+    },
+    {
+      "epoch": 46.812,
+      "grad_norm": 1.1515544652938843,
+      "learning_rate": 2e-05,
+      "loss": 0.04995457,
+      "step": 23406
+    },
+    {
+      "epoch": 46.814,
+      "grad_norm": 1.6691585779190063,
+      "learning_rate": 2e-05,
+      "loss": 0.04617448,
+      "step": 23407
+    },
+    {
+      "epoch": 46.816,
+      "grad_norm": 1.3046892881393433,
+      "learning_rate": 2e-05,
+      "loss": 0.04137983,
+      "step": 23408
+    },
+    {
+      "epoch": 46.818,
+      "grad_norm": 1.3128255605697632,
+      "learning_rate": 2e-05,
+      "loss": 0.05065984,
+      "step": 23409
+    },
+    {
+      "epoch": 46.82,
+      "grad_norm": 1.3586230278015137,
+      "learning_rate": 2e-05,
+      "loss": 0.05450898,
+      "step": 23410
+    },
+    {
+      "epoch": 46.822,
+      "grad_norm": 1.2502920627593994,
+      "learning_rate": 2e-05,
+      "loss": 0.03689632,
+      "step": 23411
+    },
+    {
+      "epoch": 46.824,
+      "grad_norm": 1.1424442529678345,
+      "learning_rate": 2e-05,
+      "loss": 0.03632981,
+      "step": 23412
+    },
+    {
+      "epoch": 46.826,
+      "grad_norm": 1.3228057622909546,
+      "learning_rate": 2e-05,
+      "loss": 0.03984486,
+      "step": 23413
+    },
+    {
+      "epoch": 46.828,
+      "grad_norm": 1.0505292415618896,
+      "learning_rate": 2e-05,
+      "loss": 0.04190434,
+      "step": 23414
+    },
+    {
+      "epoch": 46.83,
+      "grad_norm": 1.3199437856674194,
+      "learning_rate": 2e-05,
+      "loss": 0.0435336,
+      "step": 23415
+    },
+    {
+      "epoch": 46.832,
+      "grad_norm": 2.5862584114074707,
+      "learning_rate": 2e-05,
+      "loss": 0.06176166,
+      "step": 23416
+    },
+    {
+      "epoch": 46.834,
+      "grad_norm": 1.1140897274017334,
+      "learning_rate": 2e-05,
+      "loss": 0.04548195,
+      "step": 23417
+    },
+    {
+      "epoch": 46.836,
+      "grad_norm": 1.4556688070297241,
+      "learning_rate": 2e-05,
+      "loss": 0.04776576,
+      "step": 23418
+    },
+    {
+      "epoch": 46.838,
+      "grad_norm": 2.7034659385681152,
+      "learning_rate": 2e-05,
+      "loss": 0.04163261,
+      "step": 23419
+    },
+    {
+      "epoch": 46.84,
+      "grad_norm": 1.0860694646835327,
+      "learning_rate": 2e-05,
+      "loss": 0.04266633,
+      "step": 23420
+    },
+    {
+      "epoch": 46.842,
+      "grad_norm": 1.2049165964126587,
+      "learning_rate": 2e-05,
+      "loss": 0.05778315,
+      "step": 23421
+    },
+    {
+      "epoch": 46.844,
+      "grad_norm": 1.6713166236877441,
+      "learning_rate": 2e-05,
+      "loss": 0.06890272,
+      "step": 23422
+    },
+    {
+      "epoch": 46.846,
+      "grad_norm": 3.021561622619629,
+      "learning_rate": 2e-05,
+      "loss": 0.05821927,
+      "step": 23423
+    },
+    {
+      "epoch": 46.848,
+      "grad_norm": 1.4241578578948975,
+      "learning_rate": 2e-05,
+      "loss": 0.0569243,
+      "step": 23424
+    },
+    {
+      "epoch": 46.85,
+      "grad_norm": 1.1604105234146118,
+      "learning_rate": 2e-05,
+      "loss": 0.0533638,
+      "step": 23425
+    },
+    {
+      "epoch": 46.852,
+      "grad_norm": 1.0262799263000488,
+      "learning_rate": 2e-05,
+      "loss": 0.04001687,
+      "step": 23426
+    },
+    {
+      "epoch": 46.854,
+      "grad_norm": 1.0680865049362183,
+      "learning_rate": 2e-05,
+      "loss": 0.04203242,
+      "step": 23427
+    },
+    {
+      "epoch": 46.856,
+      "grad_norm": 1.7087593078613281,
+      "learning_rate": 2e-05,
+      "loss": 0.04629037,
+      "step": 23428
+    },
+    {
+      "epoch": 46.858,
+      "grad_norm": 1.3922479152679443,
+      "learning_rate": 2e-05,
+      "loss": 0.05780581,
+      "step": 23429
+    },
+    {
+      "epoch": 46.86,
+      "grad_norm": 1.3703882694244385,
+      "learning_rate": 2e-05,
+      "loss": 0.05306429,
+      "step": 23430
+    },
+    {
+      "epoch": 46.862,
+      "grad_norm": 1.1331651210784912,
+      "learning_rate": 2e-05,
+      "loss": 0.05829866,
+      "step": 23431
+    },
+    {
+      "epoch": 46.864,
+      "grad_norm": 0.9955253601074219,
+      "learning_rate": 2e-05,
+      "loss": 0.03273296,
+      "step": 23432
+    },
+    {
+      "epoch": 46.866,
+      "grad_norm": 0.9291515946388245,
+      "learning_rate": 2e-05,
+      "loss": 0.03978248,
+      "step": 23433
+    },
+    {
+      "epoch": 46.868,
+      "grad_norm": 1.4390240907669067,
+      "learning_rate": 2e-05,
+      "loss": 0.06263384,
+      "step": 23434
+    },
+    {
+      "epoch": 46.87,
+      "grad_norm": 1.2633771896362305,
+      "learning_rate": 2e-05,
+      "loss": 0.05546718,
+      "step": 23435
+    },
+    {
+      "epoch": 46.872,
+      "grad_norm": 0.8656467199325562,
+      "learning_rate": 2e-05,
+      "loss": 0.03080487,
+      "step": 23436
+    },
+    {
+      "epoch": 46.874,
+      "grad_norm": 1.1050388813018799,
+      "learning_rate": 2e-05,
+      "loss": 0.05227588,
+      "step": 23437
+    },
+    {
+      "epoch": 46.876,
+      "grad_norm": 1.1237446069717407,
+      "learning_rate": 2e-05,
+      "loss": 0.06108324,
+      "step": 23438
+    },
+    {
+      "epoch": 46.878,
+      "grad_norm": 1.0278390645980835,
+      "learning_rate": 2e-05,
+      "loss": 0.03754516,
+      "step": 23439
+    },
+    {
+      "epoch": 46.88,
+      "grad_norm": 1.2690916061401367,
+      "learning_rate": 2e-05,
+      "loss": 0.05982792,
+      "step": 23440
+    },
+    {
+      "epoch": 46.882,
+      "grad_norm": 0.9860858917236328,
+      "learning_rate": 2e-05,
+      "loss": 0.04347444,
+      "step": 23441
+    },
+    {
+      "epoch": 46.884,
+      "grad_norm": 1.0543830394744873,
+      "learning_rate": 2e-05,
+      "loss": 0.03453768,
+      "step": 23442
+    },
+    {
+      "epoch": 46.886,
+      "grad_norm": 1.214526891708374,
+      "learning_rate": 2e-05,
+      "loss": 0.04959118,
+      "step": 23443
+    },
+    {
+      "epoch": 46.888,
+      "grad_norm": 1.9631701707839966,
+      "learning_rate": 2e-05,
+      "loss": 0.05660867,
+      "step": 23444
+    },
+    {
+      "epoch": 46.89,
+      "grad_norm": 1.270965337753296,
+      "learning_rate": 2e-05,
+      "loss": 0.05391593,
+      "step": 23445
+    },
+    {
+      "epoch": 46.892,
+      "grad_norm": 1.369183897972107,
+      "learning_rate": 2e-05,
+      "loss": 0.05142535,
+      "step": 23446
+    },
+    {
+      "epoch": 46.894,
+      "grad_norm": 3.9048008918762207,
+      "learning_rate": 2e-05,
+      "loss": 0.041148,
+      "step": 23447
+    },
+    {
+      "epoch": 46.896,
+      "grad_norm": 2.134962320327759,
+      "learning_rate": 2e-05,
+      "loss": 0.03332872,
+      "step": 23448
+    },
+    {
+      "epoch": 46.898,
+      "grad_norm": 1.895607352256775,
+      "learning_rate": 2e-05,
+      "loss": 0.0488515,
+      "step": 23449
+    },
+    {
+      "epoch": 46.9,
+      "grad_norm": 1.2083396911621094,
+      "learning_rate": 2e-05,
+      "loss": 0.04162563,
+      "step": 23450
+    },
+    {
+      "epoch": 46.902,
+      "grad_norm": 1.1905925273895264,
+      "learning_rate": 2e-05,
+      "loss": 0.03953518,
+      "step": 23451
+    },
+    {
+      "epoch": 46.904,
+      "grad_norm": 1.0462309122085571,
+      "learning_rate": 2e-05,
+      "loss": 0.04261171,
+      "step": 23452
+    },
+    {
+      "epoch": 46.906,
+      "grad_norm": 1.2605634927749634,
+      "learning_rate": 2e-05,
+      "loss": 0.04457783,
+      "step": 23453
+    },
+    {
+      "epoch": 46.908,
+      "grad_norm": 1.0726561546325684,
+      "learning_rate": 2e-05,
+      "loss": 0.04535056,
+      "step": 23454
+    },
+    {
+      "epoch": 46.91,
+      "grad_norm": 1.0684914588928223,
+      "learning_rate": 2e-05,
+      "loss": 0.04435501,
+      "step": 23455
+    },
+    {
+      "epoch": 46.912,
+      "grad_norm": 1.568894386291504,
+      "learning_rate": 2e-05,
+      "loss": 0.05053505,
+      "step": 23456
+    },
+    {
+      "epoch": 46.914,
+      "grad_norm": 1.0680261850357056,
+      "learning_rate": 2e-05,
+      "loss": 0.04221952,
+      "step": 23457
+    },
+    {
+      "epoch": 46.916,
+      "grad_norm": 1.2090034484863281,
+      "learning_rate": 2e-05,
+      "loss": 0.05732998,
+      "step": 23458
+    },
+    {
+      "epoch": 46.918,
+      "grad_norm": 1.0553538799285889,
+      "learning_rate": 2e-05,
+      "loss": 0.03570461,
+      "step": 23459
+    },
+    {
+      "epoch": 46.92,
+      "grad_norm": 1.3866254091262817,
+      "learning_rate": 2e-05,
+      "loss": 0.05774738,
+      "step": 23460
+    },
+    {
+      "epoch": 46.922,
+      "grad_norm": 1.218993067741394,
+      "learning_rate": 2e-05,
+      "loss": 0.05130921,
+      "step": 23461
+    },
+    {
+      "epoch": 46.924,
+      "grad_norm": 1.7904051542282104,
+      "learning_rate": 2e-05,
+      "loss": 0.05767091,
+      "step": 23462
+    },
+    {
+      "epoch": 46.926,
+      "grad_norm": 1.2114713191986084,
+      "learning_rate": 2e-05,
+      "loss": 0.05555228,
+      "step": 23463
+    },
+    {
+      "epoch": 46.928,
+      "grad_norm": 1.6034760475158691,
+      "learning_rate": 2e-05,
+      "loss": 0.056468,
+      "step": 23464
+    },
+    {
+      "epoch": 46.93,
+      "grad_norm": 1.8197824954986572,
+      "learning_rate": 2e-05,
+      "loss": 0.03906733,
+      "step": 23465
+    },
+    {
+      "epoch": 46.932,
+      "grad_norm": 2.5520434379577637,
+      "learning_rate": 2e-05,
+      "loss": 0.05988537,
+      "step": 23466
+    },
+    {
+      "epoch": 46.934,
+      "grad_norm": 1.1971849203109741,
+      "learning_rate": 2e-05,
+      "loss": 0.05080458,
+      "step": 23467
+    },
+    {
+      "epoch": 46.936,
+      "grad_norm": 1.2232027053833008,
+      "learning_rate": 2e-05,
+      "loss": 0.0653313,
+      "step": 23468
+    },
+    {
+      "epoch": 46.938,
+      "grad_norm": 1.3607845306396484,
+      "learning_rate": 2e-05,
+      "loss": 0.04467438,
+      "step": 23469
+    },
+    {
+      "epoch": 46.94,
+      "grad_norm": 1.129961371421814,
+      "learning_rate": 2e-05,
+      "loss": 0.05106883,
+      "step": 23470
+    },
+    {
+      "epoch": 46.942,
+      "grad_norm": 1.3855695724487305,
+      "learning_rate": 2e-05,
+      "loss": 0.0506998,
+      "step": 23471
+    },
+    {
+      "epoch": 46.944,
+      "grad_norm": 1.0359809398651123,
+      "learning_rate": 2e-05,
+      "loss": 0.04331663,
+      "step": 23472
+    },
+    {
+      "epoch": 46.946,
+      "grad_norm": 0.8526789546012878,
+      "learning_rate": 2e-05,
+      "loss": 0.03156386,
+      "step": 23473
+    },
+    {
+      "epoch": 46.948,
+      "grad_norm": 1.0871293544769287,
+      "learning_rate": 2e-05,
+      "loss": 0.04656047,
+      "step": 23474
+    },
+    {
+      "epoch": 46.95,
+      "grad_norm": 1.156686782836914,
+      "learning_rate": 2e-05,
+      "loss": 0.06759372,
+      "step": 23475
+    },
+    {
+      "epoch": 46.952,
+      "grad_norm": 2.5178990364074707,
+      "learning_rate": 2e-05,
+      "loss": 0.07468967,
+      "step": 23476
+    },
+    {
+      "epoch": 46.954,
+      "grad_norm": 1.1853928565979004,
+      "learning_rate": 2e-05,
+      "loss": 0.0469472,
+      "step": 23477
+    },
+    {
+      "epoch": 46.956,
+      "grad_norm": 1.038954257965088,
+      "learning_rate": 2e-05,
+      "loss": 0.03782792,
+      "step": 23478
+    },
+    {
+      "epoch": 46.958,
+      "grad_norm": 1.2154408693313599,
+      "learning_rate": 2e-05,
+      "loss": 0.05507773,
+      "step": 23479
+    },
+    {
+      "epoch": 46.96,
+      "grad_norm": 1.1402373313903809,
+      "learning_rate": 2e-05,
+      "loss": 0.0420175,
+      "step": 23480
+    },
+    {
+      "epoch": 46.962,
+      "grad_norm": 2.3009064197540283,
+      "learning_rate": 2e-05,
+      "loss": 0.04467285,
+      "step": 23481
+    },
+    {
+      "epoch": 46.964,
+      "grad_norm": 2.1940722465515137,
+      "learning_rate": 2e-05,
+      "loss": 0.05642062,
+      "step": 23482
+    },
+    {
+      "epoch": 46.966,
+      "grad_norm": 1.6159992218017578,
+      "learning_rate": 2e-05,
+      "loss": 0.04018041,
+      "step": 23483
+    },
+    {
+      "epoch": 46.968,
+      "grad_norm": 1.1516873836517334,
+      "learning_rate": 2e-05,
+      "loss": 0.04684192,
+      "step": 23484
+    },
+    {
+      "epoch": 46.97,
+      "grad_norm": 1.2707068920135498,
+      "learning_rate": 2e-05,
+      "loss": 0.04745233,
+      "step": 23485
+    },
+    {
+      "epoch": 46.972,
+      "grad_norm": 1.360982060432434,
+      "learning_rate": 2e-05,
+      "loss": 0.04974321,
+      "step": 23486
+    },
+    {
+      "epoch": 46.974,
+      "grad_norm": 1.3548970222473145,
+      "learning_rate": 2e-05,
+      "loss": 0.04797807,
+      "step": 23487
+    },
+    {
+      "epoch": 46.976,
+      "grad_norm": 1.2003024816513062,
+      "learning_rate": 2e-05,
+      "loss": 0.04765528,
+      "step": 23488
+    },
+    {
+      "epoch": 46.978,
+      "grad_norm": 1.032074213027954,
+      "learning_rate": 2e-05,
+      "loss": 0.03950176,
+      "step": 23489
+    },
+    {
+      "epoch": 46.98,
+      "grad_norm": 1.1735929250717163,
+      "learning_rate": 2e-05,
+      "loss": 0.05224654,
+      "step": 23490
+    },
+    {
+      "epoch": 46.982,
+      "grad_norm": 1.1766208410263062,
+      "learning_rate": 2e-05,
+      "loss": 0.03919265,
+      "step": 23491
+    },
+    {
+      "epoch": 46.984,
+      "grad_norm": 1.5051755905151367,
+      "learning_rate": 2e-05,
+      "loss": 0.04944252,
+      "step": 23492
+    },
+    {
+      "epoch": 46.986,
+      "grad_norm": 1.4483623504638672,
+      "learning_rate": 2e-05,
+      "loss": 0.04749278,
+      "step": 23493
+    },
+    {
+      "epoch": 46.988,
+      "grad_norm": 1.991761565208435,
+      "learning_rate": 2e-05,
+      "loss": 0.07904573,
+      "step": 23494
+    },
+    {
+      "epoch": 46.99,
+      "grad_norm": 2.063960552215576,
+      "learning_rate": 2e-05,
+      "loss": 0.06021646,
+      "step": 23495
+    },
+    {
+      "epoch": 46.992,
+      "grad_norm": 2.799022912979126,
+      "learning_rate": 2e-05,
+      "loss": 0.04049257,
+      "step": 23496
+    },
+    {
+      "epoch": 46.994,
+      "grad_norm": 1.7096704244613647,
+      "learning_rate": 2e-05,
+      "loss": 0.04271127,
+      "step": 23497
+    },
+    {
+      "epoch": 46.996,
+      "grad_norm": 0.984707236289978,
+      "learning_rate": 2e-05,
+      "loss": 0.03668294,
+      "step": 23498
+    },
+    {
+      "epoch": 46.998,
+      "grad_norm": 1.7703429460525513,
+      "learning_rate": 2e-05,
+      "loss": 0.05086554,
+      "step": 23499
+    },
+    {
+      "epoch": 47.0,
+      "grad_norm": 1.1229311227798462,
+      "learning_rate": 2e-05,
+      "loss": 0.04655139,
+      "step": 23500
+    },
+    {
+      "epoch": 47.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.998,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9820359281437125,
+        "Equal_1": 1.0,
+        "Equal_2": 0.9880239520958084,
+        "Equal_3": 0.9880239520958084,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 0.998003992015968,
+        "LineComparison_3": 0.998003992015968,
+        "Parallel_1": 0.9939879759519038,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.992,
+        "Perpendicular_1": 0.998,
+        "Perpendicular_2": 0.998,
+        "Perpendicular_3": 0.905811623246493,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 0.9976666666666667,
+        "PointLiesOnCircle_3": 0.9936,
+        "PointLiesOnLine_1": 1.0,
+        "PointLiesOnLine_2": 0.9939879759519038,
+        "PointLiesOnLine_3": 0.9860279441117764
+      },
+      "eval_runtime": 319.7259,
+      "eval_samples_per_second": 32.841,
+      "eval_steps_per_second": 0.657,
+      "step": 23500
+    },
+    {
+      "epoch": 47.002,
+      "grad_norm": 1.2124334573745728,
+      "learning_rate": 2e-05,
+      "loss": 0.04620025,
+      "step": 23501
+    },
+    {
+      "epoch": 47.004,
+      "grad_norm": 1.2111510038375854,
+      "learning_rate": 2e-05,
+      "loss": 0.03958354,
+      "step": 23502
+    },
+    {
+      "epoch": 47.006,
+      "grad_norm": 2.0674355030059814,
+      "learning_rate": 2e-05,
+      "loss": 0.06526448,
+      "step": 23503
+    },
+    {
+      "epoch": 47.008,
+      "grad_norm": 1.139857530593872,
+      "learning_rate": 2e-05,
+      "loss": 0.0366558,
+      "step": 23504
+    },
+    {
+      "epoch": 47.01,
+      "grad_norm": 1.407399296760559,
+      "learning_rate": 2e-05,
+      "loss": 0.05581947,
+      "step": 23505
+    },
+    {
+      "epoch": 47.012,
+      "grad_norm": 1.1468596458435059,
+      "learning_rate": 2e-05,
+      "loss": 0.04368159,
+      "step": 23506
+    },
+    {
+      "epoch": 47.014,
+      "grad_norm": 1.4752362966537476,
+      "learning_rate": 2e-05,
+      "loss": 0.03284196,
+      "step": 23507
+    },
+    {
+      "epoch": 47.016,
+      "grad_norm": 1.083961009979248,
+      "learning_rate": 2e-05,
+      "loss": 0.04144663,
+      "step": 23508
+    },
+    {
+      "epoch": 47.018,
+      "grad_norm": 1.0464342832565308,
+      "learning_rate": 2e-05,
+      "loss": 0.04529237,
+      "step": 23509
+    },
+    {
+      "epoch": 47.02,
+      "grad_norm": 1.1925157308578491,
+      "learning_rate": 2e-05,
+      "loss": 0.05837334,
+      "step": 23510
+    },
+    {
+      "epoch": 47.022,
+      "grad_norm": 1.135394811630249,
+      "learning_rate": 2e-05,
+      "loss": 0.04342172,
+      "step": 23511
+    },
+    {
+      "epoch": 47.024,
+      "grad_norm": 1.2975212335586548,
+      "learning_rate": 2e-05,
+      "loss": 0.05698482,
+      "step": 23512
+    },
+    {
+      "epoch": 47.026,
+      "grad_norm": 0.9662984609603882,
+      "learning_rate": 2e-05,
+      "loss": 0.03092372,
+      "step": 23513
+    },
+    {
+      "epoch": 47.028,
+      "grad_norm": 1.3168891668319702,
+      "learning_rate": 2e-05,
+      "loss": 0.05116647,
+      "step": 23514
+    },
+    {
+      "epoch": 47.03,
+      "grad_norm": 1.0158759355545044,
+      "learning_rate": 2e-05,
+      "loss": 0.04270767,
+      "step": 23515
+    },
+    {
+      "epoch": 47.032,
+      "grad_norm": 1.0800869464874268,
+      "learning_rate": 2e-05,
+      "loss": 0.0404634,
+      "step": 23516
+    },
+    {
+      "epoch": 47.034,
+      "grad_norm": 1.3503354787826538,
+      "learning_rate": 2e-05,
+      "loss": 0.06739471,
+      "step": 23517
+    },
+    {
+      "epoch": 47.036,
+      "grad_norm": 1.1669559478759766,
+      "learning_rate": 2e-05,
+      "loss": 0.04633216,
+      "step": 23518
+    },
+    {
+      "epoch": 47.038,
+      "grad_norm": 1.211911678314209,
+      "learning_rate": 2e-05,
+      "loss": 0.05827531,
+      "step": 23519
+    },
+    {
+      "epoch": 47.04,
+      "grad_norm": 1.1868705749511719,
+      "learning_rate": 2e-05,
+      "loss": 0.04901566,
+      "step": 23520
+    },
+    {
+      "epoch": 47.042,
+      "grad_norm": 0.9578158855438232,
+      "learning_rate": 2e-05,
+      "loss": 0.04152099,
+      "step": 23521
+    },
+    {
+      "epoch": 47.044,
+      "grad_norm": 1.1845344305038452,
+      "learning_rate": 2e-05,
+      "loss": 0.04116198,
+      "step": 23522
+    },
+    {
+      "epoch": 47.046,
+      "grad_norm": 1.7209062576293945,
+      "learning_rate": 2e-05,
+      "loss": 0.05426127,
+      "step": 23523
+    },
+    {
+      "epoch": 47.048,
+      "grad_norm": 1.0993847846984863,
+      "learning_rate": 2e-05,
+      "loss": 0.04611755,
+      "step": 23524
+    },
+    {
+      "epoch": 47.05,
+      "grad_norm": 1.5714466571807861,
+      "learning_rate": 2e-05,
+      "loss": 0.06220062,
+      "step": 23525
+    },
+    {
+      "epoch": 47.052,
+      "grad_norm": 1.5425214767456055,
+      "learning_rate": 2e-05,
+      "loss": 0.04734546,
+      "step": 23526
+    },
+    {
+      "epoch": 47.054,
+      "grad_norm": 1.330129623413086,
+      "learning_rate": 2e-05,
+      "loss": 0.0763645,
+      "step": 23527
+    },
+    {
+      "epoch": 47.056,
+      "grad_norm": 1.205364465713501,
+      "learning_rate": 2e-05,
+      "loss": 0.06087791,
+      "step": 23528
+    },
+    {
+      "epoch": 47.058,
+      "grad_norm": 1.8291528224945068,
+      "learning_rate": 2e-05,
+      "loss": 0.05230067,
+      "step": 23529
+    },
+    {
+      "epoch": 47.06,
+      "grad_norm": 1.2380144596099854,
+      "learning_rate": 2e-05,
+      "loss": 0.05399077,
+      "step": 23530
+    },
+    {
+      "epoch": 47.062,
+      "grad_norm": 1.1345731019973755,
+      "learning_rate": 2e-05,
+      "loss": 0.04412289,
+      "step": 23531
+    },
+    {
+      "epoch": 47.064,
+      "grad_norm": 1.3171319961547852,
+      "learning_rate": 2e-05,
+      "loss": 0.05258146,
+      "step": 23532
+    },
+    {
+      "epoch": 47.066,
+      "grad_norm": 1.1284446716308594,
+      "learning_rate": 2e-05,
+      "loss": 0.04565362,
+      "step": 23533
+    },
+    {
+      "epoch": 47.068,
+      "grad_norm": 1.0812745094299316,
+      "learning_rate": 2e-05,
+      "loss": 0.03943041,
+      "step": 23534
+    },
+    {
+      "epoch": 47.07,
+      "grad_norm": 1.1932939291000366,
+      "learning_rate": 2e-05,
+      "loss": 0.03978365,
+      "step": 23535
+    },
+    {
+      "epoch": 47.072,
+      "grad_norm": 1.5841753482818604,
+      "learning_rate": 2e-05,
+      "loss": 0.04983898,
+      "step": 23536
+    },
+    {
+      "epoch": 47.074,
+      "grad_norm": 1.2005212306976318,
+      "learning_rate": 2e-05,
+      "loss": 0.0624929,
+      "step": 23537
+    },
+    {
+      "epoch": 47.076,
+      "grad_norm": 1.3606860637664795,
+      "learning_rate": 2e-05,
+      "loss": 0.0616478,
+      "step": 23538
+    },
+    {
+      "epoch": 47.078,
+      "grad_norm": 1.0764256715774536,
+      "learning_rate": 2e-05,
+      "loss": 0.04636963,
+      "step": 23539
+    },
+    {
+      "epoch": 47.08,
+      "grad_norm": 1.1802319288253784,
+      "learning_rate": 2e-05,
+      "loss": 0.05296212,
+      "step": 23540
+    },
+    {
+      "epoch": 47.082,
+      "grad_norm": 1.249846339225769,
+      "learning_rate": 2e-05,
+      "loss": 0.05203488,
+      "step": 23541
+    },
+    {
+      "epoch": 47.084,
+      "grad_norm": 1.3313225507736206,
+      "learning_rate": 2e-05,
+      "loss": 0.05261878,
+      "step": 23542
+    },
+    {
+      "epoch": 47.086,
+      "grad_norm": 1.1203484535217285,
+      "learning_rate": 2e-05,
+      "loss": 0.04687518,
+      "step": 23543
+    },
+    {
+      "epoch": 47.088,
+      "grad_norm": 1.1626969575881958,
+      "learning_rate": 2e-05,
+      "loss": 0.04210647,
+      "step": 23544
+    },
+    {
+      "epoch": 47.09,
+      "grad_norm": 1.0756282806396484,
+      "learning_rate": 2e-05,
+      "loss": 0.04296558,
+      "step": 23545
+    },
+    {
+      "epoch": 47.092,
+      "grad_norm": 1.29530930519104,
+      "learning_rate": 2e-05,
+      "loss": 0.05416109,
+      "step": 23546
+    },
+    {
+      "epoch": 47.094,
+      "grad_norm": 1.05868661403656,
+      "learning_rate": 2e-05,
+      "loss": 0.04070142,
+      "step": 23547
+    },
+    {
+      "epoch": 47.096,
+      "grad_norm": 1.0833359956741333,
+      "learning_rate": 2e-05,
+      "loss": 0.0563876,
+      "step": 23548
+    },
+    {
+      "epoch": 47.098,
+      "grad_norm": 1.1828356981277466,
+      "learning_rate": 2e-05,
+      "loss": 0.05474529,
+      "step": 23549
+    },
+    {
+      "epoch": 47.1,
+      "grad_norm": 1.6338591575622559,
+      "learning_rate": 2e-05,
+      "loss": 0.04851716,
+      "step": 23550
+    },
+    {
+      "epoch": 47.102,
+      "grad_norm": 1.1993439197540283,
+      "learning_rate": 2e-05,
+      "loss": 0.05251703,
+      "step": 23551
+    },
+    {
+      "epoch": 47.104,
+      "grad_norm": 1.1439919471740723,
+      "learning_rate": 2e-05,
+      "loss": 0.04419599,
+      "step": 23552
+    },
+    {
+      "epoch": 47.106,
+      "grad_norm": 1.2255439758300781,
+      "learning_rate": 2e-05,
+      "loss": 0.04737487,
+      "step": 23553
+    },
+    {
+      "epoch": 47.108,
+      "grad_norm": 1.085721492767334,
+      "learning_rate": 2e-05,
+      "loss": 0.03265755,
+      "step": 23554
+    },
+    {
+      "epoch": 47.11,
+      "grad_norm": 1.1767328977584839,
+      "learning_rate": 2e-05,
+      "loss": 0.05188051,
+      "step": 23555
+    },
+    {
+      "epoch": 47.112,
+      "grad_norm": 1.1831814050674438,
+      "learning_rate": 2e-05,
+      "loss": 0.06118751,
+      "step": 23556
+    },
+    {
+      "epoch": 47.114,
+      "grad_norm": 1.093849778175354,
+      "learning_rate": 2e-05,
+      "loss": 0.04082596,
+      "step": 23557
+    },
+    {
+      "epoch": 47.116,
+      "grad_norm": 1.1168532371520996,
+      "learning_rate": 2e-05,
+      "loss": 0.05243224,
+      "step": 23558
+    },
+    {
+      "epoch": 47.118,
+      "grad_norm": 1.548278570175171,
+      "learning_rate": 2e-05,
+      "loss": 0.05186,
+      "step": 23559
+    },
+    {
+      "epoch": 47.12,
+      "grad_norm": 1.2307542562484741,
+      "learning_rate": 2e-05,
+      "loss": 0.05896568,
+      "step": 23560
+    },
+    {
+      "epoch": 47.122,
+      "grad_norm": 1.2599307298660278,
+      "learning_rate": 2e-05,
+      "loss": 0.04969592,
+      "step": 23561
+    },
+    {
+      "epoch": 47.124,
+      "grad_norm": 1.4370466470718384,
+      "learning_rate": 2e-05,
+      "loss": 0.06255047,
+      "step": 23562
+    },
+    {
+      "epoch": 47.126,
+      "grad_norm": 1.6804131269454956,
+      "learning_rate": 2e-05,
+      "loss": 0.04512757,
+      "step": 23563
+    },
+    {
+      "epoch": 47.128,
+      "grad_norm": 1.4804704189300537,
+      "learning_rate": 2e-05,
+      "loss": 0.04683591,
+      "step": 23564
+    },
+    {
+      "epoch": 47.13,
+      "grad_norm": 2.8036322593688965,
+      "learning_rate": 2e-05,
+      "loss": 0.06730855,
+      "step": 23565
+    },
+    {
+      "epoch": 47.132,
+      "grad_norm": 0.9568971395492554,
+      "learning_rate": 2e-05,
+      "loss": 0.04147626,
+      "step": 23566
+    },
+    {
+      "epoch": 47.134,
+      "grad_norm": 1.3193092346191406,
+      "learning_rate": 2e-05,
+      "loss": 0.05314566,
+      "step": 23567
+    },
+    {
+      "epoch": 47.136,
+      "grad_norm": 1.122524380683899,
+      "learning_rate": 2e-05,
+      "loss": 0.04262283,
+      "step": 23568
+    },
+    {
+      "epoch": 47.138,
+      "grad_norm": 1.9593148231506348,
+      "learning_rate": 2e-05,
+      "loss": 0.05774532,
+      "step": 23569
+    },
+    {
+      "epoch": 47.14,
+      "grad_norm": 1.1386750936508179,
+      "learning_rate": 2e-05,
+      "loss": 0.0475467,
+      "step": 23570
+    },
+    {
+      "epoch": 47.142,
+      "grad_norm": 1.329964518547058,
+      "learning_rate": 2e-05,
+      "loss": 0.04642937,
+      "step": 23571
+    },
+    {
+      "epoch": 47.144,
+      "grad_norm": 1.109947681427002,
+      "learning_rate": 2e-05,
+      "loss": 0.05008635,
+      "step": 23572
+    },
+    {
+      "epoch": 47.146,
+      "grad_norm": 1.2006001472473145,
+      "learning_rate": 2e-05,
+      "loss": 0.05233764,
+      "step": 23573
+    },
+    {
+      "epoch": 47.148,
+      "grad_norm": 1.4829745292663574,
+      "learning_rate": 2e-05,
+      "loss": 0.05814692,
+      "step": 23574
+    },
+    {
+      "epoch": 47.15,
+      "grad_norm": 1.1940250396728516,
+      "learning_rate": 2e-05,
+      "loss": 0.05612388,
+      "step": 23575
+    },
+    {
+      "epoch": 47.152,
+      "grad_norm": 1.3208078145980835,
+      "learning_rate": 2e-05,
+      "loss": 0.04610663,
+      "step": 23576
+    },
+    {
+      "epoch": 47.154,
+      "grad_norm": 1.1997742652893066,
+      "learning_rate": 2e-05,
+      "loss": 0.04106606,
+      "step": 23577
+    },
+    {
+      "epoch": 47.156,
+      "grad_norm": 1.40494704246521,
+      "learning_rate": 2e-05,
+      "loss": 0.06252164,
+      "step": 23578
+    },
+    {
+      "epoch": 47.158,
+      "grad_norm": 0.9245737791061401,
+      "learning_rate": 2e-05,
+      "loss": 0.04033583,
+      "step": 23579
+    },
+    {
+      "epoch": 47.16,
+      "grad_norm": 1.717743992805481,
+      "learning_rate": 2e-05,
+      "loss": 0.06210865,
+      "step": 23580
+    },
+    {
+      "epoch": 47.162,
+      "grad_norm": 1.206834316253662,
+      "learning_rate": 2e-05,
+      "loss": 0.05354768,
+      "step": 23581
+    },
+    {
+      "epoch": 47.164,
+      "grad_norm": 1.2854136228561401,
+      "learning_rate": 2e-05,
+      "loss": 0.05695827,
+      "step": 23582
+    },
+    {
+      "epoch": 47.166,
+      "grad_norm": 1.6390953063964844,
+      "learning_rate": 2e-05,
+      "loss": 0.06032401,
+      "step": 23583
+    },
+    {
+      "epoch": 47.168,
+      "grad_norm": 1.2514336109161377,
+      "learning_rate": 2e-05,
+      "loss": 0.05170215,
+      "step": 23584
+    },
+    {
+      "epoch": 47.17,
+      "grad_norm": 1.274780035018921,
+      "learning_rate": 2e-05,
+      "loss": 0.03985206,
+      "step": 23585
+    },
+    {
+      "epoch": 47.172,
+      "grad_norm": 1.108774185180664,
+      "learning_rate": 2e-05,
+      "loss": 0.03357091,
+      "step": 23586
+    },
+    {
+      "epoch": 47.174,
+      "grad_norm": 1.0785712003707886,
+      "learning_rate": 2e-05,
+      "loss": 0.03320477,
+      "step": 23587
+    },
+    {
+      "epoch": 47.176,
+      "grad_norm": 1.3981683254241943,
+      "learning_rate": 2e-05,
+      "loss": 0.03950857,
+      "step": 23588
+    },
+    {
+      "epoch": 47.178,
+      "grad_norm": 1.2349064350128174,
+      "learning_rate": 2e-05,
+      "loss": 0.06858353,
+      "step": 23589
+    },
+    {
+      "epoch": 47.18,
+      "grad_norm": 2.030306100845337,
+      "learning_rate": 2e-05,
+      "loss": 0.04219834,
+      "step": 23590
+    },
+    {
+      "epoch": 47.182,
+      "grad_norm": 1.5019062757492065,
+      "learning_rate": 2e-05,
+      "loss": 0.05146636,
+      "step": 23591
+    },
+    {
+      "epoch": 47.184,
+      "grad_norm": 1.3464275598526,
+      "learning_rate": 2e-05,
+      "loss": 0.05433298,
+      "step": 23592
+    },
+    {
+      "epoch": 47.186,
+      "grad_norm": 1.466103196144104,
+      "learning_rate": 2e-05,
+      "loss": 0.06641914,
+      "step": 23593
+    },
+    {
+      "epoch": 47.188,
+      "grad_norm": 1.3546744585037231,
+      "learning_rate": 2e-05,
+      "loss": 0.05045114,
+      "step": 23594
+    },
+    {
+      "epoch": 47.19,
+      "grad_norm": 1.330115795135498,
+      "learning_rate": 2e-05,
+      "loss": 0.0545896,
+      "step": 23595
+    },
+    {
+      "epoch": 47.192,
+      "grad_norm": 1.141314148902893,
+      "learning_rate": 2e-05,
+      "loss": 0.03988175,
+      "step": 23596
+    },
+    {
+      "epoch": 47.194,
+      "grad_norm": 1.3933604955673218,
+      "learning_rate": 2e-05,
+      "loss": 0.05277719,
+      "step": 23597
+    },
+    {
+      "epoch": 47.196,
+      "grad_norm": 1.3836783170700073,
+      "learning_rate": 2e-05,
+      "loss": 0.05862165,
+      "step": 23598
+    },
+    {
+      "epoch": 47.198,
+      "grad_norm": 1.1174200773239136,
+      "learning_rate": 2e-05,
+      "loss": 0.04964589,
+      "step": 23599
+    },
+    {
+      "epoch": 47.2,
+      "grad_norm": 1.247449278831482,
+      "learning_rate": 2e-05,
+      "loss": 0.05483104,
+      "step": 23600
+    },
+    {
+      "epoch": 47.202,
+      "grad_norm": 1.1020945310592651,
+      "learning_rate": 2e-05,
+      "loss": 0.05271735,
+      "step": 23601
+    },
+    {
+      "epoch": 47.204,
+      "grad_norm": 1.5037002563476562,
+      "learning_rate": 2e-05,
+      "loss": 0.06174751,
+      "step": 23602
+    },
+    {
+      "epoch": 47.206,
+      "grad_norm": 1.4615405797958374,
+      "learning_rate": 2e-05,
+      "loss": 0.05728906,
+      "step": 23603
+    },
+    {
+      "epoch": 47.208,
+      "grad_norm": 1.1932406425476074,
+      "learning_rate": 2e-05,
+      "loss": 0.05166669,
+      "step": 23604
+    },
+    {
+      "epoch": 47.21,
+      "grad_norm": 1.133001685142517,
+      "learning_rate": 2e-05,
+      "loss": 0.05504667,
+      "step": 23605
+    },
+    {
+      "epoch": 47.212,
+      "grad_norm": 1.0907678604125977,
+      "learning_rate": 2e-05,
+      "loss": 0.04519046,
+      "step": 23606
+    },
+    {
+      "epoch": 47.214,
+      "grad_norm": 1.0519095659255981,
+      "learning_rate": 2e-05,
+      "loss": 0.0439795,
+      "step": 23607
+    },
+    {
+      "epoch": 47.216,
+      "grad_norm": 1.272680401802063,
+      "learning_rate": 2e-05,
+      "loss": 0.04971275,
+      "step": 23608
+    },
+    {
+      "epoch": 47.218,
+      "grad_norm": 1.4395430088043213,
+      "learning_rate": 2e-05,
+      "loss": 0.05114156,
+      "step": 23609
+    },
+    {
+      "epoch": 47.22,
+      "grad_norm": 1.1029596328735352,
+      "learning_rate": 2e-05,
+      "loss": 0.04121825,
+      "step": 23610
+    },
+    {
+      "epoch": 47.222,
+      "grad_norm": 1.232066035270691,
+      "learning_rate": 2e-05,
+      "loss": 0.04981194,
+      "step": 23611
+    },
+    {
+      "epoch": 47.224,
+      "grad_norm": 1.2140462398529053,
+      "learning_rate": 2e-05,
+      "loss": 0.05153929,
+      "step": 23612
+    },
+    {
+      "epoch": 47.226,
+      "grad_norm": 1.5724217891693115,
+      "learning_rate": 2e-05,
+      "loss": 0.05892528,
+      "step": 23613
+    },
+    {
+      "epoch": 47.228,
+      "grad_norm": 1.089890480041504,
+      "learning_rate": 2e-05,
+      "loss": 0.03622951,
+      "step": 23614
+    },
+    {
+      "epoch": 47.23,
+      "grad_norm": 1.137673258781433,
+      "learning_rate": 2e-05,
+      "loss": 0.05067534,
+      "step": 23615
+    },
+    {
+      "epoch": 47.232,
+      "grad_norm": 1.6163313388824463,
+      "learning_rate": 2e-05,
+      "loss": 0.07416356,
+      "step": 23616
+    },
+    {
+      "epoch": 47.234,
+      "grad_norm": 1.5656497478485107,
+      "learning_rate": 2e-05,
+      "loss": 0.04488578,
+      "step": 23617
+    },
+    {
+      "epoch": 47.236,
+      "grad_norm": 2.585951566696167,
+      "learning_rate": 2e-05,
+      "loss": 0.0466084,
+      "step": 23618
+    },
+    {
+      "epoch": 47.238,
+      "grad_norm": 1.1935782432556152,
+      "learning_rate": 2e-05,
+      "loss": 0.03517368,
+      "step": 23619
+    },
+    {
+      "epoch": 47.24,
+      "grad_norm": 0.9748817682266235,
+      "learning_rate": 2e-05,
+      "loss": 0.03498342,
+      "step": 23620
+    },
+    {
+      "epoch": 47.242,
+      "grad_norm": 1.339208960533142,
+      "learning_rate": 2e-05,
+      "loss": 0.04053495,
+      "step": 23621
+    },
+    {
+      "epoch": 47.244,
+      "grad_norm": 1.0967516899108887,
+      "learning_rate": 2e-05,
+      "loss": 0.03619142,
+      "step": 23622
+    },
+    {
+      "epoch": 47.246,
+      "grad_norm": 1.4304256439208984,
+      "learning_rate": 2e-05,
+      "loss": 0.05116164,
+      "step": 23623
+    },
+    {
+      "epoch": 47.248,
+      "grad_norm": 1.2332102060317993,
+      "learning_rate": 2e-05,
+      "loss": 0.05349259,
+      "step": 23624
+    },
+    {
+      "epoch": 47.25,
+      "grad_norm": 1.084965467453003,
+      "learning_rate": 2e-05,
+      "loss": 0.04240631,
+      "step": 23625
+    },
+    {
+      "epoch": 47.252,
+      "grad_norm": 1.075219750404358,
+      "learning_rate": 2e-05,
+      "loss": 0.04627833,
+      "step": 23626
+    },
+    {
+      "epoch": 47.254,
+      "grad_norm": 1.1400893926620483,
+      "learning_rate": 2e-05,
+      "loss": 0.04680464,
+      "step": 23627
+    },
+    {
+      "epoch": 47.256,
+      "grad_norm": 1.1270958185195923,
+      "learning_rate": 2e-05,
+      "loss": 0.04278168,
+      "step": 23628
+    },
+    {
+      "epoch": 47.258,
+      "grad_norm": 2.2311246395111084,
+      "learning_rate": 2e-05,
+      "loss": 0.05921525,
+      "step": 23629
+    },
+    {
+      "epoch": 47.26,
+      "grad_norm": 1.1150248050689697,
+      "learning_rate": 2e-05,
+      "loss": 0.04678026,
+      "step": 23630
+    },
+    {
+      "epoch": 47.262,
+      "grad_norm": 1.1364946365356445,
+      "learning_rate": 2e-05,
+      "loss": 0.05590563,
+      "step": 23631
+    },
+    {
+      "epoch": 47.264,
+      "grad_norm": 1.2034296989440918,
+      "learning_rate": 2e-05,
+      "loss": 0.05708367,
+      "step": 23632
+    },
+    {
+      "epoch": 47.266,
+      "grad_norm": 1.4349358081817627,
+      "learning_rate": 2e-05,
+      "loss": 0.06511645,
+      "step": 23633
+    },
+    {
+      "epoch": 47.268,
+      "grad_norm": 1.4762065410614014,
+      "learning_rate": 2e-05,
+      "loss": 0.06263445,
+      "step": 23634
+    },
+    {
+      "epoch": 47.27,
+      "grad_norm": 0.9478891491889954,
+      "learning_rate": 2e-05,
+      "loss": 0.03476822,
+      "step": 23635
+    },
+    {
+      "epoch": 47.272,
+      "grad_norm": 1.010331153869629,
+      "learning_rate": 2e-05,
+      "loss": 0.02933269,
+      "step": 23636
+    },
+    {
+      "epoch": 47.274,
+      "grad_norm": 1.250553011894226,
+      "learning_rate": 2e-05,
+      "loss": 0.04823877,
+      "step": 23637
+    },
+    {
+      "epoch": 47.276,
+      "grad_norm": 0.9914244413375854,
+      "learning_rate": 2e-05,
+      "loss": 0.02885609,
+      "step": 23638
+    },
+    {
+      "epoch": 47.278,
+      "grad_norm": 1.766867995262146,
+      "learning_rate": 2e-05,
+      "loss": 0.05482132,
+      "step": 23639
+    },
+    {
+      "epoch": 47.28,
+      "grad_norm": 1.9935588836669922,
+      "learning_rate": 2e-05,
+      "loss": 0.05211507,
+      "step": 23640
+    },
+    {
+      "epoch": 47.282,
+      "grad_norm": 1.3328672647476196,
+      "learning_rate": 2e-05,
+      "loss": 0.05550382,
+      "step": 23641
+    },
+    {
+      "epoch": 47.284,
+      "grad_norm": 1.2937812805175781,
+      "learning_rate": 2e-05,
+      "loss": 0.05083098,
+      "step": 23642
+    },
+    {
+      "epoch": 47.286,
+      "grad_norm": 1.1793773174285889,
+      "learning_rate": 2e-05,
+      "loss": 0.04735196,
+      "step": 23643
+    },
+    {
+      "epoch": 47.288,
+      "grad_norm": 0.9814954400062561,
+      "learning_rate": 2e-05,
+      "loss": 0.03630852,
+      "step": 23644
+    },
+    {
+      "epoch": 47.29,
+      "grad_norm": 1.104866862297058,
+      "learning_rate": 2e-05,
+      "loss": 0.04346426,
+      "step": 23645
+    },
+    {
+      "epoch": 47.292,
+      "grad_norm": 1.7783113718032837,
+      "learning_rate": 2e-05,
+      "loss": 0.05869078,
+      "step": 23646
+    },
+    {
+      "epoch": 47.294,
+      "grad_norm": 1.1085009574890137,
+      "learning_rate": 2e-05,
+      "loss": 0.05062877,
+      "step": 23647
+    },
+    {
+      "epoch": 47.296,
+      "grad_norm": 1.106187343597412,
+      "learning_rate": 2e-05,
+      "loss": 0.05129405,
+      "step": 23648
+    },
+    {
+      "epoch": 47.298,
+      "grad_norm": 1.0873476266860962,
+      "learning_rate": 2e-05,
+      "loss": 0.03707412,
+      "step": 23649
+    },
+    {
+      "epoch": 47.3,
+      "grad_norm": 1.7342414855957031,
+      "learning_rate": 2e-05,
+      "loss": 0.05554866,
+      "step": 23650
+    },
+    {
+      "epoch": 47.302,
+      "grad_norm": 0.9549355506896973,
+      "learning_rate": 2e-05,
+      "loss": 0.03941218,
+      "step": 23651
+    },
+    {
+      "epoch": 47.304,
+      "grad_norm": 1.0804848670959473,
+      "learning_rate": 2e-05,
+      "loss": 0.04693265,
+      "step": 23652
+    },
+    {
+      "epoch": 47.306,
+      "grad_norm": 1.6769800186157227,
+      "learning_rate": 2e-05,
+      "loss": 0.05842284,
+      "step": 23653
+    },
+    {
+      "epoch": 47.308,
+      "grad_norm": 1.0862559080123901,
+      "learning_rate": 2e-05,
+      "loss": 0.04118923,
+      "step": 23654
+    },
+    {
+      "epoch": 47.31,
+      "grad_norm": 1.7151005268096924,
+      "learning_rate": 2e-05,
+      "loss": 0.07205494,
+      "step": 23655
+    },
+    {
+      "epoch": 47.312,
+      "grad_norm": 1.1123970746994019,
+      "learning_rate": 2e-05,
+      "loss": 0.04154256,
+      "step": 23656
+    },
+    {
+      "epoch": 47.314,
+      "grad_norm": 1.0019702911376953,
+      "learning_rate": 2e-05,
+      "loss": 0.03406595,
+      "step": 23657
+    },
+    {
+      "epoch": 47.316,
+      "grad_norm": 1.2018071413040161,
+      "learning_rate": 2e-05,
+      "loss": 0.04904677,
+      "step": 23658
+    },
+    {
+      "epoch": 47.318,
+      "grad_norm": 1.2852320671081543,
+      "learning_rate": 2e-05,
+      "loss": 0.03975239,
+      "step": 23659
+    },
+    {
+      "epoch": 47.32,
+      "grad_norm": 1.0827425718307495,
+      "learning_rate": 2e-05,
+      "loss": 0.0492649,
+      "step": 23660
+    },
+    {
+      "epoch": 47.322,
+      "grad_norm": 1.0454564094543457,
+      "learning_rate": 2e-05,
+      "loss": 0.03652071,
+      "step": 23661
+    },
+    {
+      "epoch": 47.324,
+      "grad_norm": 1.3567827939987183,
+      "learning_rate": 2e-05,
+      "loss": 0.05563585,
+      "step": 23662
+    },
+    {
+      "epoch": 47.326,
+      "grad_norm": 2.5465247631073,
+      "learning_rate": 2e-05,
+      "loss": 0.06577208,
+      "step": 23663
+    },
+    {
+      "epoch": 47.328,
+      "grad_norm": 1.4325695037841797,
+      "learning_rate": 2e-05,
+      "loss": 0.06627315,
+      "step": 23664
+    },
+    {
+      "epoch": 47.33,
+      "grad_norm": 1.2820574045181274,
+      "learning_rate": 2e-05,
+      "loss": 0.05097567,
+      "step": 23665
+    },
+    {
+      "epoch": 47.332,
+      "grad_norm": 0.9658817648887634,
+      "learning_rate": 2e-05,
+      "loss": 0.03891777,
+      "step": 23666
+    },
+    {
+      "epoch": 47.334,
+      "grad_norm": 0.9408270716667175,
+      "learning_rate": 2e-05,
+      "loss": 0.03226583,
+      "step": 23667
+    },
+    {
+      "epoch": 47.336,
+      "grad_norm": 1.1844184398651123,
+      "learning_rate": 2e-05,
+      "loss": 0.04694714,
+      "step": 23668
+    },
+    {
+      "epoch": 47.338,
+      "grad_norm": 1.1536945104599,
+      "learning_rate": 2e-05,
+      "loss": 0.04273265,
+      "step": 23669
+    },
+    {
+      "epoch": 47.34,
+      "grad_norm": 1.163088083267212,
+      "learning_rate": 2e-05,
+      "loss": 0.03042351,
+      "step": 23670
+    },
+    {
+      "epoch": 47.342,
+      "grad_norm": 1.5534110069274902,
+      "learning_rate": 2e-05,
+      "loss": 0.04814804,
+      "step": 23671
+    },
+    {
+      "epoch": 47.344,
+      "grad_norm": 1.2012584209442139,
+      "learning_rate": 2e-05,
+      "loss": 0.04714233,
+      "step": 23672
+    },
+    {
+      "epoch": 47.346,
+      "grad_norm": 1.1810247898101807,
+      "learning_rate": 2e-05,
+      "loss": 0.03760591,
+      "step": 23673
+    },
+    {
+      "epoch": 47.348,
+      "grad_norm": 1.1035114526748657,
+      "learning_rate": 2e-05,
+      "loss": 0.04680648,
+      "step": 23674
+    },
+    {
+      "epoch": 47.35,
+      "grad_norm": 0.9210943579673767,
+      "learning_rate": 2e-05,
+      "loss": 0.03328598,
+      "step": 23675
+    },
+    {
+      "epoch": 47.352,
+      "grad_norm": 1.0501015186309814,
+      "learning_rate": 2e-05,
+      "loss": 0.03831122,
+      "step": 23676
+    },
+    {
+      "epoch": 47.354,
+      "grad_norm": 2.732465982437134,
+      "learning_rate": 2e-05,
+      "loss": 0.05034593,
+      "step": 23677
+    },
+    {
+      "epoch": 47.356,
+      "grad_norm": 0.9773300290107727,
+      "learning_rate": 2e-05,
+      "loss": 0.03243728,
+      "step": 23678
+    },
+    {
+      "epoch": 47.358,
+      "grad_norm": 1.2418444156646729,
+      "learning_rate": 2e-05,
+      "loss": 0.05027065,
+      "step": 23679
+    },
+    {
+      "epoch": 47.36,
+      "grad_norm": 1.1005994081497192,
+      "learning_rate": 2e-05,
+      "loss": 0.04392687,
+      "step": 23680
+    },
+    {
+      "epoch": 47.362,
+      "grad_norm": 1.098421573638916,
+      "learning_rate": 2e-05,
+      "loss": 0.05161924,
+      "step": 23681
+    },
+    {
+      "epoch": 47.364,
+      "grad_norm": 1.4422335624694824,
+      "learning_rate": 2e-05,
+      "loss": 0.04385792,
+      "step": 23682
+    },
+    {
+      "epoch": 47.366,
+      "grad_norm": 1.126664638519287,
+      "learning_rate": 2e-05,
+      "loss": 0.04566282,
+      "step": 23683
+    },
+    {
+      "epoch": 47.368,
+      "grad_norm": 1.4247503280639648,
+      "learning_rate": 2e-05,
+      "loss": 0.05417196,
+      "step": 23684
+    },
+    {
+      "epoch": 47.37,
+      "grad_norm": 1.1334466934204102,
+      "learning_rate": 2e-05,
+      "loss": 0.04590701,
+      "step": 23685
+    },
+    {
+      "epoch": 47.372,
+      "grad_norm": 1.1766184568405151,
+      "learning_rate": 2e-05,
+      "loss": 0.04920424,
+      "step": 23686
+    },
+    {
+      "epoch": 47.374,
+      "grad_norm": 1.0475587844848633,
+      "learning_rate": 2e-05,
+      "loss": 0.03809017,
+      "step": 23687
+    },
+    {
+      "epoch": 47.376,
+      "grad_norm": 1.0179909467697144,
+      "learning_rate": 2e-05,
+      "loss": 0.0358481,
+      "step": 23688
+    },
+    {
+      "epoch": 47.378,
+      "grad_norm": 1.281684398651123,
+      "learning_rate": 2e-05,
+      "loss": 0.05288438,
+      "step": 23689
+    },
+    {
+      "epoch": 47.38,
+      "grad_norm": 1.3577895164489746,
+      "learning_rate": 2e-05,
+      "loss": 0.05755025,
+      "step": 23690
+    },
+    {
+      "epoch": 47.382,
+      "grad_norm": 1.284982681274414,
+      "learning_rate": 2e-05,
+      "loss": 0.0437559,
+      "step": 23691
+    },
+    {
+      "epoch": 47.384,
+      "grad_norm": 0.9902780055999756,
+      "learning_rate": 2e-05,
+      "loss": 0.03365954,
+      "step": 23692
+    },
+    {
+      "epoch": 47.386,
+      "grad_norm": 2.0288844108581543,
+      "learning_rate": 2e-05,
+      "loss": 0.03784881,
+      "step": 23693
+    },
+    {
+      "epoch": 47.388,
+      "grad_norm": 1.6296567916870117,
+      "learning_rate": 2e-05,
+      "loss": 0.05740002,
+      "step": 23694
+    },
+    {
+      "epoch": 47.39,
+      "grad_norm": 1.2057472467422485,
+      "learning_rate": 2e-05,
+      "loss": 0.04737182,
+      "step": 23695
+    },
+    {
+      "epoch": 47.392,
+      "grad_norm": 2.2552409172058105,
+      "learning_rate": 2e-05,
+      "loss": 0.05677157,
+      "step": 23696
+    },
+    {
+      "epoch": 47.394,
+      "grad_norm": 1.9773105382919312,
+      "learning_rate": 2e-05,
+      "loss": 0.05616665,
+      "step": 23697
+    },
+    {
+      "epoch": 47.396,
+      "grad_norm": 1.2296985387802124,
+      "learning_rate": 2e-05,
+      "loss": 0.04554373,
+      "step": 23698
+    },
+    {
+      "epoch": 47.398,
+      "grad_norm": 1.108909010887146,
+      "learning_rate": 2e-05,
+      "loss": 0.03959118,
+      "step": 23699
+    },
+    {
+      "epoch": 47.4,
+      "grad_norm": 1.2280199527740479,
+      "learning_rate": 2e-05,
+      "loss": 0.0505251,
+      "step": 23700
+    },
+    {
+      "epoch": 47.402,
+      "grad_norm": 1.2505056858062744,
+      "learning_rate": 2e-05,
+      "loss": 0.04818206,
+      "step": 23701
+    },
+    {
+      "epoch": 47.404,
+      "grad_norm": 1.2215907573699951,
+      "learning_rate": 2e-05,
+      "loss": 0.06155346,
+      "step": 23702
+    },
+    {
+      "epoch": 47.406,
+      "grad_norm": 2.548524856567383,
+      "learning_rate": 2e-05,
+      "loss": 0.07346775,
+      "step": 23703
+    },
+    {
+      "epoch": 47.408,
+      "grad_norm": 1.2835642099380493,
+      "learning_rate": 2e-05,
+      "loss": 0.03647726,
+      "step": 23704
+    },
+    {
+      "epoch": 47.41,
+      "grad_norm": 1.0395970344543457,
+      "learning_rate": 2e-05,
+      "loss": 0.03743298,
+      "step": 23705
+    },
+    {
+      "epoch": 47.412,
+      "grad_norm": 1.1787834167480469,
+      "learning_rate": 2e-05,
+      "loss": 0.05008036,
+      "step": 23706
+    },
+    {
+      "epoch": 47.414,
+      "grad_norm": 1.2608765363693237,
+      "learning_rate": 2e-05,
+      "loss": 0.05554523,
+      "step": 23707
+    },
+    {
+      "epoch": 47.416,
+      "grad_norm": 1.1391633749008179,
+      "learning_rate": 2e-05,
+      "loss": 0.04239395,
+      "step": 23708
+    },
+    {
+      "epoch": 47.418,
+      "grad_norm": 1.310951590538025,
+      "learning_rate": 2e-05,
+      "loss": 0.05426461,
+      "step": 23709
+    },
+    {
+      "epoch": 47.42,
+      "grad_norm": 2.4088597297668457,
+      "learning_rate": 2e-05,
+      "loss": 0.06128083,
+      "step": 23710
+    },
+    {
+      "epoch": 47.422,
+      "grad_norm": 1.4820548295974731,
+      "learning_rate": 2e-05,
+      "loss": 0.07166282,
+      "step": 23711
+    },
+    {
+      "epoch": 47.424,
+      "grad_norm": 1.2579067945480347,
+      "learning_rate": 2e-05,
+      "loss": 0.05245124,
+      "step": 23712
+    },
+    {
+      "epoch": 47.426,
+      "grad_norm": 1.2548208236694336,
+      "learning_rate": 2e-05,
+      "loss": 0.04808481,
+      "step": 23713
+    },
+    {
+      "epoch": 47.428,
+      "grad_norm": 1.0751184225082397,
+      "learning_rate": 2e-05,
+      "loss": 0.0375668,
+      "step": 23714
+    },
+    {
+      "epoch": 47.43,
+      "grad_norm": 1.1384459733963013,
+      "learning_rate": 2e-05,
+      "loss": 0.03911222,
+      "step": 23715
+    },
+    {
+      "epoch": 47.432,
+      "grad_norm": 1.2228471040725708,
+      "learning_rate": 2e-05,
+      "loss": 0.05863813,
+      "step": 23716
+    },
+    {
+      "epoch": 47.434,
+      "grad_norm": 3.312714099884033,
+      "learning_rate": 2e-05,
+      "loss": 0.06117887,
+      "step": 23717
+    },
+    {
+      "epoch": 47.436,
+      "grad_norm": 1.1511445045471191,
+      "learning_rate": 2e-05,
+      "loss": 0.04497742,
+      "step": 23718
+    },
+    {
+      "epoch": 47.438,
+      "grad_norm": 1.4445806741714478,
+      "learning_rate": 2e-05,
+      "loss": 0.06244376,
+      "step": 23719
+    },
+    {
+      "epoch": 47.44,
+      "grad_norm": 1.4664429426193237,
+      "learning_rate": 2e-05,
+      "loss": 0.03415911,
+      "step": 23720
+    },
+    {
+      "epoch": 47.442,
+      "grad_norm": 1.0343490839004517,
+      "learning_rate": 2e-05,
+      "loss": 0.03623316,
+      "step": 23721
+    },
+    {
+      "epoch": 47.444,
+      "grad_norm": 1.4916282892227173,
+      "learning_rate": 2e-05,
+      "loss": 0.04197778,
+      "step": 23722
+    },
+    {
+      "epoch": 47.446,
+      "grad_norm": 1.2317581176757812,
+      "learning_rate": 2e-05,
+      "loss": 0.05793425,
+      "step": 23723
+    },
+    {
+      "epoch": 47.448,
+      "grad_norm": 2.5570762157440186,
+      "learning_rate": 2e-05,
+      "loss": 0.04695795,
+      "step": 23724
+    },
+    {
+      "epoch": 47.45,
+      "grad_norm": 0.9492204785346985,
+      "learning_rate": 2e-05,
+      "loss": 0.0334493,
+      "step": 23725
+    },
+    {
+      "epoch": 47.452,
+      "grad_norm": 1.1067724227905273,
+      "learning_rate": 2e-05,
+      "loss": 0.03886347,
+      "step": 23726
+    },
+    {
+      "epoch": 47.454,
+      "grad_norm": 1.2168097496032715,
+      "learning_rate": 2e-05,
+      "loss": 0.04640615,
+      "step": 23727
+    },
+    {
+      "epoch": 47.456,
+      "grad_norm": 1.1498087644577026,
+      "learning_rate": 2e-05,
+      "loss": 0.05414462,
+      "step": 23728
+    },
+    {
+      "epoch": 47.458,
+      "grad_norm": 1.2377431392669678,
+      "learning_rate": 2e-05,
+      "loss": 0.05543441,
+      "step": 23729
+    },
+    {
+      "epoch": 47.46,
+      "grad_norm": 1.3042511940002441,
+      "learning_rate": 2e-05,
+      "loss": 0.05268206,
+      "step": 23730
+    },
+    {
+      "epoch": 47.462,
+      "grad_norm": 1.4321894645690918,
+      "learning_rate": 2e-05,
+      "loss": 0.04099277,
+      "step": 23731
+    },
+    {
+      "epoch": 47.464,
+      "grad_norm": 1.5580233335494995,
+      "learning_rate": 2e-05,
+      "loss": 0.05169383,
+      "step": 23732
+    },
+    {
+      "epoch": 47.466,
+      "grad_norm": 1.2211647033691406,
+      "learning_rate": 2e-05,
+      "loss": 0.03869744,
+      "step": 23733
+    },
+    {
+      "epoch": 47.468,
+      "grad_norm": 1.683929681777954,
+      "learning_rate": 2e-05,
+      "loss": 0.06134295,
+      "step": 23734
+    },
+    {
+      "epoch": 47.47,
+      "grad_norm": 1.2739115953445435,
+      "learning_rate": 2e-05,
+      "loss": 0.0545798,
+      "step": 23735
+    },
+    {
+      "epoch": 47.472,
+      "grad_norm": 1.387938141822815,
+      "learning_rate": 2e-05,
+      "loss": 0.06269515,
+      "step": 23736
+    },
+    {
+      "epoch": 47.474,
+      "grad_norm": 1.0513253211975098,
+      "learning_rate": 2e-05,
+      "loss": 0.03459634,
+      "step": 23737
+    },
+    {
+      "epoch": 47.476,
+      "grad_norm": 1.5735803842544556,
+      "learning_rate": 2e-05,
+      "loss": 0.05392249,
+      "step": 23738
+    },
+    {
+      "epoch": 47.478,
+      "grad_norm": 1.1136903762817383,
+      "learning_rate": 2e-05,
+      "loss": 0.0393229,
+      "step": 23739
+    },
+    {
+      "epoch": 47.48,
+      "grad_norm": 1.2298059463500977,
+      "learning_rate": 2e-05,
+      "loss": 0.05348741,
+      "step": 23740
+    },
+    {
+      "epoch": 47.482,
+      "grad_norm": 1.1526238918304443,
+      "learning_rate": 2e-05,
+      "loss": 0.04516281,
+      "step": 23741
+    },
+    {
+      "epoch": 47.484,
+      "grad_norm": 1.8501273393630981,
+      "learning_rate": 2e-05,
+      "loss": 0.04677322,
+      "step": 23742
+    },
+    {
+      "epoch": 47.486,
+      "grad_norm": 1.4598199129104614,
+      "learning_rate": 2e-05,
+      "loss": 0.03834684,
+      "step": 23743
+    },
+    {
+      "epoch": 47.488,
+      "grad_norm": 1.539868950843811,
+      "learning_rate": 2e-05,
+      "loss": 0.04157211,
+      "step": 23744
+    },
+    {
+      "epoch": 47.49,
+      "grad_norm": 1.016363501548767,
+      "learning_rate": 2e-05,
+      "loss": 0.0401577,
+      "step": 23745
+    },
+    {
+      "epoch": 47.492,
+      "grad_norm": 1.0004823207855225,
+      "learning_rate": 2e-05,
+      "loss": 0.04216732,
+      "step": 23746
+    },
+    {
+      "epoch": 47.494,
+      "grad_norm": 0.8949176669120789,
+      "learning_rate": 2e-05,
+      "loss": 0.03267929,
+      "step": 23747
+    },
+    {
+      "epoch": 47.496,
+      "grad_norm": 1.092294692993164,
+      "learning_rate": 2e-05,
+      "loss": 0.04605146,
+      "step": 23748
+    },
+    {
+      "epoch": 47.498,
+      "grad_norm": 1.3999691009521484,
+      "learning_rate": 2e-05,
+      "loss": 0.05625776,
+      "step": 23749
+    },
+    {
+      "epoch": 47.5,
+      "grad_norm": 1.136635422706604,
+      "learning_rate": 2e-05,
+      "loss": 0.04005692,
+      "step": 23750
+    },
+    {
+      "epoch": 47.502,
+      "grad_norm": 1.1078810691833496,
+      "learning_rate": 2e-05,
+      "loss": 0.04342239,
+      "step": 23751
+    },
+    {
+      "epoch": 47.504,
+      "grad_norm": 1.279944658279419,
+      "learning_rate": 2e-05,
+      "loss": 0.04219215,
+      "step": 23752
+    },
+    {
+      "epoch": 47.506,
+      "grad_norm": 1.091110110282898,
+      "learning_rate": 2e-05,
+      "loss": 0.03880176,
+      "step": 23753
+    },
+    {
+      "epoch": 47.508,
+      "grad_norm": 4.729748249053955,
+      "learning_rate": 2e-05,
+      "loss": 0.05638552,
+      "step": 23754
+    },
+    {
+      "epoch": 47.51,
+      "grad_norm": 1.0970360040664673,
+      "learning_rate": 2e-05,
+      "loss": 0.04364794,
+      "step": 23755
+    },
+    {
+      "epoch": 47.512,
+      "grad_norm": 1.0389903783798218,
+      "learning_rate": 2e-05,
+      "loss": 0.04169441,
+      "step": 23756
+    },
+    {
+      "epoch": 47.514,
+      "grad_norm": 1.5221785306930542,
+      "learning_rate": 2e-05,
+      "loss": 0.06585445,
+      "step": 23757
+    },
+    {
+      "epoch": 47.516,
+      "grad_norm": 2.1099374294281006,
+      "learning_rate": 2e-05,
+      "loss": 0.06468124,
+      "step": 23758
+    },
+    {
+      "epoch": 47.518,
+      "grad_norm": 1.0862318277359009,
+      "learning_rate": 2e-05,
+      "loss": 0.03993117,
+      "step": 23759
+    },
+    {
+      "epoch": 47.52,
+      "grad_norm": 1.1114388704299927,
+      "learning_rate": 2e-05,
+      "loss": 0.04338128,
+      "step": 23760
+    },
+    {
+      "epoch": 47.522,
+      "grad_norm": 0.9857814908027649,
+      "learning_rate": 2e-05,
+      "loss": 0.03844403,
+      "step": 23761
+    },
+    {
+      "epoch": 47.524,
+      "grad_norm": 1.3540973663330078,
+      "learning_rate": 2e-05,
+      "loss": 0.04602545,
+      "step": 23762
+    },
+    {
+      "epoch": 47.526,
+      "grad_norm": 1.3859474658966064,
+      "learning_rate": 2e-05,
+      "loss": 0.06367904,
+      "step": 23763
+    },
+    {
+      "epoch": 47.528,
+      "grad_norm": 1.4162659645080566,
+      "learning_rate": 2e-05,
+      "loss": 0.05949718,
+      "step": 23764
+    },
+    {
+      "epoch": 47.53,
+      "grad_norm": 1.3201795816421509,
+      "learning_rate": 2e-05,
+      "loss": 0.06164296,
+      "step": 23765
+    },
+    {
+      "epoch": 47.532,
+      "grad_norm": 1.2807005643844604,
+      "learning_rate": 2e-05,
+      "loss": 0.05345618,
+      "step": 23766
+    },
+    {
+      "epoch": 47.534,
+      "grad_norm": 1.0871586799621582,
+      "learning_rate": 2e-05,
+      "loss": 0.04408365,
+      "step": 23767
+    },
+    {
+      "epoch": 47.536,
+      "grad_norm": 1.0492790937423706,
+      "learning_rate": 2e-05,
+      "loss": 0.04874376,
+      "step": 23768
+    },
+    {
+      "epoch": 47.538,
+      "grad_norm": 1.1767281293869019,
+      "learning_rate": 2e-05,
+      "loss": 0.04965818,
+      "step": 23769
+    },
+    {
+      "epoch": 47.54,
+      "grad_norm": 1.1047857999801636,
+      "learning_rate": 2e-05,
+      "loss": 0.05242363,
+      "step": 23770
+    },
+    {
+      "epoch": 47.542,
+      "grad_norm": 1.1232954263687134,
+      "learning_rate": 2e-05,
+      "loss": 0.05646368,
+      "step": 23771
+    },
+    {
+      "epoch": 47.544,
+      "grad_norm": 1.3574858903884888,
+      "learning_rate": 2e-05,
+      "loss": 0.04148731,
+      "step": 23772
+    },
+    {
+      "epoch": 47.546,
+      "grad_norm": 0.9722660183906555,
+      "learning_rate": 2e-05,
+      "loss": 0.0293147,
+      "step": 23773
+    },
+    {
+      "epoch": 47.548,
+      "grad_norm": 1.1585569381713867,
+      "learning_rate": 2e-05,
+      "loss": 0.03967403,
+      "step": 23774
+    },
+    {
+      "epoch": 47.55,
+      "grad_norm": 1.473108172416687,
+      "learning_rate": 2e-05,
+      "loss": 0.06600234,
+      "step": 23775
+    },
+    {
+      "epoch": 47.552,
+      "grad_norm": 1.2558122873306274,
+      "learning_rate": 2e-05,
+      "loss": 0.04727027,
+      "step": 23776
+    },
+    {
+      "epoch": 47.554,
+      "grad_norm": 1.1267999410629272,
+      "learning_rate": 2e-05,
+      "loss": 0.05461453,
+      "step": 23777
+    },
+    {
+      "epoch": 47.556,
+      "grad_norm": 1.1623164415359497,
+      "learning_rate": 2e-05,
+      "loss": 0.04348694,
+      "step": 23778
+    },
+    {
+      "epoch": 47.558,
+      "grad_norm": 0.9064064621925354,
+      "learning_rate": 2e-05,
+      "loss": 0.03024364,
+      "step": 23779
+    },
+    {
+      "epoch": 47.56,
+      "grad_norm": 1.4345759153366089,
+      "learning_rate": 2e-05,
+      "loss": 0.04869916,
+      "step": 23780
+    },
+    {
+      "epoch": 47.562,
+      "grad_norm": 1.3848798274993896,
+      "learning_rate": 2e-05,
+      "loss": 0.06412567,
+      "step": 23781
+    },
+    {
+      "epoch": 47.564,
+      "grad_norm": 1.2268092632293701,
+      "learning_rate": 2e-05,
+      "loss": 0.0550194,
+      "step": 23782
+    },
+    {
+      "epoch": 47.566,
+      "grad_norm": 1.1246010065078735,
+      "learning_rate": 2e-05,
+      "loss": 0.05901127,
+      "step": 23783
+    },
+    {
+      "epoch": 47.568,
+      "grad_norm": 1.1794065237045288,
+      "learning_rate": 2e-05,
+      "loss": 0.04719983,
+      "step": 23784
+    },
+    {
+      "epoch": 47.57,
+      "grad_norm": 1.2328189611434937,
+      "learning_rate": 2e-05,
+      "loss": 0.03939323,
+      "step": 23785
+    },
+    {
+      "epoch": 47.572,
+      "grad_norm": 1.12712824344635,
+      "learning_rate": 2e-05,
+      "loss": 0.04430801,
+      "step": 23786
+    },
+    {
+      "epoch": 47.574,
+      "grad_norm": 1.1956403255462646,
+      "learning_rate": 2e-05,
+      "loss": 0.05049133,
+      "step": 23787
+    },
+    {
+      "epoch": 47.576,
+      "grad_norm": 1.102604866027832,
+      "learning_rate": 2e-05,
+      "loss": 0.04754536,
+      "step": 23788
+    },
+    {
+      "epoch": 47.578,
+      "grad_norm": 1.1247801780700684,
+      "learning_rate": 2e-05,
+      "loss": 0.04530304,
+      "step": 23789
+    },
+    {
+      "epoch": 47.58,
+      "grad_norm": 1.1731112003326416,
+      "learning_rate": 2e-05,
+      "loss": 0.04685323,
+      "step": 23790
+    },
+    {
+      "epoch": 47.582,
+      "grad_norm": 1.0738763809204102,
+      "learning_rate": 2e-05,
+      "loss": 0.0424612,
+      "step": 23791
+    },
+    {
+      "epoch": 47.584,
+      "grad_norm": 0.9249600172042847,
+      "learning_rate": 2e-05,
+      "loss": 0.03432335,
+      "step": 23792
+    },
+    {
+      "epoch": 47.586,
+      "grad_norm": 0.9936639666557312,
+      "learning_rate": 2e-05,
+      "loss": 0.03993072,
+      "step": 23793
+    },
+    {
+      "epoch": 47.588,
+      "grad_norm": 1.2670681476593018,
+      "learning_rate": 2e-05,
+      "loss": 0.04661375,
+      "step": 23794
+    },
+    {
+      "epoch": 47.59,
+      "grad_norm": 1.8145248889923096,
+      "learning_rate": 2e-05,
+      "loss": 0.05395586,
+      "step": 23795
+    },
+    {
+      "epoch": 47.592,
+      "grad_norm": 1.2371265888214111,
+      "learning_rate": 2e-05,
+      "loss": 0.06457849,
+      "step": 23796
+    },
+    {
+      "epoch": 47.594,
+      "grad_norm": 1.2411563396453857,
+      "learning_rate": 2e-05,
+      "loss": 0.04113891,
+      "step": 23797
+    },
+    {
+      "epoch": 47.596,
+      "grad_norm": 1.3027358055114746,
+      "learning_rate": 2e-05,
+      "loss": 0.04472353,
+      "step": 23798
+    },
+    {
+      "epoch": 47.598,
+      "grad_norm": 1.9770257472991943,
+      "learning_rate": 2e-05,
+      "loss": 0.06172611,
+      "step": 23799
+    },
+    {
+      "epoch": 47.6,
+      "grad_norm": 1.2722480297088623,
+      "learning_rate": 2e-05,
+      "loss": 0.04975735,
+      "step": 23800
+    },
+    {
+      "epoch": 47.602,
+      "grad_norm": 0.9771064519882202,
+      "learning_rate": 2e-05,
+      "loss": 0.03003059,
+      "step": 23801
+    },
+    {
+      "epoch": 47.604,
+      "grad_norm": 1.6549073457717896,
+      "learning_rate": 2e-05,
+      "loss": 0.04169197,
+      "step": 23802
+    },
+    {
+      "epoch": 47.606,
+      "grad_norm": 0.9755164980888367,
+      "learning_rate": 2e-05,
+      "loss": 0.03423417,
+      "step": 23803
+    },
+    {
+      "epoch": 47.608,
+      "grad_norm": 1.199762225151062,
+      "learning_rate": 2e-05,
+      "loss": 0.05413834,
+      "step": 23804
+    },
+    {
+      "epoch": 47.61,
+      "grad_norm": 1.410494089126587,
+      "learning_rate": 2e-05,
+      "loss": 0.04707893,
+      "step": 23805
+    },
+    {
+      "epoch": 47.612,
+      "grad_norm": 1.3130079507827759,
+      "learning_rate": 2e-05,
+      "loss": 0.05722672,
+      "step": 23806
+    },
+    {
+      "epoch": 47.614,
+      "grad_norm": 1.174035668373108,
+      "learning_rate": 2e-05,
+      "loss": 0.05756243,
+      "step": 23807
+    },
+    {
+      "epoch": 47.616,
+      "grad_norm": 1.7935864925384521,
+      "learning_rate": 2e-05,
+      "loss": 0.07460059,
+      "step": 23808
+    },
+    {
+      "epoch": 47.618,
+      "grad_norm": 1.6251633167266846,
+      "learning_rate": 2e-05,
+      "loss": 0.03863,
+      "step": 23809
+    },
+    {
+      "epoch": 47.62,
+      "grad_norm": 1.1181731224060059,
+      "learning_rate": 2e-05,
+      "loss": 0.04170064,
+      "step": 23810
+    },
+    {
+      "epoch": 47.622,
+      "grad_norm": 1.3144530057907104,
+      "learning_rate": 2e-05,
+      "loss": 0.05499358,
+      "step": 23811
+    },
+    {
+      "epoch": 47.624,
+      "grad_norm": 1.4863946437835693,
+      "learning_rate": 2e-05,
+      "loss": 0.07462236,
+      "step": 23812
+    },
+    {
+      "epoch": 47.626,
+      "grad_norm": 1.226033329963684,
+      "learning_rate": 2e-05,
+      "loss": 0.05091349,
+      "step": 23813
+    },
+    {
+      "epoch": 47.628,
+      "grad_norm": 1.164097785949707,
+      "learning_rate": 2e-05,
+      "loss": 0.0417085,
+      "step": 23814
+    },
+    {
+      "epoch": 47.63,
+      "grad_norm": 3.6329402923583984,
+      "learning_rate": 2e-05,
+      "loss": 0.03966795,
+      "step": 23815
+    },
+    {
+      "epoch": 47.632,
+      "grad_norm": 1.1707428693771362,
+      "learning_rate": 2e-05,
+      "loss": 0.05277382,
+      "step": 23816
+    },
+    {
+      "epoch": 47.634,
+      "grad_norm": 0.8920641541481018,
+      "learning_rate": 2e-05,
+      "loss": 0.02758377,
+      "step": 23817
+    },
+    {
+      "epoch": 47.636,
+      "grad_norm": 1.1014950275421143,
+      "learning_rate": 2e-05,
+      "loss": 0.04578391,
+      "step": 23818
+    },
+    {
+      "epoch": 47.638,
+      "grad_norm": 1.0088590383529663,
+      "learning_rate": 2e-05,
+      "loss": 0.0345577,
+      "step": 23819
+    },
+    {
+      "epoch": 47.64,
+      "grad_norm": 1.6413283348083496,
+      "learning_rate": 2e-05,
+      "loss": 0.04818995,
+      "step": 23820
+    },
+    {
+      "epoch": 47.642,
+      "grad_norm": 1.2504075765609741,
+      "learning_rate": 2e-05,
+      "loss": 0.04787406,
+      "step": 23821
+    },
+    {
+      "epoch": 47.644,
+      "grad_norm": 2.165092945098877,
+      "learning_rate": 2e-05,
+      "loss": 0.03882634,
+      "step": 23822
+    },
+    {
+      "epoch": 47.646,
+      "grad_norm": 0.9500136971473694,
+      "learning_rate": 2e-05,
+      "loss": 0.04030318,
+      "step": 23823
+    },
+    {
+      "epoch": 47.648,
+      "grad_norm": 1.0927907228469849,
+      "learning_rate": 2e-05,
+      "loss": 0.03360862,
+      "step": 23824
+    },
+    {
+      "epoch": 47.65,
+      "grad_norm": 1.44313645362854,
+      "learning_rate": 2e-05,
+      "loss": 0.05937745,
+      "step": 23825
+    },
+    {
+      "epoch": 47.652,
+      "grad_norm": 1.2534681558609009,
+      "learning_rate": 2e-05,
+      "loss": 0.04984522,
+      "step": 23826
+    },
+    {
+      "epoch": 47.654,
+      "grad_norm": 2.9447219371795654,
+      "learning_rate": 2e-05,
+      "loss": 0.05307635,
+      "step": 23827
+    },
+    {
+      "epoch": 47.656,
+      "grad_norm": 1.3493146896362305,
+      "learning_rate": 2e-05,
+      "loss": 0.04327986,
+      "step": 23828
+    },
+    {
+      "epoch": 47.658,
+      "grad_norm": 1.0709292888641357,
+      "learning_rate": 2e-05,
+      "loss": 0.04765247,
+      "step": 23829
+    },
+    {
+      "epoch": 47.66,
+      "grad_norm": 1.281948447227478,
+      "learning_rate": 2e-05,
+      "loss": 0.05185589,
+      "step": 23830
+    },
+    {
+      "epoch": 47.662,
+      "grad_norm": 1.2605935335159302,
+      "learning_rate": 2e-05,
+      "loss": 0.05030036,
+      "step": 23831
+    },
+    {
+      "epoch": 47.664,
+      "grad_norm": 1.413135051727295,
+      "learning_rate": 2e-05,
+      "loss": 0.05803326,
+      "step": 23832
+    },
+    {
+      "epoch": 47.666,
+      "grad_norm": 1.2617324590682983,
+      "learning_rate": 2e-05,
+      "loss": 0.04378517,
+      "step": 23833
+    },
+    {
+      "epoch": 47.668,
+      "grad_norm": 1.1886942386627197,
+      "learning_rate": 2e-05,
+      "loss": 0.05778721,
+      "step": 23834
+    },
+    {
+      "epoch": 47.67,
+      "grad_norm": 3.0332577228546143,
+      "learning_rate": 2e-05,
+      "loss": 0.05872624,
+      "step": 23835
+    },
+    {
+      "epoch": 47.672,
+      "grad_norm": 1.4225571155548096,
+      "learning_rate": 2e-05,
+      "loss": 0.03956306,
+      "step": 23836
+    },
+    {
+      "epoch": 47.674,
+      "grad_norm": 0.9652089476585388,
+      "learning_rate": 2e-05,
+      "loss": 0.02727574,
+      "step": 23837
+    },
+    {
+      "epoch": 47.676,
+      "grad_norm": 1.103406548500061,
+      "learning_rate": 2e-05,
+      "loss": 0.04494239,
+      "step": 23838
+    },
+    {
+      "epoch": 47.678,
+      "grad_norm": 1.1904319524765015,
+      "learning_rate": 2e-05,
+      "loss": 0.04636001,
+      "step": 23839
+    },
+    {
+      "epoch": 47.68,
+      "grad_norm": 1.2117693424224854,
+      "learning_rate": 2e-05,
+      "loss": 0.04273736,
+      "step": 23840
+    },
+    {
+      "epoch": 47.682,
+      "grad_norm": 1.095442771911621,
+      "learning_rate": 2e-05,
+      "loss": 0.04069439,
+      "step": 23841
+    },
+    {
+      "epoch": 47.684,
+      "grad_norm": 1.315514326095581,
+      "learning_rate": 2e-05,
+      "loss": 0.05416162,
+      "step": 23842
+    },
+    {
+      "epoch": 47.686,
+      "grad_norm": 1.8992160558700562,
+      "learning_rate": 2e-05,
+      "loss": 0.07332712,
+      "step": 23843
+    },
+    {
+      "epoch": 47.688,
+      "grad_norm": 1.0993719100952148,
+      "learning_rate": 2e-05,
+      "loss": 0.04374916,
+      "step": 23844
+    },
+    {
+      "epoch": 47.69,
+      "grad_norm": 1.2032912969589233,
+      "learning_rate": 2e-05,
+      "loss": 0.04704859,
+      "step": 23845
+    },
+    {
+      "epoch": 47.692,
+      "grad_norm": 5.187289237976074,
+      "learning_rate": 2e-05,
+      "loss": 0.04735934,
+      "step": 23846
+    },
+    {
+      "epoch": 47.694,
+      "grad_norm": 1.0638306140899658,
+      "learning_rate": 2e-05,
+      "loss": 0.04312051,
+      "step": 23847
+    },
+    {
+      "epoch": 47.696,
+      "grad_norm": 1.0537846088409424,
+      "learning_rate": 2e-05,
+      "loss": 0.04088642,
+      "step": 23848
+    },
+    {
+      "epoch": 47.698,
+      "grad_norm": 1.1591740846633911,
+      "learning_rate": 2e-05,
+      "loss": 0.04216747,
+      "step": 23849
+    },
+    {
+      "epoch": 47.7,
+      "grad_norm": 0.8905053734779358,
+      "learning_rate": 2e-05,
+      "loss": 0.03059562,
+      "step": 23850
+    },
+    {
+      "epoch": 47.702,
+      "grad_norm": 0.9576981663703918,
+      "learning_rate": 2e-05,
+      "loss": 0.02699701,
+      "step": 23851
+    },
+    {
+      "epoch": 47.704,
+      "grad_norm": 1.3115862607955933,
+      "learning_rate": 2e-05,
+      "loss": 0.04908556,
+      "step": 23852
+    },
+    {
+      "epoch": 47.706,
+      "grad_norm": 1.1104012727737427,
+      "learning_rate": 2e-05,
+      "loss": 0.04494857,
+      "step": 23853
+    },
+    {
+      "epoch": 47.708,
+      "grad_norm": 1.0533535480499268,
+      "learning_rate": 2e-05,
+      "loss": 0.03632095,
+      "step": 23854
+    },
+    {
+      "epoch": 47.71,
+      "grad_norm": 1.0708428621292114,
+      "learning_rate": 2e-05,
+      "loss": 0.04081979,
+      "step": 23855
+    },
+    {
+      "epoch": 47.712,
+      "grad_norm": 1.3414500951766968,
+      "learning_rate": 2e-05,
+      "loss": 0.04155166,
+      "step": 23856
+    },
+    {
+      "epoch": 47.714,
+      "grad_norm": 1.131571888923645,
+      "learning_rate": 2e-05,
+      "loss": 0.03478384,
+      "step": 23857
+    },
+    {
+      "epoch": 47.716,
+      "grad_norm": 1.0607177019119263,
+      "learning_rate": 2e-05,
+      "loss": 0.04390678,
+      "step": 23858
+    },
+    {
+      "epoch": 47.718,
+      "grad_norm": 1.1620864868164062,
+      "learning_rate": 2e-05,
+      "loss": 0.03903388,
+      "step": 23859
+    },
+    {
+      "epoch": 47.72,
+      "grad_norm": 1.2598408460617065,
+      "learning_rate": 2e-05,
+      "loss": 0.04365044,
+      "step": 23860
+    },
+    {
+      "epoch": 47.722,
+      "grad_norm": 1.0379410982131958,
+      "learning_rate": 2e-05,
+      "loss": 0.03826845,
+      "step": 23861
+    },
+    {
+      "epoch": 47.724,
+      "grad_norm": 1.4189554452896118,
+      "learning_rate": 2e-05,
+      "loss": 0.05523844,
+      "step": 23862
+    },
+    {
+      "epoch": 47.726,
+      "grad_norm": 2.2879295349121094,
+      "learning_rate": 2e-05,
+      "loss": 0.04601286,
+      "step": 23863
+    },
+    {
+      "epoch": 47.728,
+      "grad_norm": 1.9662691354751587,
+      "learning_rate": 2e-05,
+      "loss": 0.04415329,
+      "step": 23864
+    },
+    {
+      "epoch": 47.73,
+      "grad_norm": 1.136714220046997,
+      "learning_rate": 2e-05,
+      "loss": 0.04704106,
+      "step": 23865
+    },
+    {
+      "epoch": 47.732,
+      "grad_norm": 1.8158776760101318,
+      "learning_rate": 2e-05,
+      "loss": 0.04839449,
+      "step": 23866
+    },
+    {
+      "epoch": 47.734,
+      "grad_norm": 1.312970519065857,
+      "learning_rate": 2e-05,
+      "loss": 0.05811097,
+      "step": 23867
+    },
+    {
+      "epoch": 47.736,
+      "grad_norm": 1.4508172273635864,
+      "learning_rate": 2e-05,
+      "loss": 0.06293134,
+      "step": 23868
+    },
+    {
+      "epoch": 47.738,
+      "grad_norm": 1.0996098518371582,
+      "learning_rate": 2e-05,
+      "loss": 0.05198246,
+      "step": 23869
+    },
+    {
+      "epoch": 47.74,
+      "grad_norm": 1.2088953256607056,
+      "learning_rate": 2e-05,
+      "loss": 0.05847076,
+      "step": 23870
+    },
+    {
+      "epoch": 47.742,
+      "grad_norm": 1.990601897239685,
+      "learning_rate": 2e-05,
+      "loss": 0.07238883,
+      "step": 23871
+    },
+    {
+      "epoch": 47.744,
+      "grad_norm": 0.990237832069397,
+      "learning_rate": 2e-05,
+      "loss": 0.03466114,
+      "step": 23872
+    },
+    {
+      "epoch": 47.746,
+      "grad_norm": 1.187274694442749,
+      "learning_rate": 2e-05,
+      "loss": 0.05454275,
+      "step": 23873
+    },
+    {
+      "epoch": 47.748,
+      "grad_norm": 1.2135357856750488,
+      "learning_rate": 2e-05,
+      "loss": 0.05944178,
+      "step": 23874
+    },
+    {
+      "epoch": 47.75,
+      "grad_norm": 1.147931694984436,
+      "learning_rate": 2e-05,
+      "loss": 0.04608715,
+      "step": 23875
+    },
+    {
+      "epoch": 47.752,
+      "grad_norm": 1.2686227560043335,
+      "learning_rate": 2e-05,
+      "loss": 0.03793816,
+      "step": 23876
+    },
+    {
+      "epoch": 47.754,
+      "grad_norm": 1.2888165712356567,
+      "learning_rate": 2e-05,
+      "loss": 0.05994112,
+      "step": 23877
+    },
+    {
+      "epoch": 47.756,
+      "grad_norm": 1.1328716278076172,
+      "learning_rate": 2e-05,
+      "loss": 0.04631865,
+      "step": 23878
+    },
+    {
+      "epoch": 47.758,
+      "grad_norm": 1.1973130702972412,
+      "learning_rate": 2e-05,
+      "loss": 0.04879654,
+      "step": 23879
+    },
+    {
+      "epoch": 47.76,
+      "grad_norm": 1.3249599933624268,
+      "learning_rate": 2e-05,
+      "loss": 0.0531996,
+      "step": 23880
+    },
+    {
+      "epoch": 47.762,
+      "grad_norm": 0.9521530866622925,
+      "learning_rate": 2e-05,
+      "loss": 0.03609097,
+      "step": 23881
+    },
+    {
+      "epoch": 47.764,
+      "grad_norm": 1.0063776969909668,
+      "learning_rate": 2e-05,
+      "loss": 0.04234225,
+      "step": 23882
+    },
+    {
+      "epoch": 47.766,
+      "grad_norm": 1.1210484504699707,
+      "learning_rate": 2e-05,
+      "loss": 0.04907482,
+      "step": 23883
+    },
+    {
+      "epoch": 47.768,
+      "grad_norm": 1.406660795211792,
+      "learning_rate": 2e-05,
+      "loss": 0.05606017,
+      "step": 23884
+    },
+    {
+      "epoch": 47.77,
+      "grad_norm": 1.1773229837417603,
+      "learning_rate": 2e-05,
+      "loss": 0.05414952,
+      "step": 23885
+    },
+    {
+      "epoch": 47.772,
+      "grad_norm": 2.4284679889678955,
+      "learning_rate": 2e-05,
+      "loss": 0.04393204,
+      "step": 23886
+    },
+    {
+      "epoch": 47.774,
+      "grad_norm": 1.0576744079589844,
+      "learning_rate": 2e-05,
+      "loss": 0.03376132,
+      "step": 23887
+    },
+    {
+      "epoch": 47.776,
+      "grad_norm": 2.838202714920044,
+      "learning_rate": 2e-05,
+      "loss": 0.06541966,
+      "step": 23888
+    },
+    {
+      "epoch": 47.778,
+      "grad_norm": 1.1831326484680176,
+      "learning_rate": 2e-05,
+      "loss": 0.04526076,
+      "step": 23889
+    },
+    {
+      "epoch": 47.78,
+      "grad_norm": 1.4545902013778687,
+      "learning_rate": 2e-05,
+      "loss": 0.05634362,
+      "step": 23890
+    },
+    {
+      "epoch": 47.782,
+      "grad_norm": 1.2465368509292603,
+      "learning_rate": 2e-05,
+      "loss": 0.04193708,
+      "step": 23891
+    },
+    {
+      "epoch": 47.784,
+      "grad_norm": 1.2062345743179321,
+      "learning_rate": 2e-05,
+      "loss": 0.05329098,
+      "step": 23892
+    },
+    {
+      "epoch": 47.786,
+      "grad_norm": 1.221840262413025,
+      "learning_rate": 2e-05,
+      "loss": 0.04324333,
+      "step": 23893
+    },
+    {
+      "epoch": 47.788,
+      "grad_norm": 1.1237281560897827,
+      "learning_rate": 2e-05,
+      "loss": 0.03751169,
+      "step": 23894
+    },
+    {
+      "epoch": 47.79,
+      "grad_norm": 1.1282539367675781,
+      "learning_rate": 2e-05,
+      "loss": 0.04889031,
+      "step": 23895
+    },
+    {
+      "epoch": 47.792,
+      "grad_norm": 1.0426477193832397,
+      "learning_rate": 2e-05,
+      "loss": 0.03937999,
+      "step": 23896
+    },
+    {
+      "epoch": 47.794,
+      "grad_norm": 2.0234551429748535,
+      "learning_rate": 2e-05,
+      "loss": 0.04139862,
+      "step": 23897
+    },
+    {
+      "epoch": 47.796,
+      "grad_norm": 1.0973756313323975,
+      "learning_rate": 2e-05,
+      "loss": 0.04653414,
+      "step": 23898
+    },
+    {
+      "epoch": 47.798,
+      "grad_norm": 1.1589323282241821,
+      "learning_rate": 2e-05,
+      "loss": 0.05205517,
+      "step": 23899
+    },
+    {
+      "epoch": 47.8,
+      "grad_norm": 1.1792693138122559,
+      "learning_rate": 2e-05,
+      "loss": 0.04807966,
+      "step": 23900
+    },
+    {
+      "epoch": 47.802,
+      "grad_norm": 0.9611416459083557,
+      "learning_rate": 2e-05,
+      "loss": 0.03486631,
+      "step": 23901
+    },
+    {
+      "epoch": 47.804,
+      "grad_norm": 1.1984083652496338,
+      "learning_rate": 2e-05,
+      "loss": 0.06004932,
+      "step": 23902
+    },
+    {
+      "epoch": 47.806,
+      "grad_norm": 1.102168083190918,
+      "learning_rate": 2e-05,
+      "loss": 0.05081422,
+      "step": 23903
+    },
+    {
+      "epoch": 47.808,
+      "grad_norm": 1.1081064939498901,
+      "learning_rate": 2e-05,
+      "loss": 0.0446634,
+      "step": 23904
+    },
+    {
+      "epoch": 47.81,
+      "grad_norm": 0.9922335147857666,
+      "learning_rate": 2e-05,
+      "loss": 0.04278713,
+      "step": 23905
+    },
+    {
+      "epoch": 47.812,
+      "grad_norm": 1.3645964860916138,
+      "learning_rate": 2e-05,
+      "loss": 0.06499512,
+      "step": 23906
+    },
+    {
+      "epoch": 47.814,
+      "grad_norm": 1.6501846313476562,
+      "learning_rate": 2e-05,
+      "loss": 0.04862681,
+      "step": 23907
+    },
+    {
+      "epoch": 47.816,
+      "grad_norm": 1.0368905067443848,
+      "learning_rate": 2e-05,
+      "loss": 0.0441791,
+      "step": 23908
+    },
+    {
+      "epoch": 47.818,
+      "grad_norm": 1.1054444313049316,
+      "learning_rate": 2e-05,
+      "loss": 0.04063675,
+      "step": 23909
+    },
+    {
+      "epoch": 47.82,
+      "grad_norm": 1.1246895790100098,
+      "learning_rate": 2e-05,
+      "loss": 0.04597709,
+      "step": 23910
+    },
+    {
+      "epoch": 47.822,
+      "grad_norm": 1.0721478462219238,
+      "learning_rate": 2e-05,
+      "loss": 0.04677291,
+      "step": 23911
+    },
+    {
+      "epoch": 47.824,
+      "grad_norm": 1.1274585723876953,
+      "learning_rate": 2e-05,
+      "loss": 0.05058762,
+      "step": 23912
+    },
+    {
+      "epoch": 47.826,
+      "grad_norm": 1.4998968839645386,
+      "learning_rate": 2e-05,
+      "loss": 0.05139451,
+      "step": 23913
+    },
+    {
+      "epoch": 47.828,
+      "grad_norm": 0.9521937370300293,
+      "learning_rate": 2e-05,
+      "loss": 0.03633155,
+      "step": 23914
+    },
+    {
+      "epoch": 47.83,
+      "grad_norm": 1.3443994522094727,
+      "learning_rate": 2e-05,
+      "loss": 0.06229488,
+      "step": 23915
+    },
+    {
+      "epoch": 47.832,
+      "grad_norm": 1.165913701057434,
+      "learning_rate": 2e-05,
+      "loss": 0.05210435,
+      "step": 23916
+    },
+    {
+      "epoch": 47.834,
+      "grad_norm": 1.4400709867477417,
+      "learning_rate": 2e-05,
+      "loss": 0.0268791,
+      "step": 23917
+    },
+    {
+      "epoch": 47.836,
+      "grad_norm": 0.978833019733429,
+      "learning_rate": 2e-05,
+      "loss": 0.03885253,
+      "step": 23918
+    },
+    {
+      "epoch": 47.838,
+      "grad_norm": 3.3101084232330322,
+      "learning_rate": 2e-05,
+      "loss": 0.06251098,
+      "step": 23919
+    },
+    {
+      "epoch": 47.84,
+      "grad_norm": 1.6580345630645752,
+      "learning_rate": 2e-05,
+      "loss": 0.04920085,
+      "step": 23920
+    },
+    {
+      "epoch": 47.842,
+      "grad_norm": 1.2298296689987183,
+      "learning_rate": 2e-05,
+      "loss": 0.04933652,
+      "step": 23921
+    },
+    {
+      "epoch": 47.844,
+      "grad_norm": 1.209776520729065,
+      "learning_rate": 2e-05,
+      "loss": 0.06445093,
+      "step": 23922
+    },
+    {
+      "epoch": 47.846,
+      "grad_norm": 1.1187968254089355,
+      "learning_rate": 2e-05,
+      "loss": 0.04598414,
+      "step": 23923
+    },
+    {
+      "epoch": 47.848,
+      "grad_norm": 1.0532989501953125,
+      "learning_rate": 2e-05,
+      "loss": 0.03271957,
+      "step": 23924
+    },
+    {
+      "epoch": 47.85,
+      "grad_norm": 1.2785584926605225,
+      "learning_rate": 2e-05,
+      "loss": 0.05796767,
+      "step": 23925
+    },
+    {
+      "epoch": 47.852,
+      "grad_norm": 1.1597468852996826,
+      "learning_rate": 2e-05,
+      "loss": 0.04433453,
+      "step": 23926
+    },
+    {
+      "epoch": 47.854,
+      "grad_norm": 1.1068360805511475,
+      "learning_rate": 2e-05,
+      "loss": 0.04815186,
+      "step": 23927
+    },
+    {
+      "epoch": 47.856,
+      "grad_norm": 1.0562989711761475,
+      "learning_rate": 2e-05,
+      "loss": 0.039212,
+      "step": 23928
+    },
+    {
+      "epoch": 47.858,
+      "grad_norm": 1.2464039325714111,
+      "learning_rate": 2e-05,
+      "loss": 0.05434022,
+      "step": 23929
+    },
+    {
+      "epoch": 47.86,
+      "grad_norm": 1.9144781827926636,
+      "learning_rate": 2e-05,
+      "loss": 0.0500843,
+      "step": 23930
+    },
+    {
+      "epoch": 47.862,
+      "grad_norm": 4.982010364532471,
+      "learning_rate": 2e-05,
+      "loss": 0.05325773,
+      "step": 23931
+    },
+    {
+      "epoch": 47.864,
+      "grad_norm": 2.1956326961517334,
+      "learning_rate": 2e-05,
+      "loss": 0.04757648,
+      "step": 23932
+    },
+    {
+      "epoch": 47.866,
+      "grad_norm": 1.3570915460586548,
+      "learning_rate": 2e-05,
+      "loss": 0.05788545,
+      "step": 23933
+    },
+    {
+      "epoch": 47.868,
+      "grad_norm": 1.1021960973739624,
+      "learning_rate": 2e-05,
+      "loss": 0.04605621,
+      "step": 23934
+    },
+    {
+      "epoch": 47.87,
+      "grad_norm": 1.2763595581054688,
+      "learning_rate": 2e-05,
+      "loss": 0.04462332,
+      "step": 23935
+    },
+    {
+      "epoch": 47.872,
+      "grad_norm": 1.2677397727966309,
+      "learning_rate": 2e-05,
+      "loss": 0.06458889,
+      "step": 23936
+    },
+    {
+      "epoch": 47.874,
+      "grad_norm": 1.571900486946106,
+      "learning_rate": 2e-05,
+      "loss": 0.04743953,
+      "step": 23937
+    },
+    {
+      "epoch": 47.876,
+      "grad_norm": 1.2645798921585083,
+      "learning_rate": 2e-05,
+      "loss": 0.04014737,
+      "step": 23938
+    },
+    {
+      "epoch": 47.878,
+      "grad_norm": 1.3748975992202759,
+      "learning_rate": 2e-05,
+      "loss": 0.05016341,
+      "step": 23939
+    },
+    {
+      "epoch": 47.88,
+      "grad_norm": 1.0344460010528564,
+      "learning_rate": 2e-05,
+      "loss": 0.03785864,
+      "step": 23940
+    },
+    {
+      "epoch": 47.882,
+      "grad_norm": 1.15352463722229,
+      "learning_rate": 2e-05,
+      "loss": 0.04408693,
+      "step": 23941
+    },
+    {
+      "epoch": 47.884,
+      "grad_norm": 0.9576112627983093,
+      "learning_rate": 2e-05,
+      "loss": 0.0283091,
+      "step": 23942
+    },
+    {
+      "epoch": 47.886,
+      "grad_norm": 1.1543420553207397,
+      "learning_rate": 2e-05,
+      "loss": 0.03587131,
+      "step": 23943
+    },
+    {
+      "epoch": 47.888,
+      "grad_norm": 1.3030697107315063,
+      "learning_rate": 2e-05,
+      "loss": 0.06125759,
+      "step": 23944
+    },
+    {
+      "epoch": 47.89,
+      "grad_norm": 1.159096598625183,
+      "learning_rate": 2e-05,
+      "loss": 0.04972298,
+      "step": 23945
+    },
+    {
+      "epoch": 47.892,
+      "grad_norm": 1.3995835781097412,
+      "learning_rate": 2e-05,
+      "loss": 0.03951022,
+      "step": 23946
+    },
+    {
+      "epoch": 47.894,
+      "grad_norm": 1.2913930416107178,
+      "learning_rate": 2e-05,
+      "loss": 0.04597832,
+      "step": 23947
+    },
+    {
+      "epoch": 47.896,
+      "grad_norm": 1.0816295146942139,
+      "learning_rate": 2e-05,
+      "loss": 0.03365443,
+      "step": 23948
+    },
+    {
+      "epoch": 47.898,
+      "grad_norm": 1.2228970527648926,
+      "learning_rate": 2e-05,
+      "loss": 0.06327063,
+      "step": 23949
+    },
+    {
+      "epoch": 47.9,
+      "grad_norm": 1.1971608400344849,
+      "learning_rate": 2e-05,
+      "loss": 0.04186807,
+      "step": 23950
+    },
+    {
+      "epoch": 47.902,
+      "grad_norm": 1.3792004585266113,
+      "learning_rate": 2e-05,
+      "loss": 0.04573726,
+      "step": 23951
+    },
+    {
+      "epoch": 47.904,
+      "grad_norm": 1.3499553203582764,
+      "learning_rate": 2e-05,
+      "loss": 0.04411193,
+      "step": 23952
+    },
+    {
+      "epoch": 47.906,
+      "grad_norm": 0.9711533188819885,
+      "learning_rate": 2e-05,
+      "loss": 0.04265768,
+      "step": 23953
+    },
+    {
+      "epoch": 47.908,
+      "grad_norm": 1.329298496246338,
+      "learning_rate": 2e-05,
+      "loss": 0.0500191,
+      "step": 23954
+    },
+    {
+      "epoch": 47.91,
+      "grad_norm": 1.0052729845046997,
+      "learning_rate": 2e-05,
+      "loss": 0.04355924,
+      "step": 23955
+    },
+    {
+      "epoch": 47.912,
+      "grad_norm": 1.1501191854476929,
+      "learning_rate": 2e-05,
+      "loss": 0.03861841,
+      "step": 23956
+    },
+    {
+      "epoch": 47.914,
+      "grad_norm": 1.1558641195297241,
+      "learning_rate": 2e-05,
+      "loss": 0.05884036,
+      "step": 23957
+    },
+    {
+      "epoch": 47.916,
+      "grad_norm": 1.0510032176971436,
+      "learning_rate": 2e-05,
+      "loss": 0.03164276,
+      "step": 23958
+    },
+    {
+      "epoch": 47.918,
+      "grad_norm": 1.3713427782058716,
+      "learning_rate": 2e-05,
+      "loss": 0.05324286,
+      "step": 23959
+    },
+    {
+      "epoch": 47.92,
+      "grad_norm": 1.4690266847610474,
+      "learning_rate": 2e-05,
+      "loss": 0.04128034,
+      "step": 23960
+    },
+    {
+      "epoch": 47.922,
+      "grad_norm": 2.2947113513946533,
+      "learning_rate": 2e-05,
+      "loss": 0.04311752,
+      "step": 23961
+    },
+    {
+      "epoch": 47.924,
+      "grad_norm": 1.1298458576202393,
+      "learning_rate": 2e-05,
+      "loss": 0.04617842,
+      "step": 23962
+    },
+    {
+      "epoch": 47.926,
+      "grad_norm": 1.2151092290878296,
+      "learning_rate": 2e-05,
+      "loss": 0.0448565,
+      "step": 23963
+    },
+    {
+      "epoch": 47.928,
+      "grad_norm": 1.210721731185913,
+      "learning_rate": 2e-05,
+      "loss": 0.03401937,
+      "step": 23964
+    },
+    {
+      "epoch": 47.93,
+      "grad_norm": 1.6079378128051758,
+      "learning_rate": 2e-05,
+      "loss": 0.04983381,
+      "step": 23965
+    },
+    {
+      "epoch": 47.932,
+      "grad_norm": 0.9416164755821228,
+      "learning_rate": 2e-05,
+      "loss": 0.0281193,
+      "step": 23966
+    },
+    {
+      "epoch": 47.934,
+      "grad_norm": 1.450137972831726,
+      "learning_rate": 2e-05,
+      "loss": 0.05513122,
+      "step": 23967
+    },
+    {
+      "epoch": 47.936,
+      "grad_norm": 1.427017092704773,
+      "learning_rate": 2e-05,
+      "loss": 0.06112292,
+      "step": 23968
+    },
+    {
+      "epoch": 47.938,
+      "grad_norm": 1.1121290922164917,
+      "learning_rate": 2e-05,
+      "loss": 0.04755298,
+      "step": 23969
+    },
+    {
+      "epoch": 47.94,
+      "grad_norm": 1.350212812423706,
+      "learning_rate": 2e-05,
+      "loss": 0.02974812,
+      "step": 23970
+    },
+    {
+      "epoch": 47.942,
+      "grad_norm": 1.009774088859558,
+      "learning_rate": 2e-05,
+      "loss": 0.03674398,
+      "step": 23971
+    },
+    {
+      "epoch": 47.944,
+      "grad_norm": 0.989896297454834,
+      "learning_rate": 2e-05,
+      "loss": 0.04537395,
+      "step": 23972
+    },
+    {
+      "epoch": 47.946,
+      "grad_norm": 1.3564207553863525,
+      "learning_rate": 2e-05,
+      "loss": 0.0425001,
+      "step": 23973
+    },
+    {
+      "epoch": 47.948,
+      "grad_norm": 2.2842447757720947,
+      "learning_rate": 2e-05,
+      "loss": 0.04634003,
+      "step": 23974
+    },
+    {
+      "epoch": 47.95,
+      "grad_norm": 1.2380462884902954,
+      "learning_rate": 2e-05,
+      "loss": 0.04033655,
+      "step": 23975
+    },
+    {
+      "epoch": 47.952,
+      "grad_norm": 1.2136390209197998,
+      "learning_rate": 2e-05,
+      "loss": 0.03208754,
+      "step": 23976
+    },
+    {
+      "epoch": 47.954,
+      "grad_norm": 1.2181363105773926,
+      "learning_rate": 2e-05,
+      "loss": 0.05851332,
+      "step": 23977
+    },
+    {
+      "epoch": 47.956,
+      "grad_norm": 1.1737416982650757,
+      "learning_rate": 2e-05,
+      "loss": 0.04869546,
+      "step": 23978
+    },
+    {
+      "epoch": 47.958,
+      "grad_norm": 1.1523841619491577,
+      "learning_rate": 2e-05,
+      "loss": 0.04287051,
+      "step": 23979
+    },
+    {
+      "epoch": 47.96,
+      "grad_norm": 1.3035601377487183,
+      "learning_rate": 2e-05,
+      "loss": 0.05795803,
+      "step": 23980
+    },
+    {
+      "epoch": 47.962,
+      "grad_norm": 1.4103642702102661,
+      "learning_rate": 2e-05,
+      "loss": 0.0504769,
+      "step": 23981
+    },
+    {
+      "epoch": 47.964,
+      "grad_norm": 1.3686590194702148,
+      "learning_rate": 2e-05,
+      "loss": 0.05023767,
+      "step": 23982
+    },
+    {
+      "epoch": 47.966,
+      "grad_norm": 1.1425793170928955,
+      "learning_rate": 2e-05,
+      "loss": 0.03801563,
+      "step": 23983
+    },
+    {
+      "epoch": 47.968,
+      "grad_norm": 1.2969728708267212,
+      "learning_rate": 2e-05,
+      "loss": 0.04899958,
+      "step": 23984
+    },
+    {
+      "epoch": 47.97,
+      "grad_norm": 1.312829613685608,
+      "learning_rate": 2e-05,
+      "loss": 0.06080573,
+      "step": 23985
+    },
+    {
+      "epoch": 47.972,
+      "grad_norm": 1.264377236366272,
+      "learning_rate": 2e-05,
+      "loss": 0.04977002,
+      "step": 23986
+    },
+    {
+      "epoch": 47.974,
+      "grad_norm": 2.843947410583496,
+      "learning_rate": 2e-05,
+      "loss": 0.06198847,
+      "step": 23987
+    },
+    {
+      "epoch": 47.976,
+      "grad_norm": 1.4334129095077515,
+      "learning_rate": 2e-05,
+      "loss": 0.07072139,
+      "step": 23988
+    },
+    {
+      "epoch": 47.978,
+      "grad_norm": 1.3132730722427368,
+      "learning_rate": 2e-05,
+      "loss": 0.06276097,
+      "step": 23989
+    },
+    {
+      "epoch": 47.98,
+      "grad_norm": 1.2689872980117798,
+      "learning_rate": 2e-05,
+      "loss": 0.04908171,
+      "step": 23990
+    },
+    {
+      "epoch": 47.982,
+      "grad_norm": 1.4487504959106445,
+      "learning_rate": 2e-05,
+      "loss": 0.04362805,
+      "step": 23991
+    },
+    {
+      "epoch": 47.984,
+      "grad_norm": 1.1339962482452393,
+      "learning_rate": 2e-05,
+      "loss": 0.05087781,
+      "step": 23992
+    },
+    {
+      "epoch": 47.986,
+      "grad_norm": 1.7450958490371704,
+      "learning_rate": 2e-05,
+      "loss": 0.05819881,
+      "step": 23993
+    },
+    {
+      "epoch": 47.988,
+      "grad_norm": 1.1396898031234741,
+      "learning_rate": 2e-05,
+      "loss": 0.04882061,
+      "step": 23994
+    },
+    {
+      "epoch": 47.99,
+      "grad_norm": 1.337996482849121,
+      "learning_rate": 2e-05,
+      "loss": 0.04781748,
+      "step": 23995
+    },
+    {
+      "epoch": 47.992,
+      "grad_norm": 1.5428389310836792,
+      "learning_rate": 2e-05,
+      "loss": 0.0593975,
+      "step": 23996
+    },
+    {
+      "epoch": 47.994,
+      "grad_norm": 1.029953956604004,
+      "learning_rate": 2e-05,
+      "loss": 0.03917014,
+      "step": 23997
+    },
+    {
+      "epoch": 47.996,
+      "grad_norm": 1.3269636631011963,
+      "learning_rate": 2e-05,
+      "loss": 0.06252966,
+      "step": 23998
+    },
+    {
+      "epoch": 47.998,
+      "grad_norm": 1.4018568992614746,
+      "learning_rate": 2e-05,
+      "loss": 0.05506238,
+      "step": 23999
+    },
+    {
+      "epoch": 48.0,
+      "grad_norm": 1.3268723487854004,
+      "learning_rate": 2e-05,
+      "loss": 0.04358728,
+      "step": 24000
+    },
+    {
+      "epoch": 48.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9840319361277445,
+        "Equal_1": 0.998,
+        "Equal_2": 0.9820359281437125,
+        "Equal_3": 0.9900199600798403,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 1.0,
+        "Parallel_1": 0.9939879759519038,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 1.0,
+        "Perpendicular_2": 1.0,
+        "Perpendicular_3": 0.9098196392785571,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 1.0,
+        "PointLiesOnCircle_3": 0.9916,
+        "PointLiesOnLine_1": 0.9979959919839679,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9940119760479041
+      },
+      "eval_runtime": 320.3246,
+      "eval_samples_per_second": 32.779,
+      "eval_steps_per_second": 0.656,
+      "step": 24000
+    },
+    {
+      "epoch": 48.002,
+      "grad_norm": 1.0694226026535034,
+      "learning_rate": 2e-05,
+      "loss": 0.03889792,
+      "step": 24001
+    },
+    {
+      "epoch": 48.004,
+      "grad_norm": 1.1457878351211548,
+      "learning_rate": 2e-05,
+      "loss": 0.04411246,
+      "step": 24002
+    },
+    {
+      "epoch": 48.006,
+      "grad_norm": 1.3555324077606201,
+      "learning_rate": 2e-05,
+      "loss": 0.05262123,
+      "step": 24003
+    },
+    {
+      "epoch": 48.008,
+      "grad_norm": 1.038233995437622,
+      "learning_rate": 2e-05,
+      "loss": 0.04188913,
+      "step": 24004
+    },
+    {
+      "epoch": 48.01,
+      "grad_norm": 3.3475518226623535,
+      "learning_rate": 2e-05,
+      "loss": 0.05884375,
+      "step": 24005
+    },
+    {
+      "epoch": 48.012,
+      "grad_norm": 1.470353364944458,
+      "learning_rate": 2e-05,
+      "loss": 0.04114807,
+      "step": 24006
+    },
+    {
+      "epoch": 48.014,
+      "grad_norm": 1.2043060064315796,
+      "learning_rate": 2e-05,
+      "loss": 0.0573847,
+      "step": 24007
+    },
+    {
+      "epoch": 48.016,
+      "grad_norm": 1.1786561012268066,
+      "learning_rate": 2e-05,
+      "loss": 0.04742495,
+      "step": 24008
+    },
+    {
+      "epoch": 48.018,
+      "grad_norm": 1.0251394510269165,
+      "learning_rate": 2e-05,
+      "loss": 0.0432762,
+      "step": 24009
+    },
+    {
+      "epoch": 48.02,
+      "grad_norm": 1.1770861148834229,
+      "learning_rate": 2e-05,
+      "loss": 0.04839217,
+      "step": 24010
+    },
+    {
+      "epoch": 48.022,
+      "grad_norm": 0.9781084060668945,
+      "learning_rate": 2e-05,
+      "loss": 0.04288932,
+      "step": 24011
+    },
+    {
+      "epoch": 48.024,
+      "grad_norm": 1.5834667682647705,
+      "learning_rate": 2e-05,
+      "loss": 0.04880485,
+      "step": 24012
+    },
+    {
+      "epoch": 48.026,
+      "grad_norm": 2.6981749534606934,
+      "learning_rate": 2e-05,
+      "loss": 0.03831647,
+      "step": 24013
+    },
+    {
+      "epoch": 48.028,
+      "grad_norm": 1.1615235805511475,
+      "learning_rate": 2e-05,
+      "loss": 0.04750589,
+      "step": 24014
+    },
+    {
+      "epoch": 48.03,
+      "grad_norm": 1.1639854907989502,
+      "learning_rate": 2e-05,
+      "loss": 0.04984055,
+      "step": 24015
+    },
+    {
+      "epoch": 48.032,
+      "grad_norm": 1.0798165798187256,
+      "learning_rate": 2e-05,
+      "loss": 0.05147992,
+      "step": 24016
+    },
+    {
+      "epoch": 48.034,
+      "grad_norm": 1.0274465084075928,
+      "learning_rate": 2e-05,
+      "loss": 0.04617003,
+      "step": 24017
+    },
+    {
+      "epoch": 48.036,
+      "grad_norm": 1.038852334022522,
+      "learning_rate": 2e-05,
+      "loss": 0.0369803,
+      "step": 24018
+    },
+    {
+      "epoch": 48.038,
+      "grad_norm": 2.6861026287078857,
+      "learning_rate": 2e-05,
+      "loss": 0.07091722,
+      "step": 24019
+    },
+    {
+      "epoch": 48.04,
+      "grad_norm": 1.3858764171600342,
+      "learning_rate": 2e-05,
+      "loss": 0.04682039,
+      "step": 24020
+    },
+    {
+      "epoch": 48.042,
+      "grad_norm": 1.1013208627700806,
+      "learning_rate": 2e-05,
+      "loss": 0.04063297,
+      "step": 24021
+    },
+    {
+      "epoch": 48.044,
+      "grad_norm": 1.4258663654327393,
+      "learning_rate": 2e-05,
+      "loss": 0.05036592,
+      "step": 24022
+    },
+    {
+      "epoch": 48.046,
+      "grad_norm": 1.1322613954544067,
+      "learning_rate": 2e-05,
+      "loss": 0.04924812,
+      "step": 24023
+    },
+    {
+      "epoch": 48.048,
+      "grad_norm": 1.7514662742614746,
+      "learning_rate": 2e-05,
+      "loss": 0.06260653,
+      "step": 24024
+    },
+    {
+      "epoch": 48.05,
+      "grad_norm": 1.3482943773269653,
+      "learning_rate": 2e-05,
+      "loss": 0.05003114,
+      "step": 24025
+    },
+    {
+      "epoch": 48.052,
+      "grad_norm": 1.5502170324325562,
+      "learning_rate": 2e-05,
+      "loss": 0.04645883,
+      "step": 24026
+    },
+    {
+      "epoch": 48.054,
+      "grad_norm": 1.1267354488372803,
+      "learning_rate": 2e-05,
+      "loss": 0.04372516,
+      "step": 24027
+    },
+    {
+      "epoch": 48.056,
+      "grad_norm": 1.668345332145691,
+      "learning_rate": 2e-05,
+      "loss": 0.07128799,
+      "step": 24028
+    },
+    {
+      "epoch": 48.058,
+      "grad_norm": 2.1013059616088867,
+      "learning_rate": 2e-05,
+      "loss": 0.04920814,
+      "step": 24029
+    },
+    {
+      "epoch": 48.06,
+      "grad_norm": 1.272568941116333,
+      "learning_rate": 2e-05,
+      "loss": 0.04338931,
+      "step": 24030
+    },
+    {
+      "epoch": 48.062,
+      "grad_norm": 1.102447748184204,
+      "learning_rate": 2e-05,
+      "loss": 0.03770321,
+      "step": 24031
+    },
+    {
+      "epoch": 48.064,
+      "grad_norm": 1.8323739767074585,
+      "learning_rate": 2e-05,
+      "loss": 0.05331714,
+      "step": 24032
+    },
+    {
+      "epoch": 48.066,
+      "grad_norm": 2.065495491027832,
+      "learning_rate": 2e-05,
+      "loss": 0.05580588,
+      "step": 24033
+    },
+    {
+      "epoch": 48.068,
+      "grad_norm": 4.448790073394775,
+      "learning_rate": 2e-05,
+      "loss": 0.04371783,
+      "step": 24034
+    },
+    {
+      "epoch": 48.07,
+      "grad_norm": 1.122683048248291,
+      "learning_rate": 2e-05,
+      "loss": 0.05487248,
+      "step": 24035
+    },
+    {
+      "epoch": 48.072,
+      "grad_norm": 1.2778655290603638,
+      "learning_rate": 2e-05,
+      "loss": 0.0487362,
+      "step": 24036
+    },
+    {
+      "epoch": 48.074,
+      "grad_norm": 1.5660276412963867,
+      "learning_rate": 2e-05,
+      "loss": 0.05149665,
+      "step": 24037
+    },
+    {
+      "epoch": 48.076,
+      "grad_norm": 1.2964324951171875,
+      "learning_rate": 2e-05,
+      "loss": 0.04056223,
+      "step": 24038
+    },
+    {
+      "epoch": 48.078,
+      "grad_norm": 1.1524419784545898,
+      "learning_rate": 2e-05,
+      "loss": 0.05273098,
+      "step": 24039
+    },
+    {
+      "epoch": 48.08,
+      "grad_norm": 1.4630237817764282,
+      "learning_rate": 2e-05,
+      "loss": 0.05465612,
+      "step": 24040
+    },
+    {
+      "epoch": 48.082,
+      "grad_norm": 1.2101308107376099,
+      "learning_rate": 2e-05,
+      "loss": 0.04670757,
+      "step": 24041
+    },
+    {
+      "epoch": 48.084,
+      "grad_norm": 1.1870620250701904,
+      "learning_rate": 2e-05,
+      "loss": 0.04672918,
+      "step": 24042
+    },
+    {
+      "epoch": 48.086,
+      "grad_norm": 1.2451860904693604,
+      "learning_rate": 2e-05,
+      "loss": 0.05255879,
+      "step": 24043
+    },
+    {
+      "epoch": 48.088,
+      "grad_norm": 1.1772750616073608,
+      "learning_rate": 2e-05,
+      "loss": 0.05022447,
+      "step": 24044
+    },
+    {
+      "epoch": 48.09,
+      "grad_norm": 1.4949736595153809,
+      "learning_rate": 2e-05,
+      "loss": 0.05462297,
+      "step": 24045
+    },
+    {
+      "epoch": 48.092,
+      "grad_norm": 1.0446524620056152,
+      "learning_rate": 2e-05,
+      "loss": 0.05005912,
+      "step": 24046
+    },
+    {
+      "epoch": 48.094,
+      "grad_norm": 1.2837433815002441,
+      "learning_rate": 2e-05,
+      "loss": 0.04919739,
+      "step": 24047
+    },
+    {
+      "epoch": 48.096,
+      "grad_norm": 1.2766021490097046,
+      "learning_rate": 2e-05,
+      "loss": 0.05800232,
+      "step": 24048
+    },
+    {
+      "epoch": 48.098,
+      "grad_norm": 1.386711835861206,
+      "learning_rate": 2e-05,
+      "loss": 0.05083595,
+      "step": 24049
+    },
+    {
+      "epoch": 48.1,
+      "grad_norm": 0.9854950308799744,
+      "learning_rate": 2e-05,
+      "loss": 0.03358131,
+      "step": 24050
+    },
+    {
+      "epoch": 48.102,
+      "grad_norm": 1.1516034603118896,
+      "learning_rate": 2e-05,
+      "loss": 0.04865019,
+      "step": 24051
+    },
+    {
+      "epoch": 48.104,
+      "grad_norm": 1.855728030204773,
+      "learning_rate": 2e-05,
+      "loss": 0.06086725,
+      "step": 24052
+    },
+    {
+      "epoch": 48.106,
+      "grad_norm": 1.3824462890625,
+      "learning_rate": 2e-05,
+      "loss": 0.03999291,
+      "step": 24053
+    },
+    {
+      "epoch": 48.108,
+      "grad_norm": 1.1911598443984985,
+      "learning_rate": 2e-05,
+      "loss": 0.04892493,
+      "step": 24054
+    },
+    {
+      "epoch": 48.11,
+      "grad_norm": 1.2059804201126099,
+      "learning_rate": 2e-05,
+      "loss": 0.05040662,
+      "step": 24055
+    },
+    {
+      "epoch": 48.112,
+      "grad_norm": 1.4151570796966553,
+      "learning_rate": 2e-05,
+      "loss": 0.05632602,
+      "step": 24056
+    },
+    {
+      "epoch": 48.114,
+      "grad_norm": 1.101078748703003,
+      "learning_rate": 2e-05,
+      "loss": 0.03802776,
+      "step": 24057
+    },
+    {
+      "epoch": 48.116,
+      "grad_norm": 1.3649169206619263,
+      "learning_rate": 2e-05,
+      "loss": 0.06694816,
+      "step": 24058
+    },
+    {
+      "epoch": 48.118,
+      "grad_norm": 1.1776071786880493,
+      "learning_rate": 2e-05,
+      "loss": 0.05049934,
+      "step": 24059
+    },
+    {
+      "epoch": 48.12,
+      "grad_norm": 1.082025408744812,
+      "learning_rate": 2e-05,
+      "loss": 0.04495796,
+      "step": 24060
+    },
+    {
+      "epoch": 48.122,
+      "grad_norm": 1.2222503423690796,
+      "learning_rate": 2e-05,
+      "loss": 0.04708505,
+      "step": 24061
+    },
+    {
+      "epoch": 48.124,
+      "grad_norm": 1.5613664388656616,
+      "learning_rate": 2e-05,
+      "loss": 0.05574992,
+      "step": 24062
+    },
+    {
+      "epoch": 48.126,
+      "grad_norm": 1.2608956098556519,
+      "learning_rate": 2e-05,
+      "loss": 0.04397441,
+      "step": 24063
+    },
+    {
+      "epoch": 48.128,
+      "grad_norm": 0.9204988479614258,
+      "learning_rate": 2e-05,
+      "loss": 0.02955709,
+      "step": 24064
+    },
+    {
+      "epoch": 48.13,
+      "grad_norm": 0.984535813331604,
+      "learning_rate": 2e-05,
+      "loss": 0.04042712,
+      "step": 24065
+    },
+    {
+      "epoch": 48.132,
+      "grad_norm": 5.602854251861572,
+      "learning_rate": 2e-05,
+      "loss": 0.046597,
+      "step": 24066
+    },
+    {
+      "epoch": 48.134,
+      "grad_norm": 1.278770089149475,
+      "learning_rate": 2e-05,
+      "loss": 0.03210206,
+      "step": 24067
+    },
+    {
+      "epoch": 48.136,
+      "grad_norm": 0.9865384697914124,
+      "learning_rate": 2e-05,
+      "loss": 0.03966504,
+      "step": 24068
+    },
+    {
+      "epoch": 48.138,
+      "grad_norm": 1.1876673698425293,
+      "learning_rate": 2e-05,
+      "loss": 0.0427718,
+      "step": 24069
+    },
+    {
+      "epoch": 48.14,
+      "grad_norm": 1.1486650705337524,
+      "learning_rate": 2e-05,
+      "loss": 0.04623136,
+      "step": 24070
+    },
+    {
+      "epoch": 48.142,
+      "grad_norm": 1.3332282304763794,
+      "learning_rate": 2e-05,
+      "loss": 0.05006913,
+      "step": 24071
+    },
+    {
+      "epoch": 48.144,
+      "grad_norm": 1.2728177309036255,
+      "learning_rate": 2e-05,
+      "loss": 0.05031462,
+      "step": 24072
+    },
+    {
+      "epoch": 48.146,
+      "grad_norm": 1.2598438262939453,
+      "learning_rate": 2e-05,
+      "loss": 0.04445124,
+      "step": 24073
+    },
+    {
+      "epoch": 48.148,
+      "grad_norm": 1.0530667304992676,
+      "learning_rate": 2e-05,
+      "loss": 0.0403815,
+      "step": 24074
+    },
+    {
+      "epoch": 48.15,
+      "grad_norm": 1.0963428020477295,
+      "learning_rate": 2e-05,
+      "loss": 0.0428313,
+      "step": 24075
+    },
+    {
+      "epoch": 48.152,
+      "grad_norm": 1.0206382274627686,
+      "learning_rate": 2e-05,
+      "loss": 0.04635064,
+      "step": 24076
+    },
+    {
+      "epoch": 48.154,
+      "grad_norm": 1.0546106100082397,
+      "learning_rate": 2e-05,
+      "loss": 0.0390348,
+      "step": 24077
+    },
+    {
+      "epoch": 48.156,
+      "grad_norm": 1.1284135580062866,
+      "learning_rate": 2e-05,
+      "loss": 0.03736102,
+      "step": 24078
+    },
+    {
+      "epoch": 48.158,
+      "grad_norm": 1.389346957206726,
+      "learning_rate": 2e-05,
+      "loss": 0.04481775,
+      "step": 24079
+    },
+    {
+      "epoch": 48.16,
+      "grad_norm": 1.1270530223846436,
+      "learning_rate": 2e-05,
+      "loss": 0.04756751,
+      "step": 24080
+    },
+    {
+      "epoch": 48.162,
+      "grad_norm": 1.2751901149749756,
+      "learning_rate": 2e-05,
+      "loss": 0.04584195,
+      "step": 24081
+    },
+    {
+      "epoch": 48.164,
+      "grad_norm": 1.0802828073501587,
+      "learning_rate": 2e-05,
+      "loss": 0.06117968,
+      "step": 24082
+    },
+    {
+      "epoch": 48.166,
+      "grad_norm": 1.2902870178222656,
+      "learning_rate": 2e-05,
+      "loss": 0.05666249,
+      "step": 24083
+    },
+    {
+      "epoch": 48.168,
+      "grad_norm": 1.07093346118927,
+      "learning_rate": 2e-05,
+      "loss": 0.04220451,
+      "step": 24084
+    },
+    {
+      "epoch": 48.17,
+      "grad_norm": 0.9996381998062134,
+      "learning_rate": 2e-05,
+      "loss": 0.03157579,
+      "step": 24085
+    },
+    {
+      "epoch": 48.172,
+      "grad_norm": 1.4622588157653809,
+      "learning_rate": 2e-05,
+      "loss": 0.04022121,
+      "step": 24086
+    },
+    {
+      "epoch": 48.174,
+      "grad_norm": 1.200003981590271,
+      "learning_rate": 2e-05,
+      "loss": 0.03526891,
+      "step": 24087
+    },
+    {
+      "epoch": 48.176,
+      "grad_norm": 1.3392802476882935,
+      "learning_rate": 2e-05,
+      "loss": 0.03042413,
+      "step": 24088
+    },
+    {
+      "epoch": 48.178,
+      "grad_norm": 1.0749857425689697,
+      "learning_rate": 2e-05,
+      "loss": 0.04232724,
+      "step": 24089
+    },
+    {
+      "epoch": 48.18,
+      "grad_norm": 1.1630322933197021,
+      "learning_rate": 2e-05,
+      "loss": 0.04984683,
+      "step": 24090
+    },
+    {
+      "epoch": 48.182,
+      "grad_norm": 1.1708325147628784,
+      "learning_rate": 2e-05,
+      "loss": 0.04399581,
+      "step": 24091
+    },
+    {
+      "epoch": 48.184,
+      "grad_norm": 2.851407527923584,
+      "learning_rate": 2e-05,
+      "loss": 0.07037769,
+      "step": 24092
+    },
+    {
+      "epoch": 48.186,
+      "grad_norm": 1.0833169221878052,
+      "learning_rate": 2e-05,
+      "loss": 0.04655275,
+      "step": 24093
+    },
+    {
+      "epoch": 48.188,
+      "grad_norm": 1.372167706489563,
+      "learning_rate": 2e-05,
+      "loss": 0.07025443,
+      "step": 24094
+    },
+    {
+      "epoch": 48.19,
+      "grad_norm": 1.2088648080825806,
+      "learning_rate": 2e-05,
+      "loss": 0.04490817,
+      "step": 24095
+    },
+    {
+      "epoch": 48.192,
+      "grad_norm": 1.134153962135315,
+      "learning_rate": 2e-05,
+      "loss": 0.05215061,
+      "step": 24096
+    },
+    {
+      "epoch": 48.194,
+      "grad_norm": 1.1450132131576538,
+      "learning_rate": 2e-05,
+      "loss": 0.05180396,
+      "step": 24097
+    },
+    {
+      "epoch": 48.196,
+      "grad_norm": 1.1347380876541138,
+      "learning_rate": 2e-05,
+      "loss": 0.04732578,
+      "step": 24098
+    },
+    {
+      "epoch": 48.198,
+      "grad_norm": 1.1348960399627686,
+      "learning_rate": 2e-05,
+      "loss": 0.05249636,
+      "step": 24099
+    },
+    {
+      "epoch": 48.2,
+      "grad_norm": 1.7801083326339722,
+      "learning_rate": 2e-05,
+      "loss": 0.05126803,
+      "step": 24100
+    },
+    {
+      "epoch": 48.202,
+      "grad_norm": 1.2715392112731934,
+      "learning_rate": 2e-05,
+      "loss": 0.05465782,
+      "step": 24101
+    },
+    {
+      "epoch": 48.204,
+      "grad_norm": 0.9739833474159241,
+      "learning_rate": 2e-05,
+      "loss": 0.03681491,
+      "step": 24102
+    },
+    {
+      "epoch": 48.206,
+      "grad_norm": 0.991855800151825,
+      "learning_rate": 2e-05,
+      "loss": 0.03075913,
+      "step": 24103
+    },
+    {
+      "epoch": 48.208,
+      "grad_norm": 1.7213369607925415,
+      "learning_rate": 2e-05,
+      "loss": 0.03983373,
+      "step": 24104
+    },
+    {
+      "epoch": 48.21,
+      "grad_norm": 1.0067009925842285,
+      "learning_rate": 2e-05,
+      "loss": 0.04427213,
+      "step": 24105
+    },
+    {
+      "epoch": 48.212,
+      "grad_norm": 1.8340460062026978,
+      "learning_rate": 2e-05,
+      "loss": 0.04303214,
+      "step": 24106
+    },
+    {
+      "epoch": 48.214,
+      "grad_norm": 1.1613703966140747,
+      "learning_rate": 2e-05,
+      "loss": 0.04921772,
+      "step": 24107
+    },
+    {
+      "epoch": 48.216,
+      "grad_norm": 2.5453414916992188,
+      "learning_rate": 2e-05,
+      "loss": 0.05552774,
+      "step": 24108
+    },
+    {
+      "epoch": 48.218,
+      "grad_norm": 1.020289659500122,
+      "learning_rate": 2e-05,
+      "loss": 0.03909015,
+      "step": 24109
+    },
+    {
+      "epoch": 48.22,
+      "grad_norm": 1.1632663011550903,
+      "learning_rate": 2e-05,
+      "loss": 0.05679201,
+      "step": 24110
+    },
+    {
+      "epoch": 48.222,
+      "grad_norm": 0.987187385559082,
+      "learning_rate": 2e-05,
+      "loss": 0.03800602,
+      "step": 24111
+    },
+    {
+      "epoch": 48.224,
+      "grad_norm": 1.2574355602264404,
+      "learning_rate": 2e-05,
+      "loss": 0.05340779,
+      "step": 24112
+    },
+    {
+      "epoch": 48.226,
+      "grad_norm": 1.076492428779602,
+      "learning_rate": 2e-05,
+      "loss": 0.04606473,
+      "step": 24113
+    },
+    {
+      "epoch": 48.228,
+      "grad_norm": 1.010301947593689,
+      "learning_rate": 2e-05,
+      "loss": 0.03338667,
+      "step": 24114
+    },
+    {
+      "epoch": 48.23,
+      "grad_norm": 1.3499643802642822,
+      "learning_rate": 2e-05,
+      "loss": 0.06098785,
+      "step": 24115
+    },
+    {
+      "epoch": 48.232,
+      "grad_norm": 1.1102203130722046,
+      "learning_rate": 2e-05,
+      "loss": 0.04781565,
+      "step": 24116
+    },
+    {
+      "epoch": 48.234,
+      "grad_norm": 0.941985011100769,
+      "learning_rate": 2e-05,
+      "loss": 0.03292421,
+      "step": 24117
+    },
+    {
+      "epoch": 48.236,
+      "grad_norm": 1.1043226718902588,
+      "learning_rate": 2e-05,
+      "loss": 0.0464811,
+      "step": 24118
+    },
+    {
+      "epoch": 48.238,
+      "grad_norm": 0.9736769795417786,
+      "learning_rate": 2e-05,
+      "loss": 0.04281894,
+      "step": 24119
+    },
+    {
+      "epoch": 48.24,
+      "grad_norm": 1.1103366613388062,
+      "learning_rate": 2e-05,
+      "loss": 0.05232271,
+      "step": 24120
+    },
+    {
+      "epoch": 48.242,
+      "grad_norm": 1.2056519985198975,
+      "learning_rate": 2e-05,
+      "loss": 0.05404781,
+      "step": 24121
+    },
+    {
+      "epoch": 48.244,
+      "grad_norm": 1.410598874092102,
+      "learning_rate": 2e-05,
+      "loss": 0.06230663,
+      "step": 24122
+    },
+    {
+      "epoch": 48.246,
+      "grad_norm": 1.136397123336792,
+      "learning_rate": 2e-05,
+      "loss": 0.05437019,
+      "step": 24123
+    },
+    {
+      "epoch": 48.248,
+      "grad_norm": 1.0891176462173462,
+      "learning_rate": 2e-05,
+      "loss": 0.04057914,
+      "step": 24124
+    },
+    {
+      "epoch": 48.25,
+      "grad_norm": 1.1873037815093994,
+      "learning_rate": 2e-05,
+      "loss": 0.04513947,
+      "step": 24125
+    },
+    {
+      "epoch": 48.252,
+      "grad_norm": 1.5150892734527588,
+      "learning_rate": 2e-05,
+      "loss": 0.0410989,
+      "step": 24126
+    },
+    {
+      "epoch": 48.254,
+      "grad_norm": 2.031759738922119,
+      "learning_rate": 2e-05,
+      "loss": 0.05381062,
+      "step": 24127
+    },
+    {
+      "epoch": 48.256,
+      "grad_norm": 1.5702145099639893,
+      "learning_rate": 2e-05,
+      "loss": 0.03662251,
+      "step": 24128
+    },
+    {
+      "epoch": 48.258,
+      "grad_norm": 1.2305831909179688,
+      "learning_rate": 2e-05,
+      "loss": 0.06105634,
+      "step": 24129
+    },
+    {
+      "epoch": 48.26,
+      "grad_norm": 1.1910518407821655,
+      "learning_rate": 2e-05,
+      "loss": 0.04331035,
+      "step": 24130
+    },
+    {
+      "epoch": 48.262,
+      "grad_norm": 1.225128412246704,
+      "learning_rate": 2e-05,
+      "loss": 0.05862859,
+      "step": 24131
+    },
+    {
+      "epoch": 48.264,
+      "grad_norm": 0.9506001472473145,
+      "learning_rate": 2e-05,
+      "loss": 0.03767806,
+      "step": 24132
+    },
+    {
+      "epoch": 48.266,
+      "grad_norm": 1.2454578876495361,
+      "learning_rate": 2e-05,
+      "loss": 0.05315366,
+      "step": 24133
+    },
+    {
+      "epoch": 48.268,
+      "grad_norm": 2.151110887527466,
+      "learning_rate": 2e-05,
+      "loss": 0.04462639,
+      "step": 24134
+    },
+    {
+      "epoch": 48.27,
+      "grad_norm": 1.1853653192520142,
+      "learning_rate": 2e-05,
+      "loss": 0.05619586,
+      "step": 24135
+    },
+    {
+      "epoch": 48.272,
+      "grad_norm": 0.8391315937042236,
+      "learning_rate": 2e-05,
+      "loss": 0.02844811,
+      "step": 24136
+    },
+    {
+      "epoch": 48.274,
+      "grad_norm": 1.102542757987976,
+      "learning_rate": 2e-05,
+      "loss": 0.04351805,
+      "step": 24137
+    },
+    {
+      "epoch": 48.276,
+      "grad_norm": 0.9882407784461975,
+      "learning_rate": 2e-05,
+      "loss": 0.0304465,
+      "step": 24138
+    },
+    {
+      "epoch": 48.278,
+      "grad_norm": 1.748734951019287,
+      "learning_rate": 2e-05,
+      "loss": 0.06221665,
+      "step": 24139
+    },
+    {
+      "epoch": 48.28,
+      "grad_norm": 1.4740245342254639,
+      "learning_rate": 2e-05,
+      "loss": 0.06243878,
+      "step": 24140
+    },
+    {
+      "epoch": 48.282,
+      "grad_norm": 1.324210524559021,
+      "learning_rate": 2e-05,
+      "loss": 0.05721395,
+      "step": 24141
+    },
+    {
+      "epoch": 48.284,
+      "grad_norm": 1.527857780456543,
+      "learning_rate": 2e-05,
+      "loss": 0.0590317,
+      "step": 24142
+    },
+    {
+      "epoch": 48.286,
+      "grad_norm": 1.0875802040100098,
+      "learning_rate": 2e-05,
+      "loss": 0.04194984,
+      "step": 24143
+    },
+    {
+      "epoch": 48.288,
+      "grad_norm": 1.2001830339431763,
+      "learning_rate": 2e-05,
+      "loss": 0.05224827,
+      "step": 24144
+    },
+    {
+      "epoch": 48.29,
+      "grad_norm": 1.3207565546035767,
+      "learning_rate": 2e-05,
+      "loss": 0.05328258,
+      "step": 24145
+    },
+    {
+      "epoch": 48.292,
+      "grad_norm": 1.3341400623321533,
+      "learning_rate": 2e-05,
+      "loss": 0.03919051,
+      "step": 24146
+    },
+    {
+      "epoch": 48.294,
+      "grad_norm": 1.3471317291259766,
+      "learning_rate": 2e-05,
+      "loss": 0.0625244,
+      "step": 24147
+    },
+    {
+      "epoch": 48.296,
+      "grad_norm": 0.9852702021598816,
+      "learning_rate": 2e-05,
+      "loss": 0.04854707,
+      "step": 24148
+    },
+    {
+      "epoch": 48.298,
+      "grad_norm": 0.9104224443435669,
+      "learning_rate": 2e-05,
+      "loss": 0.03098441,
+      "step": 24149
+    },
+    {
+      "epoch": 48.3,
+      "grad_norm": 1.0037267208099365,
+      "learning_rate": 2e-05,
+      "loss": 0.03725686,
+      "step": 24150
+    },
+    {
+      "epoch": 48.302,
+      "grad_norm": 1.1196410655975342,
+      "learning_rate": 2e-05,
+      "loss": 0.04821742,
+      "step": 24151
+    },
+    {
+      "epoch": 48.304,
+      "grad_norm": 1.17573082447052,
+      "learning_rate": 2e-05,
+      "loss": 0.04223769,
+      "step": 24152
+    },
+    {
+      "epoch": 48.306,
+      "grad_norm": 1.1599558591842651,
+      "learning_rate": 2e-05,
+      "loss": 0.04793893,
+      "step": 24153
+    },
+    {
+      "epoch": 48.308,
+      "grad_norm": 2.1056137084960938,
+      "learning_rate": 2e-05,
+      "loss": 0.04927088,
+      "step": 24154
+    },
+    {
+      "epoch": 48.31,
+      "grad_norm": 1.2018673419952393,
+      "learning_rate": 2e-05,
+      "loss": 0.04321744,
+      "step": 24155
+    },
+    {
+      "epoch": 48.312,
+      "grad_norm": 1.2046828269958496,
+      "learning_rate": 2e-05,
+      "loss": 0.0523455,
+      "step": 24156
+    },
+    {
+      "epoch": 48.314,
+      "grad_norm": 1.2521579265594482,
+      "learning_rate": 2e-05,
+      "loss": 0.04741405,
+      "step": 24157
+    },
+    {
+      "epoch": 48.316,
+      "grad_norm": 1.046248435974121,
+      "learning_rate": 2e-05,
+      "loss": 0.03981638,
+      "step": 24158
+    },
+    {
+      "epoch": 48.318,
+      "grad_norm": 1.2036545276641846,
+      "learning_rate": 2e-05,
+      "loss": 0.04452758,
+      "step": 24159
+    },
+    {
+      "epoch": 48.32,
+      "grad_norm": 1.224697470664978,
+      "learning_rate": 2e-05,
+      "loss": 0.04996022,
+      "step": 24160
+    },
+    {
+      "epoch": 48.322,
+      "grad_norm": 1.4698989391326904,
+      "learning_rate": 2e-05,
+      "loss": 0.06074246,
+      "step": 24161
+    },
+    {
+      "epoch": 48.324,
+      "grad_norm": 1.205679178237915,
+      "learning_rate": 2e-05,
+      "loss": 0.05434452,
+      "step": 24162
+    },
+    {
+      "epoch": 48.326,
+      "grad_norm": 1.2876218557357788,
+      "learning_rate": 2e-05,
+      "loss": 0.04615651,
+      "step": 24163
+    },
+    {
+      "epoch": 48.328,
+      "grad_norm": 1.218860149383545,
+      "learning_rate": 2e-05,
+      "loss": 0.05063464,
+      "step": 24164
+    },
+    {
+      "epoch": 48.33,
+      "grad_norm": 1.225708246231079,
+      "learning_rate": 2e-05,
+      "loss": 0.04554639,
+      "step": 24165
+    },
+    {
+      "epoch": 48.332,
+      "grad_norm": 1.4999538660049438,
+      "learning_rate": 2e-05,
+      "loss": 0.0535819,
+      "step": 24166
+    },
+    {
+      "epoch": 48.334,
+      "grad_norm": 1.4471505880355835,
+      "learning_rate": 2e-05,
+      "loss": 0.05369381,
+      "step": 24167
+    },
+    {
+      "epoch": 48.336,
+      "grad_norm": 1.0038902759552002,
+      "learning_rate": 2e-05,
+      "loss": 0.03418017,
+      "step": 24168
+    },
+    {
+      "epoch": 48.338,
+      "grad_norm": 1.2840253114700317,
+      "learning_rate": 2e-05,
+      "loss": 0.05745913,
+      "step": 24169
+    },
+    {
+      "epoch": 48.34,
+      "grad_norm": 1.1269090175628662,
+      "learning_rate": 2e-05,
+      "loss": 0.04826582,
+      "step": 24170
+    },
+    {
+      "epoch": 48.342,
+      "grad_norm": 0.9834097623825073,
+      "learning_rate": 2e-05,
+      "loss": 0.0321577,
+      "step": 24171
+    },
+    {
+      "epoch": 48.344,
+      "grad_norm": 1.304100513458252,
+      "learning_rate": 2e-05,
+      "loss": 0.05313345,
+      "step": 24172
+    },
+    {
+      "epoch": 48.346,
+      "grad_norm": 0.9855166077613831,
+      "learning_rate": 2e-05,
+      "loss": 0.04511321,
+      "step": 24173
+    },
+    {
+      "epoch": 48.348,
+      "grad_norm": 1.117134928703308,
+      "learning_rate": 2e-05,
+      "loss": 0.04698523,
+      "step": 24174
+    },
+    {
+      "epoch": 48.35,
+      "grad_norm": 1.2151545286178589,
+      "learning_rate": 2e-05,
+      "loss": 0.03640284,
+      "step": 24175
+    },
+    {
+      "epoch": 48.352,
+      "grad_norm": 1.690683126449585,
+      "learning_rate": 2e-05,
+      "loss": 0.05967022,
+      "step": 24176
+    },
+    {
+      "epoch": 48.354,
+      "grad_norm": 1.0497173070907593,
+      "learning_rate": 2e-05,
+      "loss": 0.03619199,
+      "step": 24177
+    },
+    {
+      "epoch": 48.356,
+      "grad_norm": 1.0658822059631348,
+      "learning_rate": 2e-05,
+      "loss": 0.04109097,
+      "step": 24178
+    },
+    {
+      "epoch": 48.358,
+      "grad_norm": 1.1863605976104736,
+      "learning_rate": 2e-05,
+      "loss": 0.05850657,
+      "step": 24179
+    },
+    {
+      "epoch": 48.36,
+      "grad_norm": 0.910224974155426,
+      "learning_rate": 2e-05,
+      "loss": 0.03389906,
+      "step": 24180
+    },
+    {
+      "epoch": 48.362,
+      "grad_norm": 1.1010740995407104,
+      "learning_rate": 2e-05,
+      "loss": 0.04562913,
+      "step": 24181
+    },
+    {
+      "epoch": 48.364,
+      "grad_norm": 1.3951750993728638,
+      "learning_rate": 2e-05,
+      "loss": 0.04543884,
+      "step": 24182
+    },
+    {
+      "epoch": 48.366,
+      "grad_norm": 1.33402681350708,
+      "learning_rate": 2e-05,
+      "loss": 0.04668194,
+      "step": 24183
+    },
+    {
+      "epoch": 48.368,
+      "grad_norm": 1.3181812763214111,
+      "learning_rate": 2e-05,
+      "loss": 0.05848655,
+      "step": 24184
+    },
+    {
+      "epoch": 48.37,
+      "grad_norm": 1.1864538192749023,
+      "learning_rate": 2e-05,
+      "loss": 0.04627438,
+      "step": 24185
+    },
+    {
+      "epoch": 48.372,
+      "grad_norm": 1.0719053745269775,
+      "learning_rate": 2e-05,
+      "loss": 0.04401678,
+      "step": 24186
+    },
+    {
+      "epoch": 48.374,
+      "grad_norm": 1.2365782260894775,
+      "learning_rate": 2e-05,
+      "loss": 0.04953314,
+      "step": 24187
+    },
+    {
+      "epoch": 48.376,
+      "grad_norm": 1.362647533416748,
+      "learning_rate": 2e-05,
+      "loss": 0.03675044,
+      "step": 24188
+    },
+    {
+      "epoch": 48.378,
+      "grad_norm": 1.1618704795837402,
+      "learning_rate": 2e-05,
+      "loss": 0.05714417,
+      "step": 24189
+    },
+    {
+      "epoch": 48.38,
+      "grad_norm": 1.2545180320739746,
+      "learning_rate": 2e-05,
+      "loss": 0.04899006,
+      "step": 24190
+    },
+    {
+      "epoch": 48.382,
+      "grad_norm": 1.5738791227340698,
+      "learning_rate": 2e-05,
+      "loss": 0.06693644,
+      "step": 24191
+    },
+    {
+      "epoch": 48.384,
+      "grad_norm": 1.2538057565689087,
+      "learning_rate": 2e-05,
+      "loss": 0.04467027,
+      "step": 24192
+    },
+    {
+      "epoch": 48.386,
+      "grad_norm": 2.4022083282470703,
+      "learning_rate": 2e-05,
+      "loss": 0.0603236,
+      "step": 24193
+    },
+    {
+      "epoch": 48.388,
+      "grad_norm": 1.0010615587234497,
+      "learning_rate": 2e-05,
+      "loss": 0.03820972,
+      "step": 24194
+    },
+    {
+      "epoch": 48.39,
+      "grad_norm": 1.7481197118759155,
+      "learning_rate": 2e-05,
+      "loss": 0.06582922,
+      "step": 24195
+    },
+    {
+      "epoch": 48.392,
+      "grad_norm": 1.1138973236083984,
+      "learning_rate": 2e-05,
+      "loss": 0.05272717,
+      "step": 24196
+    },
+    {
+      "epoch": 48.394,
+      "grad_norm": 1.0726069211959839,
+      "learning_rate": 2e-05,
+      "loss": 0.05133822,
+      "step": 24197
+    },
+    {
+      "epoch": 48.396,
+      "grad_norm": 1.7698593139648438,
+      "learning_rate": 2e-05,
+      "loss": 0.05580845,
+      "step": 24198
+    },
+    {
+      "epoch": 48.398,
+      "grad_norm": 1.3947736024856567,
+      "learning_rate": 2e-05,
+      "loss": 0.05010434,
+      "step": 24199
+    },
+    {
+      "epoch": 48.4,
+      "grad_norm": 1.0563322305679321,
+      "learning_rate": 2e-05,
+      "loss": 0.03886336,
+      "step": 24200
+    },
+    {
+      "epoch": 48.402,
+      "grad_norm": 1.6317943334579468,
+      "learning_rate": 2e-05,
+      "loss": 0.06757972,
+      "step": 24201
+    },
+    {
+      "epoch": 48.404,
+      "grad_norm": 1.1341196298599243,
+      "learning_rate": 2e-05,
+      "loss": 0.04929399,
+      "step": 24202
+    },
+    {
+      "epoch": 48.406,
+      "grad_norm": 1.2951228618621826,
+      "learning_rate": 2e-05,
+      "loss": 0.04734764,
+      "step": 24203
+    },
+    {
+      "epoch": 48.408,
+      "grad_norm": 1.2703125476837158,
+      "learning_rate": 2e-05,
+      "loss": 0.0539843,
+      "step": 24204
+    },
+    {
+      "epoch": 48.41,
+      "grad_norm": 1.4762182235717773,
+      "learning_rate": 2e-05,
+      "loss": 0.05031619,
+      "step": 24205
+    },
+    {
+      "epoch": 48.412,
+      "grad_norm": 1.1297892332077026,
+      "learning_rate": 2e-05,
+      "loss": 0.05098326,
+      "step": 24206
+    },
+    {
+      "epoch": 48.414,
+      "grad_norm": 1.776755452156067,
+      "learning_rate": 2e-05,
+      "loss": 0.05509105,
+      "step": 24207
+    },
+    {
+      "epoch": 48.416,
+      "grad_norm": 1.4835596084594727,
+      "learning_rate": 2e-05,
+      "loss": 0.05865866,
+      "step": 24208
+    },
+    {
+      "epoch": 48.418,
+      "grad_norm": 1.1328264474868774,
+      "learning_rate": 2e-05,
+      "loss": 0.0568237,
+      "step": 24209
+    },
+    {
+      "epoch": 48.42,
+      "grad_norm": 1.0949770212173462,
+      "learning_rate": 2e-05,
+      "loss": 0.04025354,
+      "step": 24210
+    },
+    {
+      "epoch": 48.422,
+      "grad_norm": 1.2721264362335205,
+      "learning_rate": 2e-05,
+      "loss": 0.05412237,
+      "step": 24211
+    },
+    {
+      "epoch": 48.424,
+      "grad_norm": 1.6875686645507812,
+      "learning_rate": 2e-05,
+      "loss": 0.05133815,
+      "step": 24212
+    },
+    {
+      "epoch": 48.426,
+      "grad_norm": 1.1843007802963257,
+      "learning_rate": 2e-05,
+      "loss": 0.04539281,
+      "step": 24213
+    },
+    {
+      "epoch": 48.428,
+      "grad_norm": 1.1561273336410522,
+      "learning_rate": 2e-05,
+      "loss": 0.05123963,
+      "step": 24214
+    },
+    {
+      "epoch": 48.43,
+      "grad_norm": 1.2568432092666626,
+      "learning_rate": 2e-05,
+      "loss": 0.05854414,
+      "step": 24215
+    },
+    {
+      "epoch": 48.432,
+      "grad_norm": 1.2234379053115845,
+      "learning_rate": 2e-05,
+      "loss": 0.03763745,
+      "step": 24216
+    },
+    {
+      "epoch": 48.434,
+      "grad_norm": 1.1725529432296753,
+      "learning_rate": 2e-05,
+      "loss": 0.05705394,
+      "step": 24217
+    },
+    {
+      "epoch": 48.436,
+      "grad_norm": 1.598198652267456,
+      "learning_rate": 2e-05,
+      "loss": 0.05314296,
+      "step": 24218
+    },
+    {
+      "epoch": 48.438,
+      "grad_norm": 1.4097914695739746,
+      "learning_rate": 2e-05,
+      "loss": 0.04881332,
+      "step": 24219
+    },
+    {
+      "epoch": 48.44,
+      "grad_norm": 1.0265836715698242,
+      "learning_rate": 2e-05,
+      "loss": 0.03001934,
+      "step": 24220
+    },
+    {
+      "epoch": 48.442,
+      "grad_norm": 1.0836626291275024,
+      "learning_rate": 2e-05,
+      "loss": 0.04115789,
+      "step": 24221
+    },
+    {
+      "epoch": 48.444,
+      "grad_norm": 0.9527786374092102,
+      "learning_rate": 2e-05,
+      "loss": 0.04305189,
+      "step": 24222
+    },
+    {
+      "epoch": 48.446,
+      "grad_norm": 2.3282713890075684,
+      "learning_rate": 2e-05,
+      "loss": 0.04773811,
+      "step": 24223
+    },
+    {
+      "epoch": 48.448,
+      "grad_norm": 1.238843321800232,
+      "learning_rate": 2e-05,
+      "loss": 0.05873473,
+      "step": 24224
+    },
+    {
+      "epoch": 48.45,
+      "grad_norm": 1.050054669380188,
+      "learning_rate": 2e-05,
+      "loss": 0.04225109,
+      "step": 24225
+    },
+    {
+      "epoch": 48.452,
+      "grad_norm": 0.9493594765663147,
+      "learning_rate": 2e-05,
+      "loss": 0.03815629,
+      "step": 24226
+    },
+    {
+      "epoch": 48.454,
+      "grad_norm": 1.4076915979385376,
+      "learning_rate": 2e-05,
+      "loss": 0.0571268,
+      "step": 24227
+    },
+    {
+      "epoch": 48.456,
+      "grad_norm": 1.1200451850891113,
+      "learning_rate": 2e-05,
+      "loss": 0.04536561,
+      "step": 24228
+    },
+    {
+      "epoch": 48.458,
+      "grad_norm": 0.9729321002960205,
+      "learning_rate": 2e-05,
+      "loss": 0.04120147,
+      "step": 24229
+    },
+    {
+      "epoch": 48.46,
+      "grad_norm": 1.1512271165847778,
+      "learning_rate": 2e-05,
+      "loss": 0.04277606,
+      "step": 24230
+    },
+    {
+      "epoch": 48.462,
+      "grad_norm": 1.1026575565338135,
+      "learning_rate": 2e-05,
+      "loss": 0.0351891,
+      "step": 24231
+    },
+    {
+      "epoch": 48.464,
+      "grad_norm": 1.0310128927230835,
+      "learning_rate": 2e-05,
+      "loss": 0.03461772,
+      "step": 24232
+    },
+    {
+      "epoch": 48.466,
+      "grad_norm": 1.412767767906189,
+      "learning_rate": 2e-05,
+      "loss": 0.04195011,
+      "step": 24233
+    },
+    {
+      "epoch": 48.468,
+      "grad_norm": 1.1256314516067505,
+      "learning_rate": 2e-05,
+      "loss": 0.05557293,
+      "step": 24234
+    },
+    {
+      "epoch": 48.47,
+      "grad_norm": 1.1693962812423706,
+      "learning_rate": 2e-05,
+      "loss": 0.03926221,
+      "step": 24235
+    },
+    {
+      "epoch": 48.472,
+      "grad_norm": 1.1517102718353271,
+      "learning_rate": 2e-05,
+      "loss": 0.05275252,
+      "step": 24236
+    },
+    {
+      "epoch": 48.474,
+      "grad_norm": 1.3744908571243286,
+      "learning_rate": 2e-05,
+      "loss": 0.04760543,
+      "step": 24237
+    },
+    {
+      "epoch": 48.476,
+      "grad_norm": 1.1861525774002075,
+      "learning_rate": 2e-05,
+      "loss": 0.04502533,
+      "step": 24238
+    },
+    {
+      "epoch": 48.478,
+      "grad_norm": 1.4491164684295654,
+      "learning_rate": 2e-05,
+      "loss": 0.05307315,
+      "step": 24239
+    },
+    {
+      "epoch": 48.48,
+      "grad_norm": 1.0572928190231323,
+      "learning_rate": 2e-05,
+      "loss": 0.04607842,
+      "step": 24240
+    },
+    {
+      "epoch": 48.482,
+      "grad_norm": 1.7253468036651611,
+      "learning_rate": 2e-05,
+      "loss": 0.0420462,
+      "step": 24241
+    },
+    {
+      "epoch": 48.484,
+      "grad_norm": 1.1973940134048462,
+      "learning_rate": 2e-05,
+      "loss": 0.06035455,
+      "step": 24242
+    },
+    {
+      "epoch": 48.486,
+      "grad_norm": 1.123047947883606,
+      "learning_rate": 2e-05,
+      "loss": 0.04807914,
+      "step": 24243
+    },
+    {
+      "epoch": 48.488,
+      "grad_norm": 3.36896014213562,
+      "learning_rate": 2e-05,
+      "loss": 0.05288967,
+      "step": 24244
+    },
+    {
+      "epoch": 48.49,
+      "grad_norm": 1.3150179386138916,
+      "learning_rate": 2e-05,
+      "loss": 0.04113879,
+      "step": 24245
+    },
+    {
+      "epoch": 48.492,
+      "grad_norm": 0.998184084892273,
+      "learning_rate": 2e-05,
+      "loss": 0.03896046,
+      "step": 24246
+    },
+    {
+      "epoch": 48.494,
+      "grad_norm": 2.140963315963745,
+      "learning_rate": 2e-05,
+      "loss": 0.05801499,
+      "step": 24247
+    },
+    {
+      "epoch": 48.496,
+      "grad_norm": 0.9742385149002075,
+      "learning_rate": 2e-05,
+      "loss": 0.04312814,
+      "step": 24248
+    },
+    {
+      "epoch": 48.498,
+      "grad_norm": 0.9801925420761108,
+      "learning_rate": 2e-05,
+      "loss": 0.04314513,
+      "step": 24249
+    },
+    {
+      "epoch": 48.5,
+      "grad_norm": 1.167195200920105,
+      "learning_rate": 2e-05,
+      "loss": 0.04025427,
+      "step": 24250
+    },
+    {
+      "epoch": 48.502,
+      "grad_norm": 1.045845627784729,
+      "learning_rate": 2e-05,
+      "loss": 0.03268411,
+      "step": 24251
+    },
+    {
+      "epoch": 48.504,
+      "grad_norm": 1.4855949878692627,
+      "learning_rate": 2e-05,
+      "loss": 0.03678501,
+      "step": 24252
+    },
+    {
+      "epoch": 48.506,
+      "grad_norm": 1.0382455587387085,
+      "learning_rate": 2e-05,
+      "loss": 0.05245716,
+      "step": 24253
+    },
+    {
+      "epoch": 48.508,
+      "grad_norm": 1.0648964643478394,
+      "learning_rate": 2e-05,
+      "loss": 0.0379638,
+      "step": 24254
+    },
+    {
+      "epoch": 48.51,
+      "grad_norm": 1.1607509851455688,
+      "learning_rate": 2e-05,
+      "loss": 0.05345634,
+      "step": 24255
+    },
+    {
+      "epoch": 48.512,
+      "grad_norm": 1.2925974130630493,
+      "learning_rate": 2e-05,
+      "loss": 0.05352456,
+      "step": 24256
+    },
+    {
+      "epoch": 48.514,
+      "grad_norm": 1.1393349170684814,
+      "learning_rate": 2e-05,
+      "loss": 0.04305478,
+      "step": 24257
+    },
+    {
+      "epoch": 48.516,
+      "grad_norm": 1.041580319404602,
+      "learning_rate": 2e-05,
+      "loss": 0.05400862,
+      "step": 24258
+    },
+    {
+      "epoch": 48.518,
+      "grad_norm": 1.0252822637557983,
+      "learning_rate": 2e-05,
+      "loss": 0.0446092,
+      "step": 24259
+    },
+    {
+      "epoch": 48.52,
+      "grad_norm": 1.2504663467407227,
+      "learning_rate": 2e-05,
+      "loss": 0.06324365,
+      "step": 24260
+    },
+    {
+      "epoch": 48.522,
+      "grad_norm": 1.246679425239563,
+      "learning_rate": 2e-05,
+      "loss": 0.05985711,
+      "step": 24261
+    },
+    {
+      "epoch": 48.524,
+      "grad_norm": 1.0827921628952026,
+      "learning_rate": 2e-05,
+      "loss": 0.04559185,
+      "step": 24262
+    },
+    {
+      "epoch": 48.526,
+      "grad_norm": 0.9944784045219421,
+      "learning_rate": 2e-05,
+      "loss": 0.04224956,
+      "step": 24263
+    },
+    {
+      "epoch": 48.528,
+      "grad_norm": 1.6731486320495605,
+      "learning_rate": 2e-05,
+      "loss": 0.06704316,
+      "step": 24264
+    },
+    {
+      "epoch": 48.53,
+      "grad_norm": 1.1593559980392456,
+      "learning_rate": 2e-05,
+      "loss": 0.04145539,
+      "step": 24265
+    },
+    {
+      "epoch": 48.532,
+      "grad_norm": 1.5469125509262085,
+      "learning_rate": 2e-05,
+      "loss": 0.05441985,
+      "step": 24266
+    },
+    {
+      "epoch": 48.534,
+      "grad_norm": 1.191585898399353,
+      "learning_rate": 2e-05,
+      "loss": 0.04084273,
+      "step": 24267
+    },
+    {
+      "epoch": 48.536,
+      "grad_norm": 2.4968080520629883,
+      "learning_rate": 2e-05,
+      "loss": 0.05761244,
+      "step": 24268
+    },
+    {
+      "epoch": 48.538,
+      "grad_norm": 1.4555327892303467,
+      "learning_rate": 2e-05,
+      "loss": 0.05669381,
+      "step": 24269
+    },
+    {
+      "epoch": 48.54,
+      "grad_norm": 1.1853517293930054,
+      "learning_rate": 2e-05,
+      "loss": 0.03782331,
+      "step": 24270
+    },
+    {
+      "epoch": 48.542,
+      "grad_norm": 1.0957677364349365,
+      "learning_rate": 2e-05,
+      "loss": 0.04011827,
+      "step": 24271
+    },
+    {
+      "epoch": 48.544,
+      "grad_norm": 1.120521903038025,
+      "learning_rate": 2e-05,
+      "loss": 0.0431511,
+      "step": 24272
+    },
+    {
+      "epoch": 48.546,
+      "grad_norm": 1.3489925861358643,
+      "learning_rate": 2e-05,
+      "loss": 0.04053114,
+      "step": 24273
+    },
+    {
+      "epoch": 48.548,
+      "grad_norm": 1.153000831604004,
+      "learning_rate": 2e-05,
+      "loss": 0.05193951,
+      "step": 24274
+    },
+    {
+      "epoch": 48.55,
+      "grad_norm": 1.4672309160232544,
+      "learning_rate": 2e-05,
+      "loss": 0.0470688,
+      "step": 24275
+    },
+    {
+      "epoch": 48.552,
+      "grad_norm": 1.2560532093048096,
+      "learning_rate": 2e-05,
+      "loss": 0.04077678,
+      "step": 24276
+    },
+    {
+      "epoch": 48.554,
+      "grad_norm": 1.1438381671905518,
+      "learning_rate": 2e-05,
+      "loss": 0.04561079,
+      "step": 24277
+    },
+    {
+      "epoch": 48.556,
+      "grad_norm": 1.103325605392456,
+      "learning_rate": 2e-05,
+      "loss": 0.04266635,
+      "step": 24278
+    },
+    {
+      "epoch": 48.558,
+      "grad_norm": 1.2053292989730835,
+      "learning_rate": 2e-05,
+      "loss": 0.05609243,
+      "step": 24279
+    },
+    {
+      "epoch": 48.56,
+      "grad_norm": 1.2866277694702148,
+      "learning_rate": 2e-05,
+      "loss": 0.05822425,
+      "step": 24280
+    },
+    {
+      "epoch": 48.562,
+      "grad_norm": 1.3046430349349976,
+      "learning_rate": 2e-05,
+      "loss": 0.05633568,
+      "step": 24281
+    },
+    {
+      "epoch": 48.564,
+      "grad_norm": 1.165405035018921,
+      "learning_rate": 2e-05,
+      "loss": 0.02543902,
+      "step": 24282
+    },
+    {
+      "epoch": 48.566,
+      "grad_norm": 1.2302770614624023,
+      "learning_rate": 2e-05,
+      "loss": 0.04642507,
+      "step": 24283
+    },
+    {
+      "epoch": 48.568,
+      "grad_norm": 1.5563623905181885,
+      "learning_rate": 2e-05,
+      "loss": 0.05143239,
+      "step": 24284
+    },
+    {
+      "epoch": 48.57,
+      "grad_norm": 1.0007959604263306,
+      "learning_rate": 2e-05,
+      "loss": 0.03177904,
+      "step": 24285
+    },
+    {
+      "epoch": 48.572,
+      "grad_norm": 1.5403399467468262,
+      "learning_rate": 2e-05,
+      "loss": 0.0840653,
+      "step": 24286
+    },
+    {
+      "epoch": 48.574,
+      "grad_norm": 2.075601577758789,
+      "learning_rate": 2e-05,
+      "loss": 0.04922469,
+      "step": 24287
+    },
+    {
+      "epoch": 48.576,
+      "grad_norm": 1.155706524848938,
+      "learning_rate": 2e-05,
+      "loss": 0.03379347,
+      "step": 24288
+    },
+    {
+      "epoch": 48.578,
+      "grad_norm": 2.421046018600464,
+      "learning_rate": 2e-05,
+      "loss": 0.03684851,
+      "step": 24289
+    },
+    {
+      "epoch": 48.58,
+      "grad_norm": 1.1234990358352661,
+      "learning_rate": 2e-05,
+      "loss": 0.04452331,
+      "step": 24290
+    },
+    {
+      "epoch": 48.582,
+      "grad_norm": 1.228397011756897,
+      "learning_rate": 2e-05,
+      "loss": 0.05986349,
+      "step": 24291
+    },
+    {
+      "epoch": 48.584,
+      "grad_norm": 1.2848201990127563,
+      "learning_rate": 2e-05,
+      "loss": 0.05497922,
+      "step": 24292
+    },
+    {
+      "epoch": 48.586,
+      "grad_norm": 1.1021811962127686,
+      "learning_rate": 2e-05,
+      "loss": 0.04066538,
+      "step": 24293
+    },
+    {
+      "epoch": 48.588,
+      "grad_norm": 1.184058427810669,
+      "learning_rate": 2e-05,
+      "loss": 0.05837499,
+      "step": 24294
+    },
+    {
+      "epoch": 48.59,
+      "grad_norm": 1.4225780963897705,
+      "learning_rate": 2e-05,
+      "loss": 0.04438188,
+      "step": 24295
+    },
+    {
+      "epoch": 48.592,
+      "grad_norm": 1.1480562686920166,
+      "learning_rate": 2e-05,
+      "loss": 0.0400277,
+      "step": 24296
+    },
+    {
+      "epoch": 48.594,
+      "grad_norm": 1.323900580406189,
+      "learning_rate": 2e-05,
+      "loss": 0.05192615,
+      "step": 24297
+    },
+    {
+      "epoch": 48.596,
+      "grad_norm": 1.441277265548706,
+      "learning_rate": 2e-05,
+      "loss": 0.03372613,
+      "step": 24298
+    },
+    {
+      "epoch": 48.598,
+      "grad_norm": 1.0926908254623413,
+      "learning_rate": 2e-05,
+      "loss": 0.05107879,
+      "step": 24299
+    },
+    {
+      "epoch": 48.6,
+      "grad_norm": 1.2348512411117554,
+      "learning_rate": 2e-05,
+      "loss": 0.05852629,
+      "step": 24300
+    },
+    {
+      "epoch": 48.602,
+      "grad_norm": 1.182897925376892,
+      "learning_rate": 2e-05,
+      "loss": 0.06499533,
+      "step": 24301
+    },
+    {
+      "epoch": 48.604,
+      "grad_norm": 1.1605639457702637,
+      "learning_rate": 2e-05,
+      "loss": 0.05055974,
+      "step": 24302
+    },
+    {
+      "epoch": 48.606,
+      "grad_norm": 1.3561877012252808,
+      "learning_rate": 2e-05,
+      "loss": 0.05069784,
+      "step": 24303
+    },
+    {
+      "epoch": 48.608,
+      "grad_norm": 1.2737542390823364,
+      "learning_rate": 2e-05,
+      "loss": 0.04829124,
+      "step": 24304
+    },
+    {
+      "epoch": 48.61,
+      "grad_norm": 1.2459218502044678,
+      "learning_rate": 2e-05,
+      "loss": 0.057886,
+      "step": 24305
+    },
+    {
+      "epoch": 48.612,
+      "grad_norm": 1.154877781867981,
+      "learning_rate": 2e-05,
+      "loss": 0.03690616,
+      "step": 24306
+    },
+    {
+      "epoch": 48.614,
+      "grad_norm": 1.1142473220825195,
+      "learning_rate": 2e-05,
+      "loss": 0.04277994,
+      "step": 24307
+    },
+    {
+      "epoch": 48.616,
+      "grad_norm": 1.2507994174957275,
+      "learning_rate": 2e-05,
+      "loss": 0.05749039,
+      "step": 24308
+    },
+    {
+      "epoch": 48.618,
+      "grad_norm": 1.2597581148147583,
+      "learning_rate": 2e-05,
+      "loss": 0.04754964,
+      "step": 24309
+    },
+    {
+      "epoch": 48.62,
+      "grad_norm": 1.1352006196975708,
+      "learning_rate": 2e-05,
+      "loss": 0.04118899,
+      "step": 24310
+    },
+    {
+      "epoch": 48.622,
+      "grad_norm": 1.7616618871688843,
+      "learning_rate": 2e-05,
+      "loss": 0.0472442,
+      "step": 24311
+    },
+    {
+      "epoch": 48.624,
+      "grad_norm": 1.2920984029769897,
+      "learning_rate": 2e-05,
+      "loss": 0.05620179,
+      "step": 24312
+    },
+    {
+      "epoch": 48.626,
+      "grad_norm": 1.0155946016311646,
+      "learning_rate": 2e-05,
+      "loss": 0.03928847,
+      "step": 24313
+    },
+    {
+      "epoch": 48.628,
+      "grad_norm": 1.8907221555709839,
+      "learning_rate": 2e-05,
+      "loss": 0.05770548,
+      "step": 24314
+    },
+    {
+      "epoch": 48.63,
+      "grad_norm": 1.3525524139404297,
+      "learning_rate": 2e-05,
+      "loss": 0.05569984,
+      "step": 24315
+    },
+    {
+      "epoch": 48.632,
+      "grad_norm": 2.0137546062469482,
+      "learning_rate": 2e-05,
+      "loss": 0.04367228,
+      "step": 24316
+    },
+    {
+      "epoch": 48.634,
+      "grad_norm": 1.1896785497665405,
+      "learning_rate": 2e-05,
+      "loss": 0.0412126,
+      "step": 24317
+    },
+    {
+      "epoch": 48.636,
+      "grad_norm": 1.2872071266174316,
+      "learning_rate": 2e-05,
+      "loss": 0.05311878,
+      "step": 24318
+    },
+    {
+      "epoch": 48.638,
+      "grad_norm": 3.229105234146118,
+      "learning_rate": 2e-05,
+      "loss": 0.05653278,
+      "step": 24319
+    },
+    {
+      "epoch": 48.64,
+      "grad_norm": 1.131352186203003,
+      "learning_rate": 2e-05,
+      "loss": 0.04109538,
+      "step": 24320
+    },
+    {
+      "epoch": 48.642,
+      "grad_norm": 1.0072386264801025,
+      "learning_rate": 2e-05,
+      "loss": 0.04287854,
+      "step": 24321
+    },
+    {
+      "epoch": 48.644,
+      "grad_norm": 1.1246684789657593,
+      "learning_rate": 2e-05,
+      "loss": 0.04119401,
+      "step": 24322
+    },
+    {
+      "epoch": 48.646,
+      "grad_norm": 0.9696083664894104,
+      "learning_rate": 2e-05,
+      "loss": 0.03462346,
+      "step": 24323
+    },
+    {
+      "epoch": 48.648,
+      "grad_norm": 1.8664556741714478,
+      "learning_rate": 2e-05,
+      "loss": 0.0543664,
+      "step": 24324
+    },
+    {
+      "epoch": 48.65,
+      "grad_norm": 2.026254653930664,
+      "learning_rate": 2e-05,
+      "loss": 0.0509354,
+      "step": 24325
+    },
+    {
+      "epoch": 48.652,
+      "grad_norm": 1.2250194549560547,
+      "learning_rate": 2e-05,
+      "loss": 0.06017506,
+      "step": 24326
+    },
+    {
+      "epoch": 48.654,
+      "grad_norm": 1.2674254179000854,
+      "learning_rate": 2e-05,
+      "loss": 0.04169001,
+      "step": 24327
+    },
+    {
+      "epoch": 48.656,
+      "grad_norm": 1.2122572660446167,
+      "learning_rate": 2e-05,
+      "loss": 0.05367371,
+      "step": 24328
+    },
+    {
+      "epoch": 48.658,
+      "grad_norm": 1.4815667867660522,
+      "learning_rate": 2e-05,
+      "loss": 0.07224436,
+      "step": 24329
+    },
+    {
+      "epoch": 48.66,
+      "grad_norm": 1.290479302406311,
+      "learning_rate": 2e-05,
+      "loss": 0.04883266,
+      "step": 24330
+    },
+    {
+      "epoch": 48.662,
+      "grad_norm": 1.1489940881729126,
+      "learning_rate": 2e-05,
+      "loss": 0.0474559,
+      "step": 24331
+    },
+    {
+      "epoch": 48.664,
+      "grad_norm": 1.302720069885254,
+      "learning_rate": 2e-05,
+      "loss": 0.05755413,
+      "step": 24332
+    },
+    {
+      "epoch": 48.666,
+      "grad_norm": 0.9729812741279602,
+      "learning_rate": 2e-05,
+      "loss": 0.0363071,
+      "step": 24333
+    },
+    {
+      "epoch": 48.668,
+      "grad_norm": 1.3361284732818604,
+      "learning_rate": 2e-05,
+      "loss": 0.04093036,
+      "step": 24334
+    },
+    {
+      "epoch": 48.67,
+      "grad_norm": 1.046409010887146,
+      "learning_rate": 2e-05,
+      "loss": 0.03625423,
+      "step": 24335
+    },
+    {
+      "epoch": 48.672,
+      "grad_norm": 1.158137321472168,
+      "learning_rate": 2e-05,
+      "loss": 0.05011062,
+      "step": 24336
+    },
+    {
+      "epoch": 48.674,
+      "grad_norm": 1.1454992294311523,
+      "learning_rate": 2e-05,
+      "loss": 0.04545939,
+      "step": 24337
+    },
+    {
+      "epoch": 48.676,
+      "grad_norm": 1.2289396524429321,
+      "learning_rate": 2e-05,
+      "loss": 0.06850711,
+      "step": 24338
+    },
+    {
+      "epoch": 48.678,
+      "grad_norm": 2.0199902057647705,
+      "learning_rate": 2e-05,
+      "loss": 0.04739082,
+      "step": 24339
+    },
+    {
+      "epoch": 48.68,
+      "grad_norm": 1.1302142143249512,
+      "learning_rate": 2e-05,
+      "loss": 0.0431182,
+      "step": 24340
+    },
+    {
+      "epoch": 48.682,
+      "grad_norm": 1.3505676984786987,
+      "learning_rate": 2e-05,
+      "loss": 0.04874783,
+      "step": 24341
+    },
+    {
+      "epoch": 48.684,
+      "grad_norm": 0.9121742248535156,
+      "learning_rate": 2e-05,
+      "loss": 0.02643291,
+      "step": 24342
+    },
+    {
+      "epoch": 48.686,
+      "grad_norm": 1.3847429752349854,
+      "learning_rate": 2e-05,
+      "loss": 0.04750867,
+      "step": 24343
+    },
+    {
+      "epoch": 48.688,
+      "grad_norm": 1.1395158767700195,
+      "learning_rate": 2e-05,
+      "loss": 0.04924957,
+      "step": 24344
+    },
+    {
+      "epoch": 48.69,
+      "grad_norm": 1.0794190168380737,
+      "learning_rate": 2e-05,
+      "loss": 0.03787911,
+      "step": 24345
+    },
+    {
+      "epoch": 48.692,
+      "grad_norm": 1.1486942768096924,
+      "learning_rate": 2e-05,
+      "loss": 0.04261864,
+      "step": 24346
+    },
+    {
+      "epoch": 48.694,
+      "grad_norm": 1.1748255491256714,
+      "learning_rate": 2e-05,
+      "loss": 0.04672168,
+      "step": 24347
+    },
+    {
+      "epoch": 48.696,
+      "grad_norm": 1.1263988018035889,
+      "learning_rate": 2e-05,
+      "loss": 0.04133555,
+      "step": 24348
+    },
+    {
+      "epoch": 48.698,
+      "grad_norm": 1.2288318872451782,
+      "learning_rate": 2e-05,
+      "loss": 0.04745194,
+      "step": 24349
+    },
+    {
+      "epoch": 48.7,
+      "grad_norm": 0.9768017530441284,
+      "learning_rate": 2e-05,
+      "loss": 0.04292785,
+      "step": 24350
+    },
+    {
+      "epoch": 48.702,
+      "grad_norm": 1.0999313592910767,
+      "learning_rate": 2e-05,
+      "loss": 0.05881019,
+      "step": 24351
+    },
+    {
+      "epoch": 48.704,
+      "grad_norm": 1.878710389137268,
+      "learning_rate": 2e-05,
+      "loss": 0.05866309,
+      "step": 24352
+    },
+    {
+      "epoch": 48.706,
+      "grad_norm": 1.63356614112854,
+      "learning_rate": 2e-05,
+      "loss": 0.05248346,
+      "step": 24353
+    },
+    {
+      "epoch": 48.708,
+      "grad_norm": 1.2071478366851807,
+      "learning_rate": 2e-05,
+      "loss": 0.05778679,
+      "step": 24354
+    },
+    {
+      "epoch": 48.71,
+      "grad_norm": 1.188579797744751,
+      "learning_rate": 2e-05,
+      "loss": 0.05696705,
+      "step": 24355
+    },
+    {
+      "epoch": 48.712,
+      "grad_norm": 0.9021721482276917,
+      "learning_rate": 2e-05,
+      "loss": 0.02676656,
+      "step": 24356
+    },
+    {
+      "epoch": 48.714,
+      "grad_norm": 1.446212887763977,
+      "learning_rate": 2e-05,
+      "loss": 0.04261808,
+      "step": 24357
+    },
+    {
+      "epoch": 48.716,
+      "grad_norm": 0.975503146648407,
+      "learning_rate": 2e-05,
+      "loss": 0.03775565,
+      "step": 24358
+    },
+    {
+      "epoch": 48.718,
+      "grad_norm": 1.1201655864715576,
+      "learning_rate": 2e-05,
+      "loss": 0.04218309,
+      "step": 24359
+    },
+    {
+      "epoch": 48.72,
+      "grad_norm": 1.2044806480407715,
+      "learning_rate": 2e-05,
+      "loss": 0.04980427,
+      "step": 24360
+    },
+    {
+      "epoch": 48.722,
+      "grad_norm": 1.0975890159606934,
+      "learning_rate": 2e-05,
+      "loss": 0.04094896,
+      "step": 24361
+    },
+    {
+      "epoch": 48.724,
+      "grad_norm": 1.5695173740386963,
+      "learning_rate": 2e-05,
+      "loss": 0.04596317,
+      "step": 24362
+    },
+    {
+      "epoch": 48.726,
+      "grad_norm": 1.0056501626968384,
+      "learning_rate": 2e-05,
+      "loss": 0.04648286,
+      "step": 24363
+    },
+    {
+      "epoch": 48.728,
+      "grad_norm": 0.9918159246444702,
+      "learning_rate": 2e-05,
+      "loss": 0.03798754,
+      "step": 24364
+    },
+    {
+      "epoch": 48.73,
+      "grad_norm": 1.287979245185852,
+      "learning_rate": 2e-05,
+      "loss": 0.05667548,
+      "step": 24365
+    },
+    {
+      "epoch": 48.732,
+      "grad_norm": 1.1386852264404297,
+      "learning_rate": 2e-05,
+      "loss": 0.05278626,
+      "step": 24366
+    },
+    {
+      "epoch": 48.734,
+      "grad_norm": 0.996176540851593,
+      "learning_rate": 2e-05,
+      "loss": 0.05246001,
+      "step": 24367
+    },
+    {
+      "epoch": 48.736,
+      "grad_norm": 2.319737434387207,
+      "learning_rate": 2e-05,
+      "loss": 0.0373604,
+      "step": 24368
+    },
+    {
+      "epoch": 48.738,
+      "grad_norm": 1.7545973062515259,
+      "learning_rate": 2e-05,
+      "loss": 0.04798623,
+      "step": 24369
+    },
+    {
+      "epoch": 48.74,
+      "grad_norm": 1.3103057146072388,
+      "learning_rate": 2e-05,
+      "loss": 0.05369485,
+      "step": 24370
+    },
+    {
+      "epoch": 48.742,
+      "grad_norm": 1.1758966445922852,
+      "learning_rate": 2e-05,
+      "loss": 0.05795531,
+      "step": 24371
+    },
+    {
+      "epoch": 48.744,
+      "grad_norm": 0.9001947641372681,
+      "learning_rate": 2e-05,
+      "loss": 0.02734162,
+      "step": 24372
+    },
+    {
+      "epoch": 48.746,
+      "grad_norm": 1.0777047872543335,
+      "learning_rate": 2e-05,
+      "loss": 0.03581995,
+      "step": 24373
+    },
+    {
+      "epoch": 48.748,
+      "grad_norm": 1.3073005676269531,
+      "learning_rate": 2e-05,
+      "loss": 0.04707327,
+      "step": 24374
+    },
+    {
+      "epoch": 48.75,
+      "grad_norm": 0.9895936250686646,
+      "learning_rate": 2e-05,
+      "loss": 0.02827165,
+      "step": 24375
+    },
+    {
+      "epoch": 48.752,
+      "grad_norm": 1.1249933242797852,
+      "learning_rate": 2e-05,
+      "loss": 0.04632932,
+      "step": 24376
+    },
+    {
+      "epoch": 48.754,
+      "grad_norm": 1.274895191192627,
+      "learning_rate": 2e-05,
+      "loss": 0.04302908,
+      "step": 24377
+    },
+    {
+      "epoch": 48.756,
+      "grad_norm": 1.0535324811935425,
+      "learning_rate": 2e-05,
+      "loss": 0.03941915,
+      "step": 24378
+    },
+    {
+      "epoch": 48.758,
+      "grad_norm": 1.3476455211639404,
+      "learning_rate": 2e-05,
+      "loss": 0.05376494,
+      "step": 24379
+    },
+    {
+      "epoch": 48.76,
+      "grad_norm": 1.2379733324050903,
+      "learning_rate": 2e-05,
+      "loss": 0.05499609,
+      "step": 24380
+    },
+    {
+      "epoch": 48.762,
+      "grad_norm": 1.1249278783798218,
+      "learning_rate": 2e-05,
+      "loss": 0.03757162,
+      "step": 24381
+    },
+    {
+      "epoch": 48.764,
+      "grad_norm": 2.2356626987457275,
+      "learning_rate": 2e-05,
+      "loss": 0.0576737,
+      "step": 24382
+    },
+    {
+      "epoch": 48.766,
+      "grad_norm": 1.1273705959320068,
+      "learning_rate": 2e-05,
+      "loss": 0.05674585,
+      "step": 24383
+    },
+    {
+      "epoch": 48.768,
+      "grad_norm": 1.3228777647018433,
+      "learning_rate": 2e-05,
+      "loss": 0.05507819,
+      "step": 24384
+    },
+    {
+      "epoch": 48.77,
+      "grad_norm": 1.3240513801574707,
+      "learning_rate": 2e-05,
+      "loss": 0.05233713,
+      "step": 24385
+    },
+    {
+      "epoch": 48.772,
+      "grad_norm": 1.2595722675323486,
+      "learning_rate": 2e-05,
+      "loss": 0.05647445,
+      "step": 24386
+    },
+    {
+      "epoch": 48.774,
+      "grad_norm": 1.1670889854431152,
+      "learning_rate": 2e-05,
+      "loss": 0.0378773,
+      "step": 24387
+    },
+    {
+      "epoch": 48.776,
+      "grad_norm": 0.9020729064941406,
+      "learning_rate": 2e-05,
+      "loss": 0.03322694,
+      "step": 24388
+    },
+    {
+      "epoch": 48.778,
+      "grad_norm": 1.0231337547302246,
+      "learning_rate": 2e-05,
+      "loss": 0.03719312,
+      "step": 24389
+    },
+    {
+      "epoch": 48.78,
+      "grad_norm": 1.0719740390777588,
+      "learning_rate": 2e-05,
+      "loss": 0.04352896,
+      "step": 24390
+    },
+    {
+      "epoch": 48.782,
+      "grad_norm": 1.1282035112380981,
+      "learning_rate": 2e-05,
+      "loss": 0.05233166,
+      "step": 24391
+    },
+    {
+      "epoch": 48.784,
+      "grad_norm": 1.2852978706359863,
+      "learning_rate": 2e-05,
+      "loss": 0.05125814,
+      "step": 24392
+    },
+    {
+      "epoch": 48.786,
+      "grad_norm": 1.1023353338241577,
+      "learning_rate": 2e-05,
+      "loss": 0.04840891,
+      "step": 24393
+    },
+    {
+      "epoch": 48.788,
+      "grad_norm": 1.5600016117095947,
+      "learning_rate": 2e-05,
+      "loss": 0.04811985,
+      "step": 24394
+    },
+    {
+      "epoch": 48.79,
+      "grad_norm": 1.1673002243041992,
+      "learning_rate": 2e-05,
+      "loss": 0.04379809,
+      "step": 24395
+    },
+    {
+      "epoch": 48.792,
+      "grad_norm": 1.3947582244873047,
+      "learning_rate": 2e-05,
+      "loss": 0.05733298,
+      "step": 24396
+    },
+    {
+      "epoch": 48.794,
+      "grad_norm": 1.2544630765914917,
+      "learning_rate": 2e-05,
+      "loss": 0.06692179,
+      "step": 24397
+    },
+    {
+      "epoch": 48.796,
+      "grad_norm": 1.9167871475219727,
+      "learning_rate": 2e-05,
+      "loss": 0.06020879,
+      "step": 24398
+    },
+    {
+      "epoch": 48.798,
+      "grad_norm": 1.5168108940124512,
+      "learning_rate": 2e-05,
+      "loss": 0.04920478,
+      "step": 24399
+    },
+    {
+      "epoch": 48.8,
+      "grad_norm": 1.047117829322815,
+      "learning_rate": 2e-05,
+      "loss": 0.03645068,
+      "step": 24400
+    },
+    {
+      "epoch": 48.802,
+      "grad_norm": 1.2618528604507446,
+      "learning_rate": 2e-05,
+      "loss": 0.05060851,
+      "step": 24401
+    },
+    {
+      "epoch": 48.804,
+      "grad_norm": 1.0257928371429443,
+      "learning_rate": 2e-05,
+      "loss": 0.04926369,
+      "step": 24402
+    },
+    {
+      "epoch": 48.806,
+      "grad_norm": 2.2728285789489746,
+      "learning_rate": 2e-05,
+      "loss": 0.05713367,
+      "step": 24403
+    },
+    {
+      "epoch": 48.808,
+      "grad_norm": 1.2449434995651245,
+      "learning_rate": 2e-05,
+      "loss": 0.05508956,
+      "step": 24404
+    },
+    {
+      "epoch": 48.81,
+      "grad_norm": 1.1658732891082764,
+      "learning_rate": 2e-05,
+      "loss": 0.05041445,
+      "step": 24405
+    },
+    {
+      "epoch": 48.812,
+      "grad_norm": 1.206467866897583,
+      "learning_rate": 2e-05,
+      "loss": 0.06293423,
+      "step": 24406
+    },
+    {
+      "epoch": 48.814,
+      "grad_norm": 1.246690034866333,
+      "learning_rate": 2e-05,
+      "loss": 0.05447911,
+      "step": 24407
+    },
+    {
+      "epoch": 48.816,
+      "grad_norm": 1.204403042793274,
+      "learning_rate": 2e-05,
+      "loss": 0.05379122,
+      "step": 24408
+    },
+    {
+      "epoch": 48.818,
+      "grad_norm": 1.2013062238693237,
+      "learning_rate": 2e-05,
+      "loss": 0.05955196,
+      "step": 24409
+    },
+    {
+      "epoch": 48.82,
+      "grad_norm": 1.0979270935058594,
+      "learning_rate": 2e-05,
+      "loss": 0.0413815,
+      "step": 24410
+    },
+    {
+      "epoch": 48.822,
+      "grad_norm": 1.224603295326233,
+      "learning_rate": 2e-05,
+      "loss": 0.05617508,
+      "step": 24411
+    },
+    {
+      "epoch": 48.824,
+      "grad_norm": 1.262998342514038,
+      "learning_rate": 2e-05,
+      "loss": 0.04537391,
+      "step": 24412
+    },
+    {
+      "epoch": 48.826,
+      "grad_norm": 1.1487383842468262,
+      "learning_rate": 2e-05,
+      "loss": 0.03556521,
+      "step": 24413
+    },
+    {
+      "epoch": 48.828,
+      "grad_norm": 0.9703657031059265,
+      "learning_rate": 2e-05,
+      "loss": 0.03395029,
+      "step": 24414
+    },
+    {
+      "epoch": 48.83,
+      "grad_norm": 1.5587289333343506,
+      "learning_rate": 2e-05,
+      "loss": 0.0705722,
+      "step": 24415
+    },
+    {
+      "epoch": 48.832,
+      "grad_norm": 1.0362234115600586,
+      "learning_rate": 2e-05,
+      "loss": 0.04194809,
+      "step": 24416
+    },
+    {
+      "epoch": 48.834,
+      "grad_norm": 1.3099713325500488,
+      "learning_rate": 2e-05,
+      "loss": 0.05039264,
+      "step": 24417
+    },
+    {
+      "epoch": 48.836,
+      "grad_norm": 2.788674831390381,
+      "learning_rate": 2e-05,
+      "loss": 0.05012584,
+      "step": 24418
+    },
+    {
+      "epoch": 48.838,
+      "grad_norm": 1.322467565536499,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 24419
+    },
+    {
+      "epoch": 48.84,
+      "grad_norm": 1.230636477470398,
+      "learning_rate": 2e-05,
+      "loss": 0.03692626,
+      "step": 24420
+    },
+    {
+      "epoch": 48.842,
+      "grad_norm": 1.2802784442901611,
+      "learning_rate": 2e-05,
+      "loss": 0.05909276,
+      "step": 24421
+    },
+    {
+      "epoch": 48.844,
+      "grad_norm": 1.3420884609222412,
+      "learning_rate": 2e-05,
+      "loss": 0.0529298,
+      "step": 24422
+    },
+    {
+      "epoch": 48.846,
+      "grad_norm": 1.0645166635513306,
+      "learning_rate": 2e-05,
+      "loss": 0.04237661,
+      "step": 24423
+    },
+    {
+      "epoch": 48.848,
+      "grad_norm": 1.489173173904419,
+      "learning_rate": 2e-05,
+      "loss": 0.04594196,
+      "step": 24424
+    },
+    {
+      "epoch": 48.85,
+      "grad_norm": 1.9845540523529053,
+      "learning_rate": 2e-05,
+      "loss": 0.06666305,
+      "step": 24425
+    },
+    {
+      "epoch": 48.852,
+      "grad_norm": 1.2646132707595825,
+      "learning_rate": 2e-05,
+      "loss": 0.06317861,
+      "step": 24426
+    },
+    {
+      "epoch": 48.854,
+      "grad_norm": 1.1693730354309082,
+      "learning_rate": 2e-05,
+      "loss": 0.03957154,
+      "step": 24427
+    },
+    {
+      "epoch": 48.856,
+      "grad_norm": 1.139262080192566,
+      "learning_rate": 2e-05,
+      "loss": 0.03852566,
+      "step": 24428
+    },
+    {
+      "epoch": 48.858,
+      "grad_norm": 1.340086579322815,
+      "learning_rate": 2e-05,
+      "loss": 0.06382296,
+      "step": 24429
+    },
+    {
+      "epoch": 48.86,
+      "grad_norm": 1.404128909111023,
+      "learning_rate": 2e-05,
+      "loss": 0.06833863,
+      "step": 24430
+    },
+    {
+      "epoch": 48.862,
+      "grad_norm": 1.1025450229644775,
+      "learning_rate": 2e-05,
+      "loss": 0.05407901,
+      "step": 24431
+    },
+    {
+      "epoch": 48.864,
+      "grad_norm": 1.1320126056671143,
+      "learning_rate": 2e-05,
+      "loss": 0.04277468,
+      "step": 24432
+    },
+    {
+      "epoch": 48.866,
+      "grad_norm": 1.192533016204834,
+      "learning_rate": 2e-05,
+      "loss": 0.04574575,
+      "step": 24433
+    },
+    {
+      "epoch": 48.868,
+      "grad_norm": 5.2781267166137695,
+      "learning_rate": 2e-05,
+      "loss": 0.06549813,
+      "step": 24434
+    },
+    {
+      "epoch": 48.87,
+      "grad_norm": 1.2239826917648315,
+      "learning_rate": 2e-05,
+      "loss": 0.05279471,
+      "step": 24435
+    },
+    {
+      "epoch": 48.872,
+      "grad_norm": 1.0941675901412964,
+      "learning_rate": 2e-05,
+      "loss": 0.04686259,
+      "step": 24436
+    },
+    {
+      "epoch": 48.874,
+      "grad_norm": 1.5225481986999512,
+      "learning_rate": 2e-05,
+      "loss": 0.05451784,
+      "step": 24437
+    },
+    {
+      "epoch": 48.876,
+      "grad_norm": 1.0985329151153564,
+      "learning_rate": 2e-05,
+      "loss": 0.03912208,
+      "step": 24438
+    },
+    {
+      "epoch": 48.878,
+      "grad_norm": 0.9841903448104858,
+      "learning_rate": 2e-05,
+      "loss": 0.03415463,
+      "step": 24439
+    },
+    {
+      "epoch": 48.88,
+      "grad_norm": 1.519121527671814,
+      "learning_rate": 2e-05,
+      "loss": 0.05647705,
+      "step": 24440
+    },
+    {
+      "epoch": 48.882,
+      "grad_norm": 1.738462209701538,
+      "learning_rate": 2e-05,
+      "loss": 0.05705027,
+      "step": 24441
+    },
+    {
+      "epoch": 48.884,
+      "grad_norm": 1.2349008321762085,
+      "learning_rate": 2e-05,
+      "loss": 0.06694788,
+      "step": 24442
+    },
+    {
+      "epoch": 48.886,
+      "grad_norm": 1.1472513675689697,
+      "learning_rate": 2e-05,
+      "loss": 0.04741376,
+      "step": 24443
+    },
+    {
+      "epoch": 48.888,
+      "grad_norm": 1.7409826517105103,
+      "learning_rate": 2e-05,
+      "loss": 0.04635547,
+      "step": 24444
+    },
+    {
+      "epoch": 48.89,
+      "grad_norm": 1.507589340209961,
+      "learning_rate": 2e-05,
+      "loss": 0.04607899,
+      "step": 24445
+    },
+    {
+      "epoch": 48.892,
+      "grad_norm": 2.211873769760132,
+      "learning_rate": 2e-05,
+      "loss": 0.05145214,
+      "step": 24446
+    },
+    {
+      "epoch": 48.894,
+      "grad_norm": 1.286216139793396,
+      "learning_rate": 2e-05,
+      "loss": 0.06177491,
+      "step": 24447
+    },
+    {
+      "epoch": 48.896,
+      "grad_norm": 1.2425750494003296,
+      "learning_rate": 2e-05,
+      "loss": 0.05383733,
+      "step": 24448
+    },
+    {
+      "epoch": 48.898,
+      "grad_norm": 1.189475655555725,
+      "learning_rate": 2e-05,
+      "loss": 0.05400949,
+      "step": 24449
+    },
+    {
+      "epoch": 48.9,
+      "grad_norm": 1.3408055305480957,
+      "learning_rate": 2e-05,
+      "loss": 0.03949302,
+      "step": 24450
+    },
+    {
+      "epoch": 48.902,
+      "grad_norm": 2.6601455211639404,
+      "learning_rate": 2e-05,
+      "loss": 0.06074892,
+      "step": 24451
+    },
+    {
+      "epoch": 48.904,
+      "grad_norm": 0.9952139854431152,
+      "learning_rate": 2e-05,
+      "loss": 0.04453053,
+      "step": 24452
+    },
+    {
+      "epoch": 48.906,
+      "grad_norm": 1.4619004726409912,
+      "learning_rate": 2e-05,
+      "loss": 0.04188272,
+      "step": 24453
+    },
+    {
+      "epoch": 48.908,
+      "grad_norm": 1.2014166116714478,
+      "learning_rate": 2e-05,
+      "loss": 0.04575384,
+      "step": 24454
+    },
+    {
+      "epoch": 48.91,
+      "grad_norm": 2.021064043045044,
+      "learning_rate": 2e-05,
+      "loss": 0.04514164,
+      "step": 24455
+    },
+    {
+      "epoch": 48.912,
+      "grad_norm": 1.2506054639816284,
+      "learning_rate": 2e-05,
+      "loss": 0.0620794,
+      "step": 24456
+    },
+    {
+      "epoch": 48.914,
+      "grad_norm": 1.2398749589920044,
+      "learning_rate": 2e-05,
+      "loss": 0.04733563,
+      "step": 24457
+    },
+    {
+      "epoch": 48.916,
+      "grad_norm": 1.095046043395996,
+      "learning_rate": 2e-05,
+      "loss": 0.04853748,
+      "step": 24458
+    },
+    {
+      "epoch": 48.918,
+      "grad_norm": 2.492252826690674,
+      "learning_rate": 2e-05,
+      "loss": 0.04443378,
+      "step": 24459
+    },
+    {
+      "epoch": 48.92,
+      "grad_norm": 1.4789276123046875,
+      "learning_rate": 2e-05,
+      "loss": 0.04718126,
+      "step": 24460
+    },
+    {
+      "epoch": 48.922,
+      "grad_norm": 1.7125012874603271,
+      "learning_rate": 2e-05,
+      "loss": 0.03764442,
+      "step": 24461
+    },
+    {
+      "epoch": 48.924,
+      "grad_norm": 1.5009963512420654,
+      "learning_rate": 2e-05,
+      "loss": 0.02679938,
+      "step": 24462
+    },
+    {
+      "epoch": 48.926,
+      "grad_norm": 1.1632426977157593,
+      "learning_rate": 2e-05,
+      "loss": 0.04027367,
+      "step": 24463
+    },
+    {
+      "epoch": 48.928,
+      "grad_norm": 1.0739227533340454,
+      "learning_rate": 2e-05,
+      "loss": 0.04538012,
+      "step": 24464
+    },
+    {
+      "epoch": 48.93,
+      "grad_norm": 1.1324819326400757,
+      "learning_rate": 2e-05,
+      "loss": 0.05008095,
+      "step": 24465
+    },
+    {
+      "epoch": 48.932,
+      "grad_norm": 1.2299511432647705,
+      "learning_rate": 2e-05,
+      "loss": 0.03868235,
+      "step": 24466
+    },
+    {
+      "epoch": 48.934,
+      "grad_norm": 1.457084059715271,
+      "learning_rate": 2e-05,
+      "loss": 0.06458244,
+      "step": 24467
+    },
+    {
+      "epoch": 48.936,
+      "grad_norm": 1.4329404830932617,
+      "learning_rate": 2e-05,
+      "loss": 0.06372182,
+      "step": 24468
+    },
+    {
+      "epoch": 48.938,
+      "grad_norm": 1.0780876874923706,
+      "learning_rate": 2e-05,
+      "loss": 0.04435993,
+      "step": 24469
+    },
+    {
+      "epoch": 48.94,
+      "grad_norm": 1.6297838687896729,
+      "learning_rate": 2e-05,
+      "loss": 0.05016711,
+      "step": 24470
+    },
+    {
+      "epoch": 48.942,
+      "grad_norm": 1.1690902709960938,
+      "learning_rate": 2e-05,
+      "loss": 0.06398173,
+      "step": 24471
+    },
+    {
+      "epoch": 48.944,
+      "grad_norm": 1.1354949474334717,
+      "learning_rate": 2e-05,
+      "loss": 0.04506849,
+      "step": 24472
+    },
+    {
+      "epoch": 48.946,
+      "grad_norm": 1.3150396347045898,
+      "learning_rate": 2e-05,
+      "loss": 0.06466317,
+      "step": 24473
+    },
+    {
+      "epoch": 48.948,
+      "grad_norm": 1.104675531387329,
+      "learning_rate": 2e-05,
+      "loss": 0.0407197,
+      "step": 24474
+    },
+    {
+      "epoch": 48.95,
+      "grad_norm": 1.0967339277267456,
+      "learning_rate": 2e-05,
+      "loss": 0.04545303,
+      "step": 24475
+    },
+    {
+      "epoch": 48.952,
+      "grad_norm": 1.0893007516860962,
+      "learning_rate": 2e-05,
+      "loss": 0.05033952,
+      "step": 24476
+    },
+    {
+      "epoch": 48.954,
+      "grad_norm": 1.1199082136154175,
+      "learning_rate": 2e-05,
+      "loss": 0.04963399,
+      "step": 24477
+    },
+    {
+      "epoch": 48.956,
+      "grad_norm": 1.1054890155792236,
+      "learning_rate": 2e-05,
+      "loss": 0.04332212,
+      "step": 24478
+    },
+    {
+      "epoch": 48.958,
+      "grad_norm": 2.2792885303497314,
+      "learning_rate": 2e-05,
+      "loss": 0.05663423,
+      "step": 24479
+    },
+    {
+      "epoch": 48.96,
+      "grad_norm": 1.1706188917160034,
+      "learning_rate": 2e-05,
+      "loss": 0.04888928,
+      "step": 24480
+    },
+    {
+      "epoch": 48.962,
+      "grad_norm": 2.348127841949463,
+      "learning_rate": 2e-05,
+      "loss": 0.05282357,
+      "step": 24481
+    },
+    {
+      "epoch": 48.964,
+      "grad_norm": 1.2429864406585693,
+      "learning_rate": 2e-05,
+      "loss": 0.05059104,
+      "step": 24482
+    },
+    {
+      "epoch": 48.966,
+      "grad_norm": 1.1958144903182983,
+      "learning_rate": 2e-05,
+      "loss": 0.03917575,
+      "step": 24483
+    },
+    {
+      "epoch": 48.968,
+      "grad_norm": 1.0921530723571777,
+      "learning_rate": 2e-05,
+      "loss": 0.03745528,
+      "step": 24484
+    },
+    {
+      "epoch": 48.97,
+      "grad_norm": 1.7767388820648193,
+      "learning_rate": 2e-05,
+      "loss": 0.06159507,
+      "step": 24485
+    },
+    {
+      "epoch": 48.972,
+      "grad_norm": 1.1890246868133545,
+      "learning_rate": 2e-05,
+      "loss": 0.04934237,
+      "step": 24486
+    },
+    {
+      "epoch": 48.974,
+      "grad_norm": 1.616060495376587,
+      "learning_rate": 2e-05,
+      "loss": 0.04842779,
+      "step": 24487
+    },
+    {
+      "epoch": 48.976,
+      "grad_norm": 2.0892298221588135,
+      "learning_rate": 2e-05,
+      "loss": 0.03829409,
+      "step": 24488
+    },
+    {
+      "epoch": 48.978,
+      "grad_norm": 1.2111293077468872,
+      "learning_rate": 2e-05,
+      "loss": 0.05161772,
+      "step": 24489
+    },
+    {
+      "epoch": 48.98,
+      "grad_norm": 1.258104920387268,
+      "learning_rate": 2e-05,
+      "loss": 0.04110197,
+      "step": 24490
+    },
+    {
+      "epoch": 48.982,
+      "grad_norm": 1.1039538383483887,
+      "learning_rate": 2e-05,
+      "loss": 0.03855299,
+      "step": 24491
+    },
+    {
+      "epoch": 48.984,
+      "grad_norm": 1.0499625205993652,
+      "learning_rate": 2e-05,
+      "loss": 0.04614359,
+      "step": 24492
+    },
+    {
+      "epoch": 48.986,
+      "grad_norm": 1.2007472515106201,
+      "learning_rate": 2e-05,
+      "loss": 0.03996876,
+      "step": 24493
+    },
+    {
+      "epoch": 48.988,
+      "grad_norm": 1.2354191541671753,
+      "learning_rate": 2e-05,
+      "loss": 0.04676054,
+      "step": 24494
+    },
+    {
+      "epoch": 48.99,
+      "grad_norm": 1.0763860940933228,
+      "learning_rate": 2e-05,
+      "loss": 0.04218274,
+      "step": 24495
+    },
+    {
+      "epoch": 48.992,
+      "grad_norm": 1.1772406101226807,
+      "learning_rate": 2e-05,
+      "loss": 0.06190931,
+      "step": 24496
+    },
+    {
+      "epoch": 48.994,
+      "grad_norm": 2.410313844680786,
+      "learning_rate": 2e-05,
+      "loss": 0.05162603,
+      "step": 24497
+    },
+    {
+      "epoch": 48.996,
+      "grad_norm": 1.293403148651123,
+      "learning_rate": 2e-05,
+      "loss": 0.04930715,
+      "step": 24498
+    },
+    {
+      "epoch": 48.998,
+      "grad_norm": 2.3302156925201416,
+      "learning_rate": 2e-05,
+      "loss": 0.05537525,
+      "step": 24499
+    },
+    {
+      "epoch": 49.0,
+      "grad_norm": 1.106102705001831,
+      "learning_rate": 2e-05,
+      "loss": 0.05640403,
+      "step": 24500
+    },
+    {
+      "epoch": 49.0,
+      "eval_performance": {
+        "AngleClassification_1": 1.0,
+        "AngleClassification_2": 1.0,
+        "AngleClassification_3": 0.9720558882235529,
+        "Equal_1": 1.0,
+        "Equal_2": 0.9800399201596807,
+        "Equal_3": 0.9860279441117764,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 1.0,
+        "Parallel_1": 0.9939879759519038,
+        "Parallel_2": 0.9939879759519038,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 1.0,
+        "Perpendicular_2": 0.998,
+        "Perpendicular_3": 0.8897795591182365,
+        "PointLiesOnCircle_1": 1.0,
+        "PointLiesOnCircle_2": 0.994,
+        "PointLiesOnCircle_3": 0.9956,
+        "PointLiesOnLine_1": 1.0,
+        "PointLiesOnLine_2": 0.9939879759519038,
+        "PointLiesOnLine_3": 0.9880239520958084
+      },
+      "eval_runtime": 320.4713,
+      "eval_samples_per_second": 32.764,
+      "eval_steps_per_second": 0.655,
+      "step": 24500
+    },
+    {
+      "epoch": 49.002,
+      "grad_norm": 1.3205698728561401,
+      "learning_rate": 2e-05,
+      "loss": 0.0536387,
+      "step": 24501
+    },
+    {
+      "epoch": 49.004,
+      "grad_norm": 1.0756568908691406,
+      "learning_rate": 2e-05,
+      "loss": 0.03509374,
+      "step": 24502
+    },
+    {
+      "epoch": 49.006,
+      "grad_norm": 1.1383591890335083,
+      "learning_rate": 2e-05,
+      "loss": 0.05303418,
+      "step": 24503
+    },
+    {
+      "epoch": 49.008,
+      "grad_norm": 1.2055492401123047,
+      "learning_rate": 2e-05,
+      "loss": 0.0511542,
+      "step": 24504
+    },
+    {
+      "epoch": 49.01,
+      "grad_norm": 1.0886911153793335,
+      "learning_rate": 2e-05,
+      "loss": 0.0437855,
+      "step": 24505
+    },
+    {
+      "epoch": 49.012,
+      "grad_norm": 1.0806195735931396,
+      "learning_rate": 2e-05,
+      "loss": 0.04402277,
+      "step": 24506
+    },
+    {
+      "epoch": 49.014,
+      "grad_norm": 0.97076815366745,
+      "learning_rate": 2e-05,
+      "loss": 0.04103847,
+      "step": 24507
+    },
+    {
+      "epoch": 49.016,
+      "grad_norm": 0.9942542910575867,
+      "learning_rate": 2e-05,
+      "loss": 0.03674922,
+      "step": 24508
+    },
+    {
+      "epoch": 49.018,
+      "grad_norm": 1.3719143867492676,
+      "learning_rate": 2e-05,
+      "loss": 0.04641896,
+      "step": 24509
+    },
+    {
+      "epoch": 49.02,
+      "grad_norm": 1.4681041240692139,
+      "learning_rate": 2e-05,
+      "loss": 0.05441123,
+      "step": 24510
+    },
+    {
+      "epoch": 49.022,
+      "grad_norm": 1.3180499076843262,
+      "learning_rate": 2e-05,
+      "loss": 0.05244955,
+      "step": 24511
+    },
+    {
+      "epoch": 49.024,
+      "grad_norm": 2.063997983932495,
+      "learning_rate": 2e-05,
+      "loss": 0.06359348,
+      "step": 24512
+    },
+    {
+      "epoch": 49.026,
+      "grad_norm": 1.1440930366516113,
+      "learning_rate": 2e-05,
+      "loss": 0.04446263,
+      "step": 24513
+    },
+    {
+      "epoch": 49.028,
+      "grad_norm": 1.2878350019454956,
+      "learning_rate": 2e-05,
+      "loss": 0.03576039,
+      "step": 24514
+    },
+    {
+      "epoch": 49.03,
+      "grad_norm": 0.9617430567741394,
+      "learning_rate": 2e-05,
+      "loss": 0.02875559,
+      "step": 24515
+    },
+    {
+      "epoch": 49.032,
+      "grad_norm": 1.2200411558151245,
+      "learning_rate": 2e-05,
+      "loss": 0.04824933,
+      "step": 24516
+    },
+    {
+      "epoch": 49.034,
+      "grad_norm": 1.4803850650787354,
+      "learning_rate": 2e-05,
+      "loss": 0.05635831,
+      "step": 24517
+    },
+    {
+      "epoch": 49.036,
+      "grad_norm": 1.9696398973464966,
+      "learning_rate": 2e-05,
+      "loss": 0.04805397,
+      "step": 24518
+    },
+    {
+      "epoch": 49.038,
+      "grad_norm": 1.0635892152786255,
+      "learning_rate": 2e-05,
+      "loss": 0.04648872,
+      "step": 24519
+    },
+    {
+      "epoch": 49.04,
+      "grad_norm": 1.0286033153533936,
+      "learning_rate": 2e-05,
+      "loss": 0.03219466,
+      "step": 24520
+    },
+    {
+      "epoch": 49.042,
+      "grad_norm": 1.2040108442306519,
+      "learning_rate": 2e-05,
+      "loss": 0.05577746,
+      "step": 24521
+    },
+    {
+      "epoch": 49.044,
+      "grad_norm": 1.2264634370803833,
+      "learning_rate": 2e-05,
+      "loss": 0.05810263,
+      "step": 24522
+    },
+    {
+      "epoch": 49.046,
+      "grad_norm": 1.1748919486999512,
+      "learning_rate": 2e-05,
+      "loss": 0.0435359,
+      "step": 24523
+    },
+    {
+      "epoch": 49.048,
+      "grad_norm": 1.623011589050293,
+      "learning_rate": 2e-05,
+      "loss": 0.06521875,
+      "step": 24524
+    },
+    {
+      "epoch": 49.05,
+      "grad_norm": 1.593483805656433,
+      "learning_rate": 2e-05,
+      "loss": 0.04450686,
+      "step": 24525
+    },
+    {
+      "epoch": 49.052,
+      "grad_norm": 1.0634883642196655,
+      "learning_rate": 2e-05,
+      "loss": 0.0392899,
+      "step": 24526
+    },
+    {
+      "epoch": 49.054,
+      "grad_norm": 0.962153971195221,
+      "learning_rate": 2e-05,
+      "loss": 0.03904355,
+      "step": 24527
+    },
+    {
+      "epoch": 49.056,
+      "grad_norm": 1.0543261766433716,
+      "learning_rate": 2e-05,
+      "loss": 0.03974721,
+      "step": 24528
+    },
+    {
+      "epoch": 49.058,
+      "grad_norm": 1.3880788087844849,
+      "learning_rate": 2e-05,
+      "loss": 0.04019085,
+      "step": 24529
+    },
+    {
+      "epoch": 49.06,
+      "grad_norm": 1.029956340789795,
+      "learning_rate": 2e-05,
+      "loss": 0.04097775,
+      "step": 24530
+    },
+    {
+      "epoch": 49.062,
+      "grad_norm": 1.2715673446655273,
+      "learning_rate": 2e-05,
+      "loss": 0.05796398,
+      "step": 24531
+    },
+    {
+      "epoch": 49.064,
+      "grad_norm": 1.1110681295394897,
+      "learning_rate": 2e-05,
+      "loss": 0.04320613,
+      "step": 24532
+    },
+    {
+      "epoch": 49.066,
+      "grad_norm": 1.2434399127960205,
+      "learning_rate": 2e-05,
+      "loss": 0.05722959,
+      "step": 24533
+    },
+    {
+      "epoch": 49.068,
+      "grad_norm": 1.0647788047790527,
+      "learning_rate": 2e-05,
+      "loss": 0.04631557,
+      "step": 24534
+    },
+    {
+      "epoch": 49.07,
+      "grad_norm": 1.1238845586776733,
+      "learning_rate": 2e-05,
+      "loss": 0.04434226,
+      "step": 24535
+    },
+    {
+      "epoch": 49.072,
+      "grad_norm": 2.4695746898651123,
+      "learning_rate": 2e-05,
+      "loss": 0.05840037,
+      "step": 24536
+    },
+    {
+      "epoch": 49.074,
+      "grad_norm": 1.5649980306625366,
+      "learning_rate": 2e-05,
+      "loss": 0.04456908,
+      "step": 24537
+    },
+    {
+      "epoch": 49.076,
+      "grad_norm": 1.1340800523757935,
+      "learning_rate": 2e-05,
+      "loss": 0.02922975,
+      "step": 24538
+    },
+    {
+      "epoch": 49.078,
+      "grad_norm": 1.6249136924743652,
+      "learning_rate": 2e-05,
+      "loss": 0.05591683,
+      "step": 24539
+    },
+    {
+      "epoch": 49.08,
+      "grad_norm": 1.0379282236099243,
+      "learning_rate": 2e-05,
+      "loss": 0.04590633,
+      "step": 24540
+    },
+    {
+      "epoch": 49.082,
+      "grad_norm": 1.1572468280792236,
+      "learning_rate": 2e-05,
+      "loss": 0.05310665,
+      "step": 24541
+    },
+    {
+      "epoch": 49.084,
+      "grad_norm": 1.3208953142166138,
+      "learning_rate": 2e-05,
+      "loss": 0.04799189,
+      "step": 24542
+    },
+    {
+      "epoch": 49.086,
+      "grad_norm": 1.1476277112960815,
+      "learning_rate": 2e-05,
+      "loss": 0.04075008,
+      "step": 24543
+    },
+    {
+      "epoch": 49.088,
+      "grad_norm": 1.445860505104065,
+      "learning_rate": 2e-05,
+      "loss": 0.04546492,
+      "step": 24544
+    },
+    {
+      "epoch": 49.09,
+      "grad_norm": 1.02455472946167,
+      "learning_rate": 2e-05,
+      "loss": 0.04289842,
+      "step": 24545
+    },
+    {
+      "epoch": 49.092,
+      "grad_norm": 1.0884724855422974,
+      "learning_rate": 2e-05,
+      "loss": 0.05313245,
+      "step": 24546
+    },
+    {
+      "epoch": 49.094,
+      "grad_norm": 1.2504005432128906,
+      "learning_rate": 2e-05,
+      "loss": 0.05837908,
+      "step": 24547
+    },
+    {
+      "epoch": 49.096,
+      "grad_norm": 1.050110101699829,
+      "learning_rate": 2e-05,
+      "loss": 0.04570572,
+      "step": 24548
+    },
+    {
+      "epoch": 49.098,
+      "grad_norm": 1.3742461204528809,
+      "learning_rate": 2e-05,
+      "loss": 0.04856823,
+      "step": 24549
+    },
+    {
+      "epoch": 49.1,
+      "grad_norm": 1.0349839925765991,
+      "learning_rate": 2e-05,
+      "loss": 0.04231741,
+      "step": 24550
+    },
+    {
+      "epoch": 49.102,
+      "grad_norm": 1.7095385789871216,
+      "learning_rate": 2e-05,
+      "loss": 0.05369782,
+      "step": 24551
+    },
+    {
+      "epoch": 49.104,
+      "grad_norm": 1.103822946548462,
+      "learning_rate": 2e-05,
+      "loss": 0.04244398,
+      "step": 24552
+    },
+    {
+      "epoch": 49.106,
+      "grad_norm": 1.2387681007385254,
+      "learning_rate": 2e-05,
+      "loss": 0.05892359,
+      "step": 24553
+    },
+    {
+      "epoch": 49.108,
+      "grad_norm": 1.1725491285324097,
+      "learning_rate": 2e-05,
+      "loss": 0.05246655,
+      "step": 24554
+    },
+    {
+      "epoch": 49.11,
+      "grad_norm": 1.0159834623336792,
+      "learning_rate": 2e-05,
+      "loss": 0.04580885,
+      "step": 24555
+    },
+    {
+      "epoch": 49.112,
+      "grad_norm": 1.1978039741516113,
+      "learning_rate": 2e-05,
+      "loss": 0.05257396,
+      "step": 24556
+    },
+    {
+      "epoch": 49.114,
+      "grad_norm": 1.2739366292953491,
+      "learning_rate": 2e-05,
+      "loss": 0.06457013,
+      "step": 24557
+    },
+    {
+      "epoch": 49.116,
+      "grad_norm": 1.2462983131408691,
+      "learning_rate": 2e-05,
+      "loss": 0.05257029,
+      "step": 24558
+    },
+    {
+      "epoch": 49.118,
+      "grad_norm": 1.0246502161026,
+      "learning_rate": 2e-05,
+      "loss": 0.03672457,
+      "step": 24559
+    },
+    {
+      "epoch": 49.12,
+      "grad_norm": 1.0243715047836304,
+      "learning_rate": 2e-05,
+      "loss": 0.03848898,
+      "step": 24560
+    },
+    {
+      "epoch": 49.122,
+      "grad_norm": 1.0890052318572998,
+      "learning_rate": 2e-05,
+      "loss": 0.04123619,
+      "step": 24561
+    },
+    {
+      "epoch": 49.124,
+      "grad_norm": 1.0428330898284912,
+      "learning_rate": 2e-05,
+      "loss": 0.04471182,
+      "step": 24562
+    },
+    {
+      "epoch": 49.126,
+      "grad_norm": 1.534914255142212,
+      "learning_rate": 2e-05,
+      "loss": 0.03503163,
+      "step": 24563
+    },
+    {
+      "epoch": 49.128,
+      "grad_norm": 3.263657808303833,
+      "learning_rate": 2e-05,
+      "loss": 0.06536023,
+      "step": 24564
+    },
+    {
+      "epoch": 49.13,
+      "grad_norm": 0.8561310768127441,
+      "learning_rate": 2e-05,
+      "loss": 0.02339423,
+      "step": 24565
+    },
+    {
+      "epoch": 49.132,
+      "grad_norm": 1.2668267488479614,
+      "learning_rate": 2e-05,
+      "loss": 0.05148039,
+      "step": 24566
+    },
+    {
+      "epoch": 49.134,
+      "grad_norm": 1.4245041608810425,
+      "learning_rate": 2e-05,
+      "loss": 0.04260141,
+      "step": 24567
+    },
+    {
+      "epoch": 49.136,
+      "grad_norm": 1.2553406953811646,
+      "learning_rate": 2e-05,
+      "loss": 0.05330418,
+      "step": 24568
+    },
+    {
+      "epoch": 49.138,
+      "grad_norm": 1.5819698572158813,
+      "learning_rate": 2e-05,
+      "loss": 0.0418847,
+      "step": 24569
+    },
+    {
+      "epoch": 49.14,
+      "grad_norm": 1.157503604888916,
+      "learning_rate": 2e-05,
+      "loss": 0.0456868,
+      "step": 24570
+    },
+    {
+      "epoch": 49.142,
+      "grad_norm": 1.1057533025741577,
+      "learning_rate": 2e-05,
+      "loss": 0.04578763,
+      "step": 24571
+    },
+    {
+      "epoch": 49.144,
+      "grad_norm": 1.3727549314498901,
+      "learning_rate": 2e-05,
+      "loss": 0.06042334,
+      "step": 24572
+    },
+    {
+      "epoch": 49.146,
+      "grad_norm": 1.2736660242080688,
+      "learning_rate": 2e-05,
+      "loss": 0.05840679,
+      "step": 24573
+    },
+    {
+      "epoch": 49.148,
+      "grad_norm": 1.6220810413360596,
+      "learning_rate": 2e-05,
+      "loss": 0.04685524,
+      "step": 24574
+    },
+    {
+      "epoch": 49.15,
+      "grad_norm": 1.0651113986968994,
+      "learning_rate": 2e-05,
+      "loss": 0.0451685,
+      "step": 24575
+    },
+    {
+      "epoch": 49.152,
+      "grad_norm": 1.3879947662353516,
+      "learning_rate": 2e-05,
+      "loss": 0.05266923,
+      "step": 24576
+    },
+    {
+      "epoch": 49.154,
+      "grad_norm": 1.164844036102295,
+      "learning_rate": 2e-05,
+      "loss": 0.04246272,
+      "step": 24577
+    },
+    {
+      "epoch": 49.156,
+      "grad_norm": 1.1619504690170288,
+      "learning_rate": 2e-05,
+      "loss": 0.04659127,
+      "step": 24578
+    },
+    {
+      "epoch": 49.158,
+      "grad_norm": 1.3078287839889526,
+      "learning_rate": 2e-05,
+      "loss": 0.04308105,
+      "step": 24579
+    },
+    {
+      "epoch": 49.16,
+      "grad_norm": 1.6431167125701904,
+      "learning_rate": 2e-05,
+      "loss": 0.04275879,
+      "step": 24580
+    },
+    {
+      "epoch": 49.162,
+      "grad_norm": 0.956802248954773,
+      "learning_rate": 2e-05,
+      "loss": 0.03798681,
+      "step": 24581
+    },
+    {
+      "epoch": 49.164,
+      "grad_norm": 1.2058719396591187,
+      "learning_rate": 2e-05,
+      "loss": 0.04575798,
+      "step": 24582
+    },
+    {
+      "epoch": 49.166,
+      "grad_norm": 0.9033730626106262,
+      "learning_rate": 2e-05,
+      "loss": 0.03232187,
+      "step": 24583
+    },
+    {
+      "epoch": 49.168,
+      "grad_norm": 1.3325138092041016,
+      "learning_rate": 2e-05,
+      "loss": 0.03885297,
+      "step": 24584
+    },
+    {
+      "epoch": 49.17,
+      "grad_norm": 1.1684311628341675,
+      "learning_rate": 2e-05,
+      "loss": 0.04936481,
+      "step": 24585
+    },
+    {
+      "epoch": 49.172,
+      "grad_norm": 0.9386561512947083,
+      "learning_rate": 2e-05,
+      "loss": 0.02889916,
+      "step": 24586
+    },
+    {
+      "epoch": 49.174,
+      "grad_norm": 1.115565538406372,
+      "learning_rate": 2e-05,
+      "loss": 0.04214649,
+      "step": 24587
+    },
+    {
+      "epoch": 49.176,
+      "grad_norm": 1.8959245681762695,
+      "learning_rate": 2e-05,
+      "loss": 0.05413285,
+      "step": 24588
+    },
+    {
+      "epoch": 49.178,
+      "grad_norm": 1.2707425355911255,
+      "learning_rate": 2e-05,
+      "loss": 0.04919063,
+      "step": 24589
+    },
+    {
+      "epoch": 49.18,
+      "grad_norm": 1.1527551412582397,
+      "learning_rate": 2e-05,
+      "loss": 0.04787771,
+      "step": 24590
+    },
+    {
+      "epoch": 49.182,
+      "grad_norm": 1.5040851831436157,
+      "learning_rate": 2e-05,
+      "loss": 0.0654645,
+      "step": 24591
+    },
+    {
+      "epoch": 49.184,
+      "grad_norm": 1.2260459661483765,
+      "learning_rate": 2e-05,
+      "loss": 0.04879155,
+      "step": 24592
+    },
+    {
+      "epoch": 49.186,
+      "grad_norm": 1.2306815385818481,
+      "learning_rate": 2e-05,
+      "loss": 0.03491394,
+      "step": 24593
+    },
+    {
+      "epoch": 49.188,
+      "grad_norm": 1.220257043838501,
+      "learning_rate": 2e-05,
+      "loss": 0.04896056,
+      "step": 24594
+    },
+    {
+      "epoch": 49.19,
+      "grad_norm": 1.178658366203308,
+      "learning_rate": 2e-05,
+      "loss": 0.03527681,
+      "step": 24595
+    },
+    {
+      "epoch": 49.192,
+      "grad_norm": 1.795593023300171,
+      "learning_rate": 2e-05,
+      "loss": 0.07905695,
+      "step": 24596
+    },
+    {
+      "epoch": 49.194,
+      "grad_norm": 1.1190385818481445,
+      "learning_rate": 2e-05,
+      "loss": 0.04868895,
+      "step": 24597
+    },
+    {
+      "epoch": 49.196,
+      "grad_norm": 1.0901833772659302,
+      "learning_rate": 2e-05,
+      "loss": 0.04650778,
+      "step": 24598
+    },
+    {
+      "epoch": 49.198,
+      "grad_norm": 1.1784275770187378,
+      "learning_rate": 2e-05,
+      "loss": 0.03697899,
+      "step": 24599
+    },
+    {
+      "epoch": 49.2,
+      "grad_norm": 0.9963816404342651,
+      "learning_rate": 2e-05,
+      "loss": 0.04271566,
+      "step": 24600
+    },
+    {
+      "epoch": 49.202,
+      "grad_norm": 1.3513808250427246,
+      "learning_rate": 2e-05,
+      "loss": 0.03827177,
+      "step": 24601
+    },
+    {
+      "epoch": 49.204,
+      "grad_norm": 1.297082781791687,
+      "learning_rate": 2e-05,
+      "loss": 0.05176802,
+      "step": 24602
+    },
+    {
+      "epoch": 49.206,
+      "grad_norm": 1.3902502059936523,
+      "learning_rate": 2e-05,
+      "loss": 0.04333564,
+      "step": 24603
+    },
+    {
+      "epoch": 49.208,
+      "grad_norm": 1.5116182565689087,
+      "learning_rate": 2e-05,
+      "loss": 0.06067524,
+      "step": 24604
+    },
+    {
+      "epoch": 49.21,
+      "grad_norm": 1.116183876991272,
+      "learning_rate": 2e-05,
+      "loss": 0.04102014,
+      "step": 24605
+    },
+    {
+      "epoch": 49.212,
+      "grad_norm": 1.3426848649978638,
+      "learning_rate": 2e-05,
+      "loss": 0.06735122,
+      "step": 24606
+    },
+    {
+      "epoch": 49.214,
+      "grad_norm": 1.37204110622406,
+      "learning_rate": 2e-05,
+      "loss": 0.04117677,
+      "step": 24607
+    },
+    {
+      "epoch": 49.216,
+      "grad_norm": 1.1000261306762695,
+      "learning_rate": 2e-05,
+      "loss": 0.03981378,
+      "step": 24608
+    },
+    {
+      "epoch": 49.218,
+      "grad_norm": 0.9247246980667114,
+      "learning_rate": 2e-05,
+      "loss": 0.02853769,
+      "step": 24609
+    },
+    {
+      "epoch": 49.22,
+      "grad_norm": 1.3698744773864746,
+      "learning_rate": 2e-05,
+      "loss": 0.05477355,
+      "step": 24610
+    },
+    {
+      "epoch": 49.222,
+      "grad_norm": 7.3837504386901855,
+      "learning_rate": 2e-05,
+      "loss": 0.06142946,
+      "step": 24611
+    },
+    {
+      "epoch": 49.224,
+      "grad_norm": 1.7267507314682007,
+      "learning_rate": 2e-05,
+      "loss": 0.05547143,
+      "step": 24612
+    },
+    {
+      "epoch": 49.226,
+      "grad_norm": 1.1767046451568604,
+      "learning_rate": 2e-05,
+      "loss": 0.05019828,
+      "step": 24613
+    },
+    {
+      "epoch": 49.228,
+      "grad_norm": 1.5185580253601074,
+      "learning_rate": 2e-05,
+      "loss": 0.05122478,
+      "step": 24614
+    },
+    {
+      "epoch": 49.23,
+      "grad_norm": 1.286940097808838,
+      "learning_rate": 2e-05,
+      "loss": 0.05315614,
+      "step": 24615
+    },
+    {
+      "epoch": 49.232,
+      "grad_norm": 1.2825911045074463,
+      "learning_rate": 2e-05,
+      "loss": 0.04709144,
+      "step": 24616
+    },
+    {
+      "epoch": 49.234,
+      "grad_norm": 1.1364567279815674,
+      "learning_rate": 2e-05,
+      "loss": 0.06106904,
+      "step": 24617
+    },
+    {
+      "epoch": 49.236,
+      "grad_norm": 1.2616055011749268,
+      "learning_rate": 2e-05,
+      "loss": 0.05195578,
+      "step": 24618
+    },
+    {
+      "epoch": 49.238,
+      "grad_norm": 1.2682206630706787,
+      "learning_rate": 2e-05,
+      "loss": 0.0589853,
+      "step": 24619
+    },
+    {
+      "epoch": 49.24,
+      "grad_norm": 1.1738789081573486,
+      "learning_rate": 2e-05,
+      "loss": 0.04862665,
+      "step": 24620
+    },
+    {
+      "epoch": 49.242,
+      "grad_norm": 1.5461301803588867,
+      "learning_rate": 2e-05,
+      "loss": 0.05598284,
+      "step": 24621
+    },
+    {
+      "epoch": 49.244,
+      "grad_norm": 1.336150050163269,
+      "learning_rate": 2e-05,
+      "loss": 0.05503397,
+      "step": 24622
+    },
+    {
+      "epoch": 49.246,
+      "grad_norm": 1.2071928977966309,
+      "learning_rate": 2e-05,
+      "loss": 0.05489289,
+      "step": 24623
+    },
+    {
+      "epoch": 49.248,
+      "grad_norm": 1.186057448387146,
+      "learning_rate": 2e-05,
+      "loss": 0.04869892,
+      "step": 24624
+    },
+    {
+      "epoch": 49.25,
+      "grad_norm": 1.1905598640441895,
+      "learning_rate": 2e-05,
+      "loss": 0.04225285,
+      "step": 24625
+    },
+    {
+      "epoch": 49.252,
+      "grad_norm": 1.1007153987884521,
+      "learning_rate": 2e-05,
+      "loss": 0.05079801,
+      "step": 24626
+    },
+    {
+      "epoch": 49.254,
+      "grad_norm": 1.4382884502410889,
+      "learning_rate": 2e-05,
+      "loss": 0.03811382,
+      "step": 24627
+    },
+    {
+      "epoch": 49.256,
+      "grad_norm": 1.2123656272888184,
+      "learning_rate": 2e-05,
+      "loss": 0.05187578,
+      "step": 24628
+    },
+    {
+      "epoch": 49.258,
+      "grad_norm": 1.222488284111023,
+      "learning_rate": 2e-05,
+      "loss": 0.05206077,
+      "step": 24629
+    },
+    {
+      "epoch": 49.26,
+      "grad_norm": 1.4503012895584106,
+      "learning_rate": 2e-05,
+      "loss": 0.04863203,
+      "step": 24630
+    },
+    {
+      "epoch": 49.262,
+      "grad_norm": 1.278603196144104,
+      "learning_rate": 2e-05,
+      "loss": 0.05332947,
+      "step": 24631
+    },
+    {
+      "epoch": 49.264,
+      "grad_norm": 1.2204537391662598,
+      "learning_rate": 2e-05,
+      "loss": 0.0511545,
+      "step": 24632
+    },
+    {
+      "epoch": 49.266,
+      "grad_norm": 1.0226802825927734,
+      "learning_rate": 2e-05,
+      "loss": 0.0365659,
+      "step": 24633
+    },
+    {
+      "epoch": 49.268,
+      "grad_norm": 0.9728658199310303,
+      "learning_rate": 2e-05,
+      "loss": 0.03656091,
+      "step": 24634
+    },
+    {
+      "epoch": 49.27,
+      "grad_norm": 1.2572722434997559,
+      "learning_rate": 2e-05,
+      "loss": 0.05211511,
+      "step": 24635
+    },
+    {
+      "epoch": 49.272,
+      "grad_norm": 1.153205394744873,
+      "learning_rate": 2e-05,
+      "loss": 0.06159973,
+      "step": 24636
+    },
+    {
+      "epoch": 49.274,
+      "grad_norm": 1.1002930402755737,
+      "learning_rate": 2e-05,
+      "loss": 0.04466134,
+      "step": 24637
+    },
+    {
+      "epoch": 49.276,
+      "grad_norm": 1.2222155332565308,
+      "learning_rate": 2e-05,
+      "loss": 0.05036249,
+      "step": 24638
+    },
+    {
+      "epoch": 49.278,
+      "grad_norm": 3.585354804992676,
+      "learning_rate": 2e-05,
+      "loss": 0.03442348,
+      "step": 24639
+    },
+    {
+      "epoch": 49.28,
+      "grad_norm": 1.2959743738174438,
+      "learning_rate": 2e-05,
+      "loss": 0.05108596,
+      "step": 24640
+    },
+    {
+      "epoch": 49.282,
+      "grad_norm": 1.3540737628936768,
+      "learning_rate": 2e-05,
+      "loss": 0.05969388,
+      "step": 24641
+    },
+    {
+      "epoch": 49.284,
+      "grad_norm": 1.276310920715332,
+      "learning_rate": 2e-05,
+      "loss": 0.05358229,
+      "step": 24642
+    },
+    {
+      "epoch": 49.286,
+      "grad_norm": 1.1716383695602417,
+      "learning_rate": 2e-05,
+      "loss": 0.03722314,
+      "step": 24643
+    },
+    {
+      "epoch": 49.288,
+      "grad_norm": 1.0765808820724487,
+      "learning_rate": 2e-05,
+      "loss": 0.03878718,
+      "step": 24644
+    },
+    {
+      "epoch": 49.29,
+      "grad_norm": 1.1529293060302734,
+      "learning_rate": 2e-05,
+      "loss": 0.05101361,
+      "step": 24645
+    },
+    {
+      "epoch": 49.292,
+      "grad_norm": 1.209089994430542,
+      "learning_rate": 2e-05,
+      "loss": 0.03981159,
+      "step": 24646
+    },
+    {
+      "epoch": 49.294,
+      "grad_norm": 1.1552914381027222,
+      "learning_rate": 2e-05,
+      "loss": 0.03878843,
+      "step": 24647
+    },
+    {
+      "epoch": 49.296,
+      "grad_norm": 3.3828413486480713,
+      "learning_rate": 2e-05,
+      "loss": 0.06451204,
+      "step": 24648
+    },
+    {
+      "epoch": 49.298,
+      "grad_norm": 1.4841753244400024,
+      "learning_rate": 2e-05,
+      "loss": 0.04134528,
+      "step": 24649
+    },
+    {
+      "epoch": 49.3,
+      "grad_norm": 1.1813883781433105,
+      "learning_rate": 2e-05,
+      "loss": 0.04356892,
+      "step": 24650
+    },
+    {
+      "epoch": 49.302,
+      "grad_norm": 1.3516215085983276,
+      "learning_rate": 2e-05,
+      "loss": 0.06003821,
+      "step": 24651
+    },
+    {
+      "epoch": 49.304,
+      "grad_norm": 1.0883997678756714,
+      "learning_rate": 2e-05,
+      "loss": 0.04874385,
+      "step": 24652
+    },
+    {
+      "epoch": 49.306,
+      "grad_norm": 1.3258439302444458,
+      "learning_rate": 2e-05,
+      "loss": 0.06162676,
+      "step": 24653
+    },
+    {
+      "epoch": 49.308,
+      "grad_norm": 0.9306918382644653,
+      "learning_rate": 2e-05,
+      "loss": 0.04051726,
+      "step": 24654
+    },
+    {
+      "epoch": 49.31,
+      "grad_norm": 1.3069220781326294,
+      "learning_rate": 2e-05,
+      "loss": 0.06446021,
+      "step": 24655
+    },
+    {
+      "epoch": 49.312,
+      "grad_norm": 1.4380950927734375,
+      "learning_rate": 2e-05,
+      "loss": 0.06534431,
+      "step": 24656
+    },
+    {
+      "epoch": 49.314,
+      "grad_norm": 1.293108582496643,
+      "learning_rate": 2e-05,
+      "loss": 0.05122873,
+      "step": 24657
+    },
+    {
+      "epoch": 49.316,
+      "grad_norm": 1.5020869970321655,
+      "learning_rate": 2e-05,
+      "loss": 0.04365094,
+      "step": 24658
+    },
+    {
+      "epoch": 49.318,
+      "grad_norm": 1.0571807622909546,
+      "learning_rate": 2e-05,
+      "loss": 0.04433767,
+      "step": 24659
+    },
+    {
+      "epoch": 49.32,
+      "grad_norm": 1.6660947799682617,
+      "learning_rate": 2e-05,
+      "loss": 0.05386006,
+      "step": 24660
+    },
+    {
+      "epoch": 49.322,
+      "grad_norm": 1.2687824964523315,
+      "learning_rate": 2e-05,
+      "loss": 0.05124846,
+      "step": 24661
+    },
+    {
+      "epoch": 49.324,
+      "grad_norm": 1.3088194131851196,
+      "learning_rate": 2e-05,
+      "loss": 0.05232558,
+      "step": 24662
+    },
+    {
+      "epoch": 49.326,
+      "grad_norm": 1.0739314556121826,
+      "learning_rate": 2e-05,
+      "loss": 0.04608125,
+      "step": 24663
+    },
+    {
+      "epoch": 49.328,
+      "grad_norm": 1.3387324810028076,
+      "learning_rate": 2e-05,
+      "loss": 0.0492764,
+      "step": 24664
+    },
+    {
+      "epoch": 49.33,
+      "grad_norm": 1.2844595909118652,
+      "learning_rate": 2e-05,
+      "loss": 0.04417478,
+      "step": 24665
+    },
+    {
+      "epoch": 49.332,
+      "grad_norm": 1.1230336427688599,
+      "learning_rate": 2e-05,
+      "loss": 0.04758064,
+      "step": 24666
+    },
+    {
+      "epoch": 49.334,
+      "grad_norm": 5.419571876525879,
+      "learning_rate": 2e-05,
+      "loss": 0.04511326,
+      "step": 24667
+    },
+    {
+      "epoch": 49.336,
+      "grad_norm": 1.1522164344787598,
+      "learning_rate": 2e-05,
+      "loss": 0.04697948,
+      "step": 24668
+    },
+    {
+      "epoch": 49.338,
+      "grad_norm": 1.43909752368927,
+      "learning_rate": 2e-05,
+      "loss": 0.05124324,
+      "step": 24669
+    },
+    {
+      "epoch": 49.34,
+      "grad_norm": 1.290650486946106,
+      "learning_rate": 2e-05,
+      "loss": 0.03891211,
+      "step": 24670
+    },
+    {
+      "epoch": 49.342,
+      "grad_norm": 1.0777543783187866,
+      "learning_rate": 2e-05,
+      "loss": 0.03897686,
+      "step": 24671
+    },
+    {
+      "epoch": 49.344,
+      "grad_norm": 1.2240030765533447,
+      "learning_rate": 2e-05,
+      "loss": 0.04273064,
+      "step": 24672
+    },
+    {
+      "epoch": 49.346,
+      "grad_norm": 1.2230967283248901,
+      "learning_rate": 2e-05,
+      "loss": 0.05533133,
+      "step": 24673
+    },
+    {
+      "epoch": 49.348,
+      "grad_norm": 1.7442550659179688,
+      "learning_rate": 2e-05,
+      "loss": 0.04444986,
+      "step": 24674
+    },
+    {
+      "epoch": 49.35,
+      "grad_norm": 1.193528413772583,
+      "learning_rate": 2e-05,
+      "loss": 0.02971295,
+      "step": 24675
+    },
+    {
+      "epoch": 49.352,
+      "grad_norm": 1.548940896987915,
+      "learning_rate": 2e-05,
+      "loss": 0.05257022,
+      "step": 24676
+    },
+    {
+      "epoch": 49.354,
+      "grad_norm": 1.4281851053237915,
+      "learning_rate": 2e-05,
+      "loss": 0.05808748,
+      "step": 24677
+    },
+    {
+      "epoch": 49.356,
+      "grad_norm": 1.1668903827667236,
+      "learning_rate": 2e-05,
+      "loss": 0.039745,
+      "step": 24678
+    },
+    {
+      "epoch": 49.358,
+      "grad_norm": 2.4779348373413086,
+      "learning_rate": 2e-05,
+      "loss": 0.04287686,
+      "step": 24679
+    },
+    {
+      "epoch": 49.36,
+      "grad_norm": 1.0857728719711304,
+      "learning_rate": 2e-05,
+      "loss": 0.03546745,
+      "step": 24680
+    },
+    {
+      "epoch": 49.362,
+      "grad_norm": 1.758777141571045,
+      "learning_rate": 2e-05,
+      "loss": 0.03802776,
+      "step": 24681
+    },
+    {
+      "epoch": 49.364,
+      "grad_norm": 0.9436732530593872,
+      "learning_rate": 2e-05,
+      "loss": 0.03220242,
+      "step": 24682
+    },
+    {
+      "epoch": 49.366,
+      "grad_norm": 1.0815670490264893,
+      "learning_rate": 2e-05,
+      "loss": 0.04280725,
+      "step": 24683
+    },
+    {
+      "epoch": 49.368,
+      "grad_norm": 1.114959955215454,
+      "learning_rate": 2e-05,
+      "loss": 0.04519267,
+      "step": 24684
+    },
+    {
+      "epoch": 49.37,
+      "grad_norm": 1.2002875804901123,
+      "learning_rate": 2e-05,
+      "loss": 0.04737946,
+      "step": 24685
+    },
+    {
+      "epoch": 49.372,
+      "grad_norm": 1.381123661994934,
+      "learning_rate": 2e-05,
+      "loss": 0.0593003,
+      "step": 24686
+    },
+    {
+      "epoch": 49.374,
+      "grad_norm": 2.0842089653015137,
+      "learning_rate": 2e-05,
+      "loss": 0.0516322,
+      "step": 24687
+    },
+    {
+      "epoch": 49.376,
+      "grad_norm": 1.0045135021209717,
+      "learning_rate": 2e-05,
+      "loss": 0.03573417,
+      "step": 24688
+    },
+    {
+      "epoch": 49.378,
+      "grad_norm": 1.3091294765472412,
+      "learning_rate": 2e-05,
+      "loss": 0.04659965,
+      "step": 24689
+    },
+    {
+      "epoch": 49.38,
+      "grad_norm": 1.3777577877044678,
+      "learning_rate": 2e-05,
+      "loss": 0.0343422,
+      "step": 24690
+    },
+    {
+      "epoch": 49.382,
+      "grad_norm": 1.5019209384918213,
+      "learning_rate": 2e-05,
+      "loss": 0.05598423,
+      "step": 24691
+    },
+    {
+      "epoch": 49.384,
+      "grad_norm": 1.168333888053894,
+      "learning_rate": 2e-05,
+      "loss": 0.04606739,
+      "step": 24692
+    },
+    {
+      "epoch": 49.386,
+      "grad_norm": 1.3140724897384644,
+      "learning_rate": 2e-05,
+      "loss": 0.05491744,
+      "step": 24693
+    },
+    {
+      "epoch": 49.388,
+      "grad_norm": 1.1507341861724854,
+      "learning_rate": 2e-05,
+      "loss": 0.04055898,
+      "step": 24694
+    },
+    {
+      "epoch": 49.39,
+      "grad_norm": 1.2338950634002686,
+      "learning_rate": 2e-05,
+      "loss": 0.04385046,
+      "step": 24695
+    },
+    {
+      "epoch": 49.392,
+      "grad_norm": 1.2507073879241943,
+      "learning_rate": 2e-05,
+      "loss": 0.06624331,
+      "step": 24696
+    },
+    {
+      "epoch": 49.394,
+      "grad_norm": 1.172641634941101,
+      "learning_rate": 2e-05,
+      "loss": 0.05282669,
+      "step": 24697
+    },
+    {
+      "epoch": 49.396,
+      "grad_norm": 1.290877103805542,
+      "learning_rate": 2e-05,
+      "loss": 0.0552482,
+      "step": 24698
+    },
+    {
+      "epoch": 49.398,
+      "grad_norm": 1.3046849966049194,
+      "learning_rate": 2e-05,
+      "loss": 0.05484886,
+      "step": 24699
+    },
+    {
+      "epoch": 49.4,
+      "grad_norm": 1.0632675886154175,
+      "learning_rate": 2e-05,
+      "loss": 0.04052684,
+      "step": 24700
+    },
+    {
+      "epoch": 49.402,
+      "grad_norm": 1.0683989524841309,
+      "learning_rate": 2e-05,
+      "loss": 0.04582777,
+      "step": 24701
+    },
+    {
+      "epoch": 49.404,
+      "grad_norm": 1.6077747344970703,
+      "learning_rate": 2e-05,
+      "loss": 0.04787633,
+      "step": 24702
+    },
+    {
+      "epoch": 49.406,
+      "grad_norm": 1.207194209098816,
+      "learning_rate": 2e-05,
+      "loss": 0.05000635,
+      "step": 24703
+    },
+    {
+      "epoch": 49.408,
+      "grad_norm": 1.1203255653381348,
+      "learning_rate": 2e-05,
+      "loss": 0.05599294,
+      "step": 24704
+    },
+    {
+      "epoch": 49.41,
+      "grad_norm": 1.4439189434051514,
+      "learning_rate": 2e-05,
+      "loss": 0.04717576,
+      "step": 24705
+    },
+    {
+      "epoch": 49.412,
+      "grad_norm": 1.1799260377883911,
+      "learning_rate": 2e-05,
+      "loss": 0.04708055,
+      "step": 24706
+    },
+    {
+      "epoch": 49.414,
+      "grad_norm": 1.2697405815124512,
+      "learning_rate": 2e-05,
+      "loss": 0.05073357,
+      "step": 24707
+    },
+    {
+      "epoch": 49.416,
+      "grad_norm": 1.1728367805480957,
+      "learning_rate": 2e-05,
+      "loss": 0.05580144,
+      "step": 24708
+    },
+    {
+      "epoch": 49.418,
+      "grad_norm": 1.9259074926376343,
+      "learning_rate": 2e-05,
+      "loss": 0.04573488,
+      "step": 24709
+    },
+    {
+      "epoch": 49.42,
+      "grad_norm": 1.285468578338623,
+      "learning_rate": 2e-05,
+      "loss": 0.06362525,
+      "step": 24710
+    },
+    {
+      "epoch": 49.422,
+      "grad_norm": 1.0827348232269287,
+      "learning_rate": 2e-05,
+      "loss": 0.0446627,
+      "step": 24711
+    },
+    {
+      "epoch": 49.424,
+      "grad_norm": 1.354174256324768,
+      "learning_rate": 2e-05,
+      "loss": 0.06585463,
+      "step": 24712
+    },
+    {
+      "epoch": 49.426,
+      "grad_norm": 1.2439528703689575,
+      "learning_rate": 2e-05,
+      "loss": 0.04843699,
+      "step": 24713
+    },
+    {
+      "epoch": 49.428,
+      "grad_norm": 1.024624228477478,
+      "learning_rate": 2e-05,
+      "loss": 0.04217855,
+      "step": 24714
+    },
+    {
+      "epoch": 49.43,
+      "grad_norm": 1.0700979232788086,
+      "learning_rate": 2e-05,
+      "loss": 0.04749613,
+      "step": 24715
+    },
+    {
+      "epoch": 49.432,
+      "grad_norm": 1.496504306793213,
+      "learning_rate": 2e-05,
+      "loss": 0.0489388,
+      "step": 24716
+    },
+    {
+      "epoch": 49.434,
+      "grad_norm": 1.1693992614746094,
+      "learning_rate": 2e-05,
+      "loss": 0.04739952,
+      "step": 24717
+    },
+    {
+      "epoch": 49.436,
+      "grad_norm": 1.1603397130966187,
+      "learning_rate": 2e-05,
+      "loss": 0.04808056,
+      "step": 24718
+    },
+    {
+      "epoch": 49.438,
+      "grad_norm": 1.1455094814300537,
+      "learning_rate": 2e-05,
+      "loss": 0.05102035,
+      "step": 24719
+    },
+    {
+      "epoch": 49.44,
+      "grad_norm": 1.1179344654083252,
+      "learning_rate": 2e-05,
+      "loss": 0.04601201,
+      "step": 24720
+    },
+    {
+      "epoch": 49.442,
+      "grad_norm": 1.0473073720932007,
+      "learning_rate": 2e-05,
+      "loss": 0.04930402,
+      "step": 24721
+    },
+    {
+      "epoch": 49.444,
+      "grad_norm": 1.236930251121521,
+      "learning_rate": 2e-05,
+      "loss": 0.04765532,
+      "step": 24722
+    },
+    {
+      "epoch": 49.446,
+      "grad_norm": 1.2558659315109253,
+      "learning_rate": 2e-05,
+      "loss": 0.04867148,
+      "step": 24723
+    },
+    {
+      "epoch": 49.448,
+      "grad_norm": 1.1845966577529907,
+      "learning_rate": 2e-05,
+      "loss": 0.04930171,
+      "step": 24724
+    },
+    {
+      "epoch": 49.45,
+      "grad_norm": 1.112603783607483,
+      "learning_rate": 2e-05,
+      "loss": 0.04704773,
+      "step": 24725
+    },
+    {
+      "epoch": 49.452,
+      "grad_norm": 1.188025712966919,
+      "learning_rate": 2e-05,
+      "loss": 0.04383368,
+      "step": 24726
+    },
+    {
+      "epoch": 49.454,
+      "grad_norm": 1.1775012016296387,
+      "learning_rate": 2e-05,
+      "loss": 0.04882312,
+      "step": 24727
+    },
+    {
+      "epoch": 49.456,
+      "grad_norm": 1.2298845052719116,
+      "learning_rate": 2e-05,
+      "loss": 0.04555701,
+      "step": 24728
+    },
+    {
+      "epoch": 49.458,
+      "grad_norm": 1.373660683631897,
+      "learning_rate": 2e-05,
+      "loss": 0.05526736,
+      "step": 24729
+    },
+    {
+      "epoch": 49.46,
+      "grad_norm": 1.3296241760253906,
+      "learning_rate": 2e-05,
+      "loss": 0.06211912,
+      "step": 24730
+    },
+    {
+      "epoch": 49.462,
+      "grad_norm": 1.1016606092453003,
+      "learning_rate": 2e-05,
+      "loss": 0.05170915,
+      "step": 24731
+    },
+    {
+      "epoch": 49.464,
+      "grad_norm": 1.3367371559143066,
+      "learning_rate": 2e-05,
+      "loss": 0.04138402,
+      "step": 24732
+    },
+    {
+      "epoch": 49.466,
+      "grad_norm": 3.0442302227020264,
+      "learning_rate": 2e-05,
+      "loss": 0.07401902,
+      "step": 24733
+    },
+    {
+      "epoch": 49.468,
+      "grad_norm": 1.304378867149353,
+      "learning_rate": 2e-05,
+      "loss": 0.04841861,
+      "step": 24734
+    },
+    {
+      "epoch": 49.47,
+      "grad_norm": 1.211232304573059,
+      "learning_rate": 2e-05,
+      "loss": 0.04923613,
+      "step": 24735
+    },
+    {
+      "epoch": 49.472,
+      "grad_norm": 1.0207058191299438,
+      "learning_rate": 2e-05,
+      "loss": 0.03922756,
+      "step": 24736
+    },
+    {
+      "epoch": 49.474,
+      "grad_norm": 1.7041534185409546,
+      "learning_rate": 2e-05,
+      "loss": 0.04910336,
+      "step": 24737
+    },
+    {
+      "epoch": 49.476,
+      "grad_norm": 1.1709401607513428,
+      "learning_rate": 2e-05,
+      "loss": 0.04106373,
+      "step": 24738
+    },
+    {
+      "epoch": 49.478,
+      "grad_norm": 2.144960880279541,
+      "learning_rate": 2e-05,
+      "loss": 0.05597088,
+      "step": 24739
+    },
+    {
+      "epoch": 49.48,
+      "grad_norm": 1.052988052368164,
+      "learning_rate": 2e-05,
+      "loss": 0.04293213,
+      "step": 24740
+    },
+    {
+      "epoch": 49.482,
+      "grad_norm": 1.4180502891540527,
+      "learning_rate": 2e-05,
+      "loss": 0.04159679,
+      "step": 24741
+    },
+    {
+      "epoch": 49.484,
+      "grad_norm": 1.2406322956085205,
+      "learning_rate": 2e-05,
+      "loss": 0.05517141,
+      "step": 24742
+    },
+    {
+      "epoch": 49.486,
+      "grad_norm": 1.1686590909957886,
+      "learning_rate": 2e-05,
+      "loss": 0.04570251,
+      "step": 24743
+    },
+    {
+      "epoch": 49.488,
+      "grad_norm": 1.2575324773788452,
+      "learning_rate": 2e-05,
+      "loss": 0.03278139,
+      "step": 24744
+    },
+    {
+      "epoch": 49.49,
+      "grad_norm": 1.0804585218429565,
+      "learning_rate": 2e-05,
+      "loss": 0.03636895,
+      "step": 24745
+    },
+    {
+      "epoch": 49.492,
+      "grad_norm": 1.3195343017578125,
+      "learning_rate": 2e-05,
+      "loss": 0.05228486,
+      "step": 24746
+    },
+    {
+      "epoch": 49.494,
+      "grad_norm": 1.1343023777008057,
+      "learning_rate": 2e-05,
+      "loss": 0.04752772,
+      "step": 24747
+    },
+    {
+      "epoch": 49.496,
+      "grad_norm": 1.1542292833328247,
+      "learning_rate": 2e-05,
+      "loss": 0.04346114,
+      "step": 24748
+    },
+    {
+      "epoch": 49.498,
+      "grad_norm": 1.0267248153686523,
+      "learning_rate": 2e-05,
+      "loss": 0.04148855,
+      "step": 24749
+    },
+    {
+      "epoch": 49.5,
+      "grad_norm": 1.197594404220581,
+      "learning_rate": 2e-05,
+      "loss": 0.05338103,
+      "step": 24750
+    },
+    {
+      "epoch": 49.502,
+      "grad_norm": 1.3752979040145874,
+      "learning_rate": 2e-05,
+      "loss": 0.05251706,
+      "step": 24751
+    },
+    {
+      "epoch": 49.504,
+      "grad_norm": 2.483018636703491,
+      "learning_rate": 2e-05,
+      "loss": 0.05616256,
+      "step": 24752
+    },
+    {
+      "epoch": 49.506,
+      "grad_norm": 0.9340147972106934,
+      "learning_rate": 2e-05,
+      "loss": 0.0262505,
+      "step": 24753
+    },
+    {
+      "epoch": 49.508,
+      "grad_norm": 1.1258265972137451,
+      "learning_rate": 2e-05,
+      "loss": 0.03799858,
+      "step": 24754
+    },
+    {
+      "epoch": 49.51,
+      "grad_norm": 1.6418697834014893,
+      "learning_rate": 2e-05,
+      "loss": 0.06251703,
+      "step": 24755
+    },
+    {
+      "epoch": 49.512,
+      "grad_norm": 1.4351956844329834,
+      "learning_rate": 2e-05,
+      "loss": 0.04956281,
+      "step": 24756
+    },
+    {
+      "epoch": 49.514,
+      "grad_norm": 1.3971424102783203,
+      "learning_rate": 2e-05,
+      "loss": 0.07184118,
+      "step": 24757
+    },
+    {
+      "epoch": 49.516,
+      "grad_norm": 1.242491602897644,
+      "learning_rate": 2e-05,
+      "loss": 0.04963819,
+      "step": 24758
+    },
+    {
+      "epoch": 49.518,
+      "grad_norm": 1.8168903589248657,
+      "learning_rate": 2e-05,
+      "loss": 0.05463743,
+      "step": 24759
+    },
+    {
+      "epoch": 49.52,
+      "grad_norm": 1.0832699537277222,
+      "learning_rate": 2e-05,
+      "loss": 0.04424177,
+      "step": 24760
+    },
+    {
+      "epoch": 49.522,
+      "grad_norm": 0.9689600467681885,
+      "learning_rate": 2e-05,
+      "loss": 0.04252536,
+      "step": 24761
+    },
+    {
+      "epoch": 49.524,
+      "grad_norm": 1.111344337463379,
+      "learning_rate": 2e-05,
+      "loss": 0.04760269,
+      "step": 24762
+    },
+    {
+      "epoch": 49.526,
+      "grad_norm": 1.120718240737915,
+      "learning_rate": 2e-05,
+      "loss": 0.04254458,
+      "step": 24763
+    },
+    {
+      "epoch": 49.528,
+      "grad_norm": 1.1055079698562622,
+      "learning_rate": 2e-05,
+      "loss": 0.04112123,
+      "step": 24764
+    },
+    {
+      "epoch": 49.53,
+      "grad_norm": 1.8217605352401733,
+      "learning_rate": 2e-05,
+      "loss": 0.05603817,
+      "step": 24765
+    },
+    {
+      "epoch": 49.532,
+      "grad_norm": 1.008670449256897,
+      "learning_rate": 2e-05,
+      "loss": 0.03235183,
+      "step": 24766
+    },
+    {
+      "epoch": 49.534,
+      "grad_norm": 1.1165664196014404,
+      "learning_rate": 2e-05,
+      "loss": 0.0455579,
+      "step": 24767
+    },
+    {
+      "epoch": 49.536,
+      "grad_norm": 1.0880547761917114,
+      "learning_rate": 2e-05,
+      "loss": 0.04338773,
+      "step": 24768
+    },
+    {
+      "epoch": 49.538,
+      "grad_norm": 0.9217278361320496,
+      "learning_rate": 2e-05,
+      "loss": 0.03157655,
+      "step": 24769
+    },
+    {
+      "epoch": 49.54,
+      "grad_norm": 1.095906376838684,
+      "learning_rate": 2e-05,
+      "loss": 0.04428188,
+      "step": 24770
+    },
+    {
+      "epoch": 49.542,
+      "grad_norm": 1.5616658926010132,
+      "learning_rate": 2e-05,
+      "loss": 0.03784732,
+      "step": 24771
+    },
+    {
+      "epoch": 49.544,
+      "grad_norm": 1.0943976640701294,
+      "learning_rate": 2e-05,
+      "loss": 0.04282872,
+      "step": 24772
+    },
+    {
+      "epoch": 49.546,
+      "grad_norm": 1.0541517734527588,
+      "learning_rate": 2e-05,
+      "loss": 0.04491699,
+      "step": 24773
+    },
+    {
+      "epoch": 49.548,
+      "grad_norm": 3.1224377155303955,
+      "learning_rate": 2e-05,
+      "loss": 0.04894248,
+      "step": 24774
+    },
+    {
+      "epoch": 49.55,
+      "grad_norm": 1.100438117980957,
+      "learning_rate": 2e-05,
+      "loss": 0.04886558,
+      "step": 24775
+    },
+    {
+      "epoch": 49.552,
+      "grad_norm": 1.7008203268051147,
+      "learning_rate": 2e-05,
+      "loss": 0.03542287,
+      "step": 24776
+    },
+    {
+      "epoch": 49.554,
+      "grad_norm": 1.2163397073745728,
+      "learning_rate": 2e-05,
+      "loss": 0.05112882,
+      "step": 24777
+    },
+    {
+      "epoch": 49.556,
+      "grad_norm": 1.178572177886963,
+      "learning_rate": 2e-05,
+      "loss": 0.05226102,
+      "step": 24778
+    },
+    {
+      "epoch": 49.558,
+      "grad_norm": 1.07505202293396,
+      "learning_rate": 2e-05,
+      "loss": 0.04829258,
+      "step": 24779
+    },
+    {
+      "epoch": 49.56,
+      "grad_norm": 1.545498013496399,
+      "learning_rate": 2e-05,
+      "loss": 0.05802028,
+      "step": 24780
+    },
+    {
+      "epoch": 49.562,
+      "grad_norm": 1.0516809225082397,
+      "learning_rate": 2e-05,
+      "loss": 0.03485567,
+      "step": 24781
+    },
+    {
+      "epoch": 49.564,
+      "grad_norm": 1.0159010887145996,
+      "learning_rate": 2e-05,
+      "loss": 0.04023987,
+      "step": 24782
+    },
+    {
+      "epoch": 49.566,
+      "grad_norm": 1.1692757606506348,
+      "learning_rate": 2e-05,
+      "loss": 0.04836097,
+      "step": 24783
+    },
+    {
+      "epoch": 49.568,
+      "grad_norm": 1.2164714336395264,
+      "learning_rate": 2e-05,
+      "loss": 0.05162116,
+      "step": 24784
+    },
+    {
+      "epoch": 49.57,
+      "grad_norm": 1.0095146894454956,
+      "learning_rate": 2e-05,
+      "loss": 0.04396236,
+      "step": 24785
+    },
+    {
+      "epoch": 49.572,
+      "grad_norm": 1.3631678819656372,
+      "learning_rate": 2e-05,
+      "loss": 0.05843265,
+      "step": 24786
+    },
+    {
+      "epoch": 49.574,
+      "grad_norm": 1.3029148578643799,
+      "learning_rate": 2e-05,
+      "loss": 0.04724909,
+      "step": 24787
+    },
+    {
+      "epoch": 49.576,
+      "grad_norm": 1.3042387962341309,
+      "learning_rate": 2e-05,
+      "loss": 0.04229445,
+      "step": 24788
+    },
+    {
+      "epoch": 49.578,
+      "grad_norm": 1.0347017049789429,
+      "learning_rate": 2e-05,
+      "loss": 0.03812275,
+      "step": 24789
+    },
+    {
+      "epoch": 49.58,
+      "grad_norm": 1.7310672998428345,
+      "learning_rate": 2e-05,
+      "loss": 0.06088685,
+      "step": 24790
+    },
+    {
+      "epoch": 49.582,
+      "grad_norm": 0.9168398976325989,
+      "learning_rate": 2e-05,
+      "loss": 0.02849315,
+      "step": 24791
+    },
+    {
+      "epoch": 49.584,
+      "grad_norm": 1.1964490413665771,
+      "learning_rate": 2e-05,
+      "loss": 0.0535724,
+      "step": 24792
+    },
+    {
+      "epoch": 49.586,
+      "grad_norm": 1.1546590328216553,
+      "learning_rate": 2e-05,
+      "loss": 0.03581116,
+      "step": 24793
+    },
+    {
+      "epoch": 49.588,
+      "grad_norm": 1.0855379104614258,
+      "learning_rate": 2e-05,
+      "loss": 0.05509954,
+      "step": 24794
+    },
+    {
+      "epoch": 49.59,
+      "grad_norm": 1.2587318420410156,
+      "learning_rate": 2e-05,
+      "loss": 0.04235309,
+      "step": 24795
+    },
+    {
+      "epoch": 49.592,
+      "grad_norm": 1.2183408737182617,
+      "learning_rate": 2e-05,
+      "loss": 0.05718917,
+      "step": 24796
+    },
+    {
+      "epoch": 49.594,
+      "grad_norm": 1.1804263591766357,
+      "learning_rate": 2e-05,
+      "loss": 0.04849926,
+      "step": 24797
+    },
+    {
+      "epoch": 49.596,
+      "grad_norm": 1.4819769859313965,
+      "learning_rate": 2e-05,
+      "loss": 0.05709495,
+      "step": 24798
+    },
+    {
+      "epoch": 49.598,
+      "grad_norm": 1.1413755416870117,
+      "learning_rate": 2e-05,
+      "loss": 0.0554666,
+      "step": 24799
+    },
+    {
+      "epoch": 49.6,
+      "grad_norm": 1.1274505853652954,
+      "learning_rate": 2e-05,
+      "loss": 0.04003352,
+      "step": 24800
+    },
+    {
+      "epoch": 49.602,
+      "grad_norm": 1.2199150323867798,
+      "learning_rate": 2e-05,
+      "loss": 0.04663592,
+      "step": 24801
+    },
+    {
+      "epoch": 49.604,
+      "grad_norm": 1.1105495691299438,
+      "learning_rate": 2e-05,
+      "loss": 0.04194749,
+      "step": 24802
+    },
+    {
+      "epoch": 49.606,
+      "grad_norm": 1.1821727752685547,
+      "learning_rate": 2e-05,
+      "loss": 0.05207427,
+      "step": 24803
+    },
+    {
+      "epoch": 49.608,
+      "grad_norm": 1.0401041507720947,
+      "learning_rate": 2e-05,
+      "loss": 0.04428156,
+      "step": 24804
+    },
+    {
+      "epoch": 49.61,
+      "grad_norm": 1.840406060218811,
+      "learning_rate": 2e-05,
+      "loss": 0.05603269,
+      "step": 24805
+    },
+    {
+      "epoch": 49.612,
+      "grad_norm": 1.3533574342727661,
+      "learning_rate": 2e-05,
+      "loss": 0.03703146,
+      "step": 24806
+    },
+    {
+      "epoch": 49.614,
+      "grad_norm": 1.0640056133270264,
+      "learning_rate": 2e-05,
+      "loss": 0.03546765,
+      "step": 24807
+    },
+    {
+      "epoch": 49.616,
+      "grad_norm": 1.206377625465393,
+      "learning_rate": 2e-05,
+      "loss": 0.05672641,
+      "step": 24808
+    },
+    {
+      "epoch": 49.618,
+      "grad_norm": 1.0968784093856812,
+      "learning_rate": 2e-05,
+      "loss": 0.03934178,
+      "step": 24809
+    },
+    {
+      "epoch": 49.62,
+      "grad_norm": 1.4784859418869019,
+      "learning_rate": 2e-05,
+      "loss": 0.05061619,
+      "step": 24810
+    },
+    {
+      "epoch": 49.622,
+      "grad_norm": 1.2458291053771973,
+      "learning_rate": 2e-05,
+      "loss": 0.04952991,
+      "step": 24811
+    },
+    {
+      "epoch": 49.624,
+      "grad_norm": 0.917413055896759,
+      "learning_rate": 2e-05,
+      "loss": 0.03567467,
+      "step": 24812
+    },
+    {
+      "epoch": 49.626,
+      "grad_norm": 1.093457818031311,
+      "learning_rate": 2e-05,
+      "loss": 0.04676176,
+      "step": 24813
+    },
+    {
+      "epoch": 49.628,
+      "grad_norm": 1.19669771194458,
+      "learning_rate": 2e-05,
+      "loss": 0.05353126,
+      "step": 24814
+    },
+    {
+      "epoch": 49.63,
+      "grad_norm": 1.258165955543518,
+      "learning_rate": 2e-05,
+      "loss": 0.05881431,
+      "step": 24815
+    },
+    {
+      "epoch": 49.632,
+      "grad_norm": 1.8352237939834595,
+      "learning_rate": 2e-05,
+      "loss": 0.03418227,
+      "step": 24816
+    },
+    {
+      "epoch": 49.634,
+      "grad_norm": 1.2149909734725952,
+      "learning_rate": 2e-05,
+      "loss": 0.0405385,
+      "step": 24817
+    },
+    {
+      "epoch": 49.636,
+      "grad_norm": 1.1387282609939575,
+      "learning_rate": 2e-05,
+      "loss": 0.04924862,
+      "step": 24818
+    },
+    {
+      "epoch": 49.638,
+      "grad_norm": 1.0884685516357422,
+      "learning_rate": 2e-05,
+      "loss": 0.03455062,
+      "step": 24819
+    },
+    {
+      "epoch": 49.64,
+      "grad_norm": 1.1216976642608643,
+      "learning_rate": 2e-05,
+      "loss": 0.04718848,
+      "step": 24820
+    },
+    {
+      "epoch": 49.642,
+      "grad_norm": 1.1525574922561646,
+      "learning_rate": 2e-05,
+      "loss": 0.0507764,
+      "step": 24821
+    },
+    {
+      "epoch": 49.644,
+      "grad_norm": 1.040356993675232,
+      "learning_rate": 2e-05,
+      "loss": 0.03835782,
+      "step": 24822
+    },
+    {
+      "epoch": 49.646,
+      "grad_norm": 1.6863218545913696,
+      "learning_rate": 2e-05,
+      "loss": 0.05974036,
+      "step": 24823
+    },
+    {
+      "epoch": 49.648,
+      "grad_norm": 1.5963078737258911,
+      "learning_rate": 2e-05,
+      "loss": 0.05577124,
+      "step": 24824
+    },
+    {
+      "epoch": 49.65,
+      "grad_norm": 1.9898076057434082,
+      "learning_rate": 2e-05,
+      "loss": 0.04162847,
+      "step": 24825
+    },
+    {
+      "epoch": 49.652,
+      "grad_norm": 1.7517437934875488,
+      "learning_rate": 2e-05,
+      "loss": 0.04477061,
+      "step": 24826
+    },
+    {
+      "epoch": 49.654,
+      "grad_norm": 1.0563920736312866,
+      "learning_rate": 2e-05,
+      "loss": 0.04491702,
+      "step": 24827
+    },
+    {
+      "epoch": 49.656,
+      "grad_norm": 1.1382147073745728,
+      "learning_rate": 2e-05,
+      "loss": 0.04465196,
+      "step": 24828
+    },
+    {
+      "epoch": 49.658,
+      "grad_norm": 1.1406420469284058,
+      "learning_rate": 2e-05,
+      "loss": 0.04676701,
+      "step": 24829
+    },
+    {
+      "epoch": 49.66,
+      "grad_norm": 0.9442988038063049,
+      "learning_rate": 2e-05,
+      "loss": 0.03124127,
+      "step": 24830
+    },
+    {
+      "epoch": 49.662,
+      "grad_norm": 1.63547682762146,
+      "learning_rate": 2e-05,
+      "loss": 0.05501399,
+      "step": 24831
+    },
+    {
+      "epoch": 49.664,
+      "grad_norm": 0.9990745782852173,
+      "learning_rate": 2e-05,
+      "loss": 0.03285347,
+      "step": 24832
+    },
+    {
+      "epoch": 49.666,
+      "grad_norm": 1.0436017513275146,
+      "learning_rate": 2e-05,
+      "loss": 0.0376052,
+      "step": 24833
+    },
+    {
+      "epoch": 49.668,
+      "grad_norm": 1.4094048738479614,
+      "learning_rate": 2e-05,
+      "loss": 0.05660111,
+      "step": 24834
+    },
+    {
+      "epoch": 49.67,
+      "grad_norm": 1.181092381477356,
+      "learning_rate": 2e-05,
+      "loss": 0.04279668,
+      "step": 24835
+    },
+    {
+      "epoch": 49.672,
+      "grad_norm": 1.2665365934371948,
+      "learning_rate": 2e-05,
+      "loss": 0.05205832,
+      "step": 24836
+    },
+    {
+      "epoch": 49.674,
+      "grad_norm": 1.061452031135559,
+      "learning_rate": 2e-05,
+      "loss": 0.03486584,
+      "step": 24837
+    },
+    {
+      "epoch": 49.676,
+      "grad_norm": 1.312247395515442,
+      "learning_rate": 2e-05,
+      "loss": 0.05554381,
+      "step": 24838
+    },
+    {
+      "epoch": 49.678,
+      "grad_norm": 1.2573567628860474,
+      "learning_rate": 2e-05,
+      "loss": 0.05710577,
+      "step": 24839
+    },
+    {
+      "epoch": 49.68,
+      "grad_norm": 1.2652338743209839,
+      "learning_rate": 2e-05,
+      "loss": 0.04693561,
+      "step": 24840
+    },
+    {
+      "epoch": 49.682,
+      "grad_norm": 1.076812505722046,
+      "learning_rate": 2e-05,
+      "loss": 0.04891666,
+      "step": 24841
+    },
+    {
+      "epoch": 49.684,
+      "grad_norm": 1.0280793905258179,
+      "learning_rate": 2e-05,
+      "loss": 0.03525931,
+      "step": 24842
+    },
+    {
+      "epoch": 49.686,
+      "grad_norm": 1.2222983837127686,
+      "learning_rate": 2e-05,
+      "loss": 0.03659509,
+      "step": 24843
+    },
+    {
+      "epoch": 49.688,
+      "grad_norm": 1.1637095212936401,
+      "learning_rate": 2e-05,
+      "loss": 0.04860913,
+      "step": 24844
+    },
+    {
+      "epoch": 49.69,
+      "grad_norm": 1.232571005821228,
+      "learning_rate": 2e-05,
+      "loss": 0.04756819,
+      "step": 24845
+    },
+    {
+      "epoch": 49.692,
+      "grad_norm": 1.7061697244644165,
+      "learning_rate": 2e-05,
+      "loss": 0.05409347,
+      "step": 24846
+    },
+    {
+      "epoch": 49.694,
+      "grad_norm": 3.033412218093872,
+      "learning_rate": 2e-05,
+      "loss": 0.05069696,
+      "step": 24847
+    },
+    {
+      "epoch": 49.696,
+      "grad_norm": 1.3082060813903809,
+      "learning_rate": 2e-05,
+      "loss": 0.06048907,
+      "step": 24848
+    },
+    {
+      "epoch": 49.698,
+      "grad_norm": 1.3506269454956055,
+      "learning_rate": 2e-05,
+      "loss": 0.04115216,
+      "step": 24849
+    },
+    {
+      "epoch": 49.7,
+      "grad_norm": 1.5505048036575317,
+      "learning_rate": 2e-05,
+      "loss": 0.04728535,
+      "step": 24850
+    },
+    {
+      "epoch": 49.702,
+      "grad_norm": 1.154044270515442,
+      "learning_rate": 2e-05,
+      "loss": 0.03792886,
+      "step": 24851
+    },
+    {
+      "epoch": 49.704,
+      "grad_norm": 1.337306261062622,
+      "learning_rate": 2e-05,
+      "loss": 0.0570382,
+      "step": 24852
+    },
+    {
+      "epoch": 49.706,
+      "grad_norm": 1.206878900527954,
+      "learning_rate": 2e-05,
+      "loss": 0.05738255,
+      "step": 24853
+    },
+    {
+      "epoch": 49.708,
+      "grad_norm": 1.0860217809677124,
+      "learning_rate": 2e-05,
+      "loss": 0.04733476,
+      "step": 24854
+    },
+    {
+      "epoch": 49.71,
+      "grad_norm": 1.1992672681808472,
+      "learning_rate": 2e-05,
+      "loss": 0.05342748,
+      "step": 24855
+    },
+    {
+      "epoch": 49.712,
+      "grad_norm": 1.2483594417572021,
+      "learning_rate": 2e-05,
+      "loss": 0.03885884,
+      "step": 24856
+    },
+    {
+      "epoch": 49.714,
+      "grad_norm": 1.1189866065979004,
+      "learning_rate": 2e-05,
+      "loss": 0.03883381,
+      "step": 24857
+    },
+    {
+      "epoch": 49.716,
+      "grad_norm": 3.194491386413574,
+      "learning_rate": 2e-05,
+      "loss": 0.05132719,
+      "step": 24858
+    },
+    {
+      "epoch": 49.718,
+      "grad_norm": 2.8470568656921387,
+      "learning_rate": 2e-05,
+      "loss": 0.04896916,
+      "step": 24859
+    },
+    {
+      "epoch": 49.72,
+      "grad_norm": 1.2007850408554077,
+      "learning_rate": 2e-05,
+      "loss": 0.04390725,
+      "step": 24860
+    },
+    {
+      "epoch": 49.722,
+      "grad_norm": 1.5928409099578857,
+      "learning_rate": 2e-05,
+      "loss": 0.06557049,
+      "step": 24861
+    },
+    {
+      "epoch": 49.724,
+      "grad_norm": 1.5397547483444214,
+      "learning_rate": 2e-05,
+      "loss": 0.05117305,
+      "step": 24862
+    },
+    {
+      "epoch": 49.726,
+      "grad_norm": 1.1637266874313354,
+      "learning_rate": 2e-05,
+      "loss": 0.0384597,
+      "step": 24863
+    },
+    {
+      "epoch": 49.728,
+      "grad_norm": 1.3207765817642212,
+      "learning_rate": 2e-05,
+      "loss": 0.04964681,
+      "step": 24864
+    },
+    {
+      "epoch": 49.73,
+      "grad_norm": 1.3087096214294434,
+      "learning_rate": 2e-05,
+      "loss": 0.05499934,
+      "step": 24865
+    },
+    {
+      "epoch": 49.732,
+      "grad_norm": 1.8449457883834839,
+      "learning_rate": 2e-05,
+      "loss": 0.053078,
+      "step": 24866
+    },
+    {
+      "epoch": 49.734,
+      "grad_norm": 1.1697624921798706,
+      "learning_rate": 2e-05,
+      "loss": 0.04434161,
+      "step": 24867
+    },
+    {
+      "epoch": 49.736,
+      "grad_norm": 1.2733197212219238,
+      "learning_rate": 2e-05,
+      "loss": 0.05949679,
+      "step": 24868
+    },
+    {
+      "epoch": 49.738,
+      "grad_norm": 1.6024494171142578,
+      "learning_rate": 2e-05,
+      "loss": 0.05078848,
+      "step": 24869
+    },
+    {
+      "epoch": 49.74,
+      "grad_norm": 1.3443810939788818,
+      "learning_rate": 2e-05,
+      "loss": 0.04602218,
+      "step": 24870
+    },
+    {
+      "epoch": 49.742,
+      "grad_norm": 1.3979158401489258,
+      "learning_rate": 2e-05,
+      "loss": 0.05351568,
+      "step": 24871
+    },
+    {
+      "epoch": 49.744,
+      "grad_norm": 1.2029260396957397,
+      "learning_rate": 2e-05,
+      "loss": 0.0432038,
+      "step": 24872
+    },
+    {
+      "epoch": 49.746,
+      "grad_norm": 1.2258926630020142,
+      "learning_rate": 2e-05,
+      "loss": 0.05626082,
+      "step": 24873
+    },
+    {
+      "epoch": 49.748,
+      "grad_norm": 1.232932686805725,
+      "learning_rate": 2e-05,
+      "loss": 0.05177718,
+      "step": 24874
+    },
+    {
+      "epoch": 49.75,
+      "grad_norm": 1.1512670516967773,
+      "learning_rate": 2e-05,
+      "loss": 0.04957498,
+      "step": 24875
+    },
+    {
+      "epoch": 49.752,
+      "grad_norm": 1.1977880001068115,
+      "learning_rate": 2e-05,
+      "loss": 0.04884469,
+      "step": 24876
+    },
+    {
+      "epoch": 49.754,
+      "grad_norm": 1.363495111465454,
+      "learning_rate": 2e-05,
+      "loss": 0.04705457,
+      "step": 24877
+    },
+    {
+      "epoch": 49.756,
+      "grad_norm": 1.2528479099273682,
+      "learning_rate": 2e-05,
+      "loss": 0.04862735,
+      "step": 24878
+    },
+    {
+      "epoch": 49.758,
+      "grad_norm": 2.5792670249938965,
+      "learning_rate": 2e-05,
+      "loss": 0.0410705,
+      "step": 24879
+    },
+    {
+      "epoch": 49.76,
+      "grad_norm": 2.7361626625061035,
+      "learning_rate": 2e-05,
+      "loss": 0.05311979,
+      "step": 24880
+    },
+    {
+      "epoch": 49.762,
+      "grad_norm": 2.464141845703125,
+      "learning_rate": 2e-05,
+      "loss": 0.05702599,
+      "step": 24881
+    },
+    {
+      "epoch": 49.764,
+      "grad_norm": 1.099612832069397,
+      "learning_rate": 2e-05,
+      "loss": 0.03931236,
+      "step": 24882
+    },
+    {
+      "epoch": 49.766,
+      "grad_norm": 1.0781171321868896,
+      "learning_rate": 2e-05,
+      "loss": 0.04257463,
+      "step": 24883
+    },
+    {
+      "epoch": 49.768,
+      "grad_norm": 4.444637775421143,
+      "learning_rate": 2e-05,
+      "loss": 0.05754374,
+      "step": 24884
+    },
+    {
+      "epoch": 49.77,
+      "grad_norm": 1.3171765804290771,
+      "learning_rate": 2e-05,
+      "loss": 0.05533032,
+      "step": 24885
+    },
+    {
+      "epoch": 49.772,
+      "grad_norm": 1.1796399354934692,
+      "learning_rate": 2e-05,
+      "loss": 0.03754626,
+      "step": 24886
+    },
+    {
+      "epoch": 49.774,
+      "grad_norm": 1.154632806777954,
+      "learning_rate": 2e-05,
+      "loss": 0.04902484,
+      "step": 24887
+    },
+    {
+      "epoch": 49.776,
+      "grad_norm": 1.1982251405715942,
+      "learning_rate": 2e-05,
+      "loss": 0.04432965,
+      "step": 24888
+    },
+    {
+      "epoch": 49.778,
+      "grad_norm": 1.4120286703109741,
+      "learning_rate": 2e-05,
+      "loss": 0.04318996,
+      "step": 24889
+    },
+    {
+      "epoch": 49.78,
+      "grad_norm": 1.3181231021881104,
+      "learning_rate": 2e-05,
+      "loss": 0.04140037,
+      "step": 24890
+    },
+    {
+      "epoch": 49.782,
+      "grad_norm": 1.1879209280014038,
+      "learning_rate": 2e-05,
+      "loss": 0.04180463,
+      "step": 24891
+    },
+    {
+      "epoch": 49.784,
+      "grad_norm": 1.2485787868499756,
+      "learning_rate": 2e-05,
+      "loss": 0.06122434,
+      "step": 24892
+    },
+    {
+      "epoch": 49.786,
+      "grad_norm": 1.3617092370986938,
+      "learning_rate": 2e-05,
+      "loss": 0.04306863,
+      "step": 24893
+    },
+    {
+      "epoch": 49.788,
+      "grad_norm": 1.1197565793991089,
+      "learning_rate": 2e-05,
+      "loss": 0.04167994,
+      "step": 24894
+    },
+    {
+      "epoch": 49.79,
+      "grad_norm": 1.2155925035476685,
+      "learning_rate": 2e-05,
+      "loss": 0.05426242,
+      "step": 24895
+    },
+    {
+      "epoch": 49.792,
+      "grad_norm": 1.309443712234497,
+      "learning_rate": 2e-05,
+      "loss": 0.05209014,
+      "step": 24896
+    },
+    {
+      "epoch": 49.794,
+      "grad_norm": 2.3679568767547607,
+      "learning_rate": 2e-05,
+      "loss": 0.0490745,
+      "step": 24897
+    },
+    {
+      "epoch": 49.796,
+      "grad_norm": 1.1244512796401978,
+      "learning_rate": 2e-05,
+      "loss": 0.04849419,
+      "step": 24898
+    },
+    {
+      "epoch": 49.798,
+      "grad_norm": 1.0218008756637573,
+      "learning_rate": 2e-05,
+      "loss": 0.04120092,
+      "step": 24899
+    },
+    {
+      "epoch": 49.8,
+      "grad_norm": 1.2458765506744385,
+      "learning_rate": 2e-05,
+      "loss": 0.05435371,
+      "step": 24900
+    },
+    {
+      "epoch": 49.802,
+      "grad_norm": 1.1713382005691528,
+      "learning_rate": 2e-05,
+      "loss": 0.04432591,
+      "step": 24901
+    },
+    {
+      "epoch": 49.804,
+      "grad_norm": 1.043931007385254,
+      "learning_rate": 2e-05,
+      "loss": 0.03500161,
+      "step": 24902
+    },
+    {
+      "epoch": 49.806,
+      "grad_norm": 1.1524564027786255,
+      "learning_rate": 2e-05,
+      "loss": 0.05594937,
+      "step": 24903
+    },
+    {
+      "epoch": 49.808,
+      "grad_norm": 1.2194099426269531,
+      "learning_rate": 2e-05,
+      "loss": 0.06552596,
+      "step": 24904
+    },
+    {
+      "epoch": 49.81,
+      "grad_norm": 1.5245249271392822,
+      "learning_rate": 2e-05,
+      "loss": 0.04009125,
+      "step": 24905
+    },
+    {
+      "epoch": 49.812,
+      "grad_norm": 1.0253095626831055,
+      "learning_rate": 2e-05,
+      "loss": 0.04843939,
+      "step": 24906
+    },
+    {
+      "epoch": 49.814,
+      "grad_norm": 1.9079041481018066,
+      "learning_rate": 2e-05,
+      "loss": 0.05018196,
+      "step": 24907
+    },
+    {
+      "epoch": 49.816,
+      "grad_norm": 1.1070141792297363,
+      "learning_rate": 2e-05,
+      "loss": 0.04317448,
+      "step": 24908
+    },
+    {
+      "epoch": 49.818,
+      "grad_norm": 1.1777336597442627,
+      "learning_rate": 2e-05,
+      "loss": 0.05528118,
+      "step": 24909
+    },
+    {
+      "epoch": 49.82,
+      "grad_norm": 1.162487268447876,
+      "learning_rate": 2e-05,
+      "loss": 0.04349917,
+      "step": 24910
+    },
+    {
+      "epoch": 49.822,
+      "grad_norm": 1.0818195343017578,
+      "learning_rate": 2e-05,
+      "loss": 0.03660588,
+      "step": 24911
+    },
+    {
+      "epoch": 49.824,
+      "grad_norm": 1.178435206413269,
+      "learning_rate": 2e-05,
+      "loss": 0.04270165,
+      "step": 24912
+    },
+    {
+      "epoch": 49.826,
+      "grad_norm": 1.0772615671157837,
+      "learning_rate": 2e-05,
+      "loss": 0.0409967,
+      "step": 24913
+    },
+    {
+      "epoch": 49.828,
+      "grad_norm": 1.5055285692214966,
+      "learning_rate": 2e-05,
+      "loss": 0.0552909,
+      "step": 24914
+    },
+    {
+      "epoch": 49.83,
+      "grad_norm": 1.929326057434082,
+      "learning_rate": 2e-05,
+      "loss": 0.05568858,
+      "step": 24915
+    },
+    {
+      "epoch": 49.832,
+      "grad_norm": 1.478667974472046,
+      "learning_rate": 2e-05,
+      "loss": 0.06444258,
+      "step": 24916
+    },
+    {
+      "epoch": 49.834,
+      "grad_norm": 2.3012428283691406,
+      "learning_rate": 2e-05,
+      "loss": 0.06190757,
+      "step": 24917
+    },
+    {
+      "epoch": 49.836,
+      "grad_norm": 1.908669114112854,
+      "learning_rate": 2e-05,
+      "loss": 0.07899611,
+      "step": 24918
+    },
+    {
+      "epoch": 49.838,
+      "grad_norm": 1.1932661533355713,
+      "learning_rate": 2e-05,
+      "loss": 0.05439082,
+      "step": 24919
+    },
+    {
+      "epoch": 49.84,
+      "grad_norm": 1.1718835830688477,
+      "learning_rate": 2e-05,
+      "loss": 0.04695781,
+      "step": 24920
+    },
+    {
+      "epoch": 49.842,
+      "grad_norm": 1.2461529970169067,
+      "learning_rate": 2e-05,
+      "loss": 0.06060567,
+      "step": 24921
+    },
+    {
+      "epoch": 49.844,
+      "grad_norm": 1.0214687585830688,
+      "learning_rate": 2e-05,
+      "loss": 0.03788566,
+      "step": 24922
+    },
+    {
+      "epoch": 49.846,
+      "grad_norm": 1.10343599319458,
+      "learning_rate": 2e-05,
+      "loss": 0.05159609,
+      "step": 24923
+    },
+    {
+      "epoch": 49.848,
+      "grad_norm": 1.017777919769287,
+      "learning_rate": 2e-05,
+      "loss": 0.0479437,
+      "step": 24924
+    },
+    {
+      "epoch": 49.85,
+      "grad_norm": 1.4957668781280518,
+      "learning_rate": 2e-05,
+      "loss": 0.04876685,
+      "step": 24925
+    },
+    {
+      "epoch": 49.852,
+      "grad_norm": 1.3842114210128784,
+      "learning_rate": 2e-05,
+      "loss": 0.05079174,
+      "step": 24926
+    },
+    {
+      "epoch": 49.854,
+      "grad_norm": 1.3459488153457642,
+      "learning_rate": 2e-05,
+      "loss": 0.05015388,
+      "step": 24927
+    },
+    {
+      "epoch": 49.856,
+      "grad_norm": 1.1398237943649292,
+      "learning_rate": 2e-05,
+      "loss": 0.04445317,
+      "step": 24928
+    },
+    {
+      "epoch": 49.858,
+      "grad_norm": 1.344321370124817,
+      "learning_rate": 2e-05,
+      "loss": 0.05805134,
+      "step": 24929
+    },
+    {
+      "epoch": 49.86,
+      "grad_norm": 1.3307359218597412,
+      "learning_rate": 2e-05,
+      "loss": 0.04095314,
+      "step": 24930
+    },
+    {
+      "epoch": 49.862,
+      "grad_norm": 1.3429630994796753,
+      "learning_rate": 2e-05,
+      "loss": 0.04923171,
+      "step": 24931
+    },
+    {
+      "epoch": 49.864,
+      "grad_norm": 1.1111401319503784,
+      "learning_rate": 2e-05,
+      "loss": 0.05133491,
+      "step": 24932
+    },
+    {
+      "epoch": 49.866,
+      "grad_norm": 1.182901382446289,
+      "learning_rate": 2e-05,
+      "loss": 0.05458375,
+      "step": 24933
+    },
+    {
+      "epoch": 49.868,
+      "grad_norm": 0.9391345977783203,
+      "learning_rate": 2e-05,
+      "loss": 0.03023691,
+      "step": 24934
+    },
+    {
+      "epoch": 49.87,
+      "grad_norm": 2.566356658935547,
+      "learning_rate": 2e-05,
+      "loss": 0.06601861,
+      "step": 24935
+    },
+    {
+      "epoch": 49.872,
+      "grad_norm": 1.2246321439743042,
+      "learning_rate": 2e-05,
+      "loss": 0.05346267,
+      "step": 24936
+    },
+    {
+      "epoch": 49.874,
+      "grad_norm": 1.1111116409301758,
+      "learning_rate": 2e-05,
+      "loss": 0.04999265,
+      "step": 24937
+    },
+    {
+      "epoch": 49.876,
+      "grad_norm": 1.101694941520691,
+      "learning_rate": 2e-05,
+      "loss": 0.04546491,
+      "step": 24938
+    },
+    {
+      "epoch": 49.878,
+      "grad_norm": 1.1672554016113281,
+      "learning_rate": 2e-05,
+      "loss": 0.0529165,
+      "step": 24939
+    },
+    {
+      "epoch": 49.88,
+      "grad_norm": 1.219642996788025,
+      "learning_rate": 2e-05,
+      "loss": 0.03548888,
+      "step": 24940
+    },
+    {
+      "epoch": 49.882,
+      "grad_norm": 1.246626853942871,
+      "learning_rate": 2e-05,
+      "loss": 0.05844071,
+      "step": 24941
+    },
+    {
+      "epoch": 49.884,
+      "grad_norm": 1.2160290479660034,
+      "learning_rate": 2e-05,
+      "loss": 0.05645484,
+      "step": 24942
+    },
+    {
+      "epoch": 49.886,
+      "grad_norm": 2.756503105163574,
+      "learning_rate": 2e-05,
+      "loss": 0.05539309,
+      "step": 24943
+    },
+    {
+      "epoch": 49.888,
+      "grad_norm": 2.0938100814819336,
+      "learning_rate": 2e-05,
+      "loss": 0.04288248,
+      "step": 24944
+    },
+    {
+      "epoch": 49.89,
+      "grad_norm": 1.1077803373336792,
+      "learning_rate": 2e-05,
+      "loss": 0.04419069,
+      "step": 24945
+    },
+    {
+      "epoch": 49.892,
+      "grad_norm": 1.1682178974151611,
+      "learning_rate": 2e-05,
+      "loss": 0.04008082,
+      "step": 24946
+    },
+    {
+      "epoch": 49.894,
+      "grad_norm": 1.7397586107254028,
+      "learning_rate": 2e-05,
+      "loss": 0.04435687,
+      "step": 24947
+    },
+    {
+      "epoch": 49.896,
+      "grad_norm": 0.9368448257446289,
+      "learning_rate": 2e-05,
+      "loss": 0.04006282,
+      "step": 24948
+    },
+    {
+      "epoch": 49.898,
+      "grad_norm": 2.1141977310180664,
+      "learning_rate": 2e-05,
+      "loss": 0.04719079,
+      "step": 24949
+    },
+    {
+      "epoch": 49.9,
+      "grad_norm": 1.1511861085891724,
+      "learning_rate": 2e-05,
+      "loss": 0.03775914,
+      "step": 24950
+    },
+    {
+      "epoch": 49.902,
+      "grad_norm": 1.1330093145370483,
+      "learning_rate": 2e-05,
+      "loss": 0.04169647,
+      "step": 24951
+    },
+    {
+      "epoch": 49.904,
+      "grad_norm": 1.1492831707000732,
+      "learning_rate": 2e-05,
+      "loss": 0.04711498,
+      "step": 24952
+    },
+    {
+      "epoch": 49.906,
+      "grad_norm": 1.6958730220794678,
+      "learning_rate": 2e-05,
+      "loss": 0.05036974,
+      "step": 24953
+    },
+    {
+      "epoch": 49.908,
+      "grad_norm": 1.0818232297897339,
+      "learning_rate": 2e-05,
+      "loss": 0.0415853,
+      "step": 24954
+    },
+    {
+      "epoch": 49.91,
+      "grad_norm": 1.1111513376235962,
+      "learning_rate": 2e-05,
+      "loss": 0.04978892,
+      "step": 24955
+    },
+    {
+      "epoch": 49.912,
+      "grad_norm": 1.2296136617660522,
+      "learning_rate": 2e-05,
+      "loss": 0.04639769,
+      "step": 24956
+    },
+    {
+      "epoch": 49.914,
+      "grad_norm": 1.226378321647644,
+      "learning_rate": 2e-05,
+      "loss": 0.05798329,
+      "step": 24957
+    },
+    {
+      "epoch": 49.916,
+      "grad_norm": 1.0596457719802856,
+      "learning_rate": 2e-05,
+      "loss": 0.04839835,
+      "step": 24958
+    },
+    {
+      "epoch": 49.918,
+      "grad_norm": 1.3038498163223267,
+      "learning_rate": 2e-05,
+      "loss": 0.05488671,
+      "step": 24959
+    },
+    {
+      "epoch": 49.92,
+      "grad_norm": 1.0039633512496948,
+      "learning_rate": 2e-05,
+      "loss": 0.03758329,
+      "step": 24960
+    },
+    {
+      "epoch": 49.922,
+      "grad_norm": 1.0808930397033691,
+      "learning_rate": 2e-05,
+      "loss": 0.03469931,
+      "step": 24961
+    },
+    {
+      "epoch": 49.924,
+      "grad_norm": 1.2433420419692993,
+      "learning_rate": 2e-05,
+      "loss": 0.04504021,
+      "step": 24962
+    },
+    {
+      "epoch": 49.926,
+      "grad_norm": 1.254826545715332,
+      "learning_rate": 2e-05,
+      "loss": 0.05490943,
+      "step": 24963
+    },
+    {
+      "epoch": 49.928,
+      "grad_norm": 0.9941253066062927,
+      "learning_rate": 2e-05,
+      "loss": 0.05116001,
+      "step": 24964
+    },
+    {
+      "epoch": 49.93,
+      "grad_norm": 1.0954303741455078,
+      "learning_rate": 2e-05,
+      "loss": 0.0452432,
+      "step": 24965
+    },
+    {
+      "epoch": 49.932,
+      "grad_norm": 1.0251585245132446,
+      "learning_rate": 2e-05,
+      "loss": 0.04928794,
+      "step": 24966
+    },
+    {
+      "epoch": 49.934,
+      "grad_norm": 1.1478753089904785,
+      "learning_rate": 2e-05,
+      "loss": 0.0425126,
+      "step": 24967
+    },
+    {
+      "epoch": 49.936,
+      "grad_norm": 1.222156047821045,
+      "learning_rate": 2e-05,
+      "loss": 0.04921298,
+      "step": 24968
+    },
+    {
+      "epoch": 49.938,
+      "grad_norm": 1.1650967597961426,
+      "learning_rate": 2e-05,
+      "loss": 0.04439209,
+      "step": 24969
+    },
+    {
+      "epoch": 49.94,
+      "grad_norm": 1.2680879831314087,
+      "learning_rate": 2e-05,
+      "loss": 0.05739405,
+      "step": 24970
+    },
+    {
+      "epoch": 49.942,
+      "grad_norm": 1.1213183403015137,
+      "learning_rate": 2e-05,
+      "loss": 0.05398349,
+      "step": 24971
+    },
+    {
+      "epoch": 49.944,
+      "grad_norm": 1.088508129119873,
+      "learning_rate": 2e-05,
+      "loss": 0.04370025,
+      "step": 24972
+    },
+    {
+      "epoch": 49.946,
+      "grad_norm": 1.314470887184143,
+      "learning_rate": 2e-05,
+      "loss": 0.05276281,
+      "step": 24973
+    },
+    {
+      "epoch": 49.948,
+      "grad_norm": 1.5439000129699707,
+      "learning_rate": 2e-05,
+      "loss": 0.04817521,
+      "step": 24974
+    },
+    {
+      "epoch": 49.95,
+      "grad_norm": 1.5129501819610596,
+      "learning_rate": 2e-05,
+      "loss": 0.05268683,
+      "step": 24975
+    },
+    {
+      "epoch": 49.952,
+      "grad_norm": 1.2225157022476196,
+      "learning_rate": 2e-05,
+      "loss": 0.05818931,
+      "step": 24976
+    },
+    {
+      "epoch": 49.954,
+      "grad_norm": 2.1078433990478516,
+      "learning_rate": 2e-05,
+      "loss": 0.05812836,
+      "step": 24977
+    },
+    {
+      "epoch": 49.956,
+      "grad_norm": 1.2115638256072998,
+      "learning_rate": 2e-05,
+      "loss": 0.04945941,
+      "step": 24978
+    },
+    {
+      "epoch": 49.958,
+      "grad_norm": 1.2102251052856445,
+      "learning_rate": 2e-05,
+      "loss": 0.04883428,
+      "step": 24979
+    },
+    {
+      "epoch": 49.96,
+      "grad_norm": 1.0502766370773315,
+      "learning_rate": 2e-05,
+      "loss": 0.03550196,
+      "step": 24980
+    },
+    {
+      "epoch": 49.962,
+      "grad_norm": 1.2148598432540894,
+      "learning_rate": 2e-05,
+      "loss": 0.048013,
+      "step": 24981
+    },
+    {
+      "epoch": 49.964,
+      "grad_norm": 1.189860224723816,
+      "learning_rate": 2e-05,
+      "loss": 0.06181513,
+      "step": 24982
+    },
+    {
+      "epoch": 49.966,
+      "grad_norm": 1.195906639099121,
+      "learning_rate": 2e-05,
+      "loss": 0.04937382,
+      "step": 24983
+    },
+    {
+      "epoch": 49.968,
+      "grad_norm": 3.3317697048187256,
+      "learning_rate": 2e-05,
+      "loss": 0.05652332,
+      "step": 24984
+    },
+    {
+      "epoch": 49.97,
+      "grad_norm": 0.9888531565666199,
+      "learning_rate": 2e-05,
+      "loss": 0.04144951,
+      "step": 24985
+    },
+    {
+      "epoch": 49.972,
+      "grad_norm": 1.15919828414917,
+      "learning_rate": 2e-05,
+      "loss": 0.04281298,
+      "step": 24986
+    },
+    {
+      "epoch": 49.974,
+      "grad_norm": 1.744892954826355,
+      "learning_rate": 2e-05,
+      "loss": 0.04002345,
+      "step": 24987
+    },
+    {
+      "epoch": 49.976,
+      "grad_norm": 1.2224351167678833,
+      "learning_rate": 2e-05,
+      "loss": 0.04044309,
+      "step": 24988
+    },
+    {
+      "epoch": 49.978,
+      "grad_norm": 1.1010823249816895,
+      "learning_rate": 2e-05,
+      "loss": 0.04040832,
+      "step": 24989
+    },
+    {
+      "epoch": 49.98,
+      "grad_norm": 1.0111849308013916,
+      "learning_rate": 2e-05,
+      "loss": 0.03537663,
+      "step": 24990
+    },
+    {
+      "epoch": 49.982,
+      "grad_norm": 1.0763754844665527,
+      "learning_rate": 2e-05,
+      "loss": 0.0413045,
+      "step": 24991
+    },
+    {
+      "epoch": 49.984,
+      "grad_norm": 1.0886006355285645,
+      "learning_rate": 2e-05,
+      "loss": 0.05110417,
+      "step": 24992
+    },
+    {
+      "epoch": 49.986,
+      "grad_norm": 1.0088766813278198,
+      "learning_rate": 2e-05,
+      "loss": 0.03378244,
+      "step": 24993
+    },
+    {
+      "epoch": 49.988,
+      "grad_norm": 1.1878137588500977,
+      "learning_rate": 2e-05,
+      "loss": 0.05160052,
+      "step": 24994
+    },
+    {
+      "epoch": 49.99,
+      "grad_norm": 1.188334584236145,
+      "learning_rate": 2e-05,
+      "loss": 0.03924968,
+      "step": 24995
+    },
+    {
+      "epoch": 49.992,
+      "grad_norm": 1.1772879362106323,
+      "learning_rate": 2e-05,
+      "loss": 0.05064984,
+      "step": 24996
+    },
+    {
+      "epoch": 49.994,
+      "grad_norm": 0.7895515561103821,
+      "learning_rate": 2e-05,
+      "loss": 0.02828157,
+      "step": 24997
+    },
+    {
+      "epoch": 49.996,
+      "grad_norm": 0.9504234790802002,
+      "learning_rate": 2e-05,
+      "loss": 0.02684129,
+      "step": 24998
+    },
+    {
+      "epoch": 49.998,
+      "grad_norm": 1.4684773683547974,
+      "learning_rate": 2e-05,
+      "loss": 0.05880807,
+      "step": 24999
+    },
+    {
+      "epoch": 50.0,
+      "grad_norm": 1.0322144031524658,
+      "learning_rate": 2e-05,
+      "loss": 0.03352503,
+      "step": 25000
+    },
+    {
+      "epoch": 50.0,
+      "eval_performance": {
+        "AngleClassification_1": 0.998,
+        "AngleClassification_2": 0.998,
+        "AngleClassification_3": 0.9840319361277445,
+        "Equal_1": 1.0,
+        "Equal_2": 0.9840319361277445,
+        "Equal_3": 0.9940119760479041,
+        "LineComparison_1": 1.0,
+        "LineComparison_2": 1.0,
+        "LineComparison_3": 1.0,
+        "Parallel_1": 0.9919839679358717,
+        "Parallel_2": 0.9919839679358717,
+        "Parallel_3": 0.994,
+        "Perpendicular_1": 1.0,
+        "Perpendicular_2": 0.996,
+        "Perpendicular_3": 0.9078156312625251,
+        "PointLiesOnCircle_1": 0.9979959919839679,
+        "PointLiesOnCircle_2": 1.0,
+        "PointLiesOnCircle_3": 0.99,
+        "PointLiesOnLine_1": 1.0,
+        "PointLiesOnLine_2": 0.9959919839679359,
+        "PointLiesOnLine_3": 0.9880239520958084
+      },
+      "eval_runtime": 321.2431,
+      "eval_samples_per_second": 32.686,
+      "eval_steps_per_second": 0.654,
+      "step": 25000
+    },
+    {
+      "epoch": 50.0,
+      "step": 25000,
+      "total_flos": 9.912786476979978e+17,
+      "train_loss": 0.05224685758773238,
+      "train_runtime": 92925.7559,
+      "train_samples_per_second": 17.218,
+      "train_steps_per_second": 0.269
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 25000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 50,
+  "save_steps": 5000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9.912786476979978e+17,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}